Keywords: Model-based Control, Process and Control Monitoring, Process Applications
Abstract: Every ten years or so, Shell has looked to refresh its Advanced Process Control (APC) technology. The last major technology upgrade occurred in 2003 when Shell along, with our APC alliance partner, Yokogawa, released SMOCPro (MPC) and RQEPro (quality estimation). In 2011, Shell and Yokogawa agreed to initiate the development of our next-generation APC technology. Brought to the market in 2015, the Platform for Advanced Control leveraging our long combined experience in APC.

Keywords: Energy Processes and Control, Model-based Control, Optimization and Scheduling
Abstract: Ancillary services constitute the cornerstone of the power grid. They allow for an efficient system operation, provide resilience to uncertainties and establish safeguards against unprecedented events. Their importance is growing due to the rise of grid decentralisation and integration of intermittent, renewable power sources, which lead to more variability and uncertainty in the system. Today, the vast share of ancillary services is provided by large generating units. An ongoing effort by research and business entities focuses on using variation of loads connected to the power grid in order to increase significantly the provision of such services, hopefully at a reduced cost. We examine here, from an economic perspective, the use of commercial buildings as ancillary service providers based on real prices from the Swiss electricity market. We calculate the effect of retail electrical prices on the economic performance of a building and find that for the rates charged in the least expensive cantons a single building can reduce its overall energy costs, when participating in the ancillary services market. For the high end of prices this gradually becomes prohibitive but can be alleviated for a building that has a need for electricity during nighttime hours, as well as daytime. Finally, we show, the counter-intuitive result that providing ancillary services can increase the comfort levels of a building at a decreased cost.

Keywords: Optimization and Scheduling, Process Applications
Abstract: We consider an apparently oversimplified first order model of an anaerobic digester operated as a CSTR, where the dilution rate is the controlled input and the biogas production rate is the measured output. The parameters of this model are considered slowly time-varying. The output function depends on the only state (the substrate), and at any instant has a unique maximum. We propose a simple output-feedback controller based on the super-twisting algorithm combined with a state machine, which converges in a practical sense to this maximum. The controller was tested by simulations of an anarobic digester, maximizing the biogas production rate, showing very good results.

Keywords: Model-based Control, Optimization and Scheduling, Energy Processes and Control
Abstract: In this work, a novel approach to construct the terminal cost for economic model predictive control (EMPC) is developed. In the proposed approach, an auxiliary nonlinear controller which renders the desired optimal steady state asymptotically stable is taken advantage of. The proposed terminal cost is constructed such that it reflects the economic performance of the auxiliary controller over sufficiently large but finite steps. The proposed EMPC design provides guaranteed closed-loop economic performance which does not depend on the length of the prediction horizon. This means that the proposed EMPC algorithm could be very computationally efficient. The proposed EMPC algorithm is applied to an oilsand primary separation vessel. The simulation results demonstrate the effectiveness of the proposed approach.

Keywords: Model-based Control, Process Applications, Optimization and Scheduling
Abstract: The development and application of distributed economic model predictive control (DEMPC) methodologies to a catalytic reactor is considered. Two DEMPC methodologies are designed for sequential and iterative implementation, respectively. The DEMPC architectures are evaluated on the basis of the closed-loop performance and on-line computation time requirements compared to a centralized EMPC approach. For the catalytic reactor considered, DEMPC proves to be a viable option as it is able to give similar closed-loop performance while reducing the on-line computation time requirements relative to a centralized EMPC strategy.

Keywords: Model-based Control, Process Applications
Abstract: Two-stage-riser fluidized catalytic pyrolysis for maximizing propylene yield (TMP) process focuses on propylene production meanwhile without significant losses on gasoline/diesel yields. The complex nonlinear behaviour of this process calls for advanced control system to tackle various disturbances to achieve optimal operation. This paper is devoted to presenting an economics-oriented NMPC scheme that could maximize propylene production while satisfying process operation constraints. In the proposed scheme, those variables, which are difficult to be measured online, including product yields as well as uncertain model parameters are estimated by an unscented transformation based Kalman filter. Potential economic benefits and robustness associated with the proposed controller scheme are illustrated through simulations.

Keywords: Model-based Control, Process Applications
Abstract: Nonlinear Model predictive control (NMPC) is a popular control strategy for highly nonlinear chemical processes. The ability to handle safety and environmental constraints along with the use of an economic objective makes NMPC highly appealing to industries. The performance of NMPC depends strongly on the accuracy of the model. In reality, there always are plant-model mismatch and state estimation errors. Hence the NMPC controller must be robust to uncertainties in the model as well as against estimation errors. Among the several approaches presented in the literature, the scenario-tree based multi-stage NMPC approach is a non-conservative and efficient formulation. In this approach, the evolutions of the plant for different realizations of the uncertainties are considered as different scenarios and the optimization problem is formulated as a multi-stage stochastic programming problem with recourse. In this work, we consider multi-stage output feedback NMPC using the Unscented Kalman Filter (UKF) where the nonlinearities are represented using deterministically chosen sigma points for state estimation. In the control problem, we explicitly consider the UKF estimation equations to predict the future evolution of the system. The proposed approach is illustrated by simulation results of fed-batch chemical reactor with an economic cost function.

Keywords: Model-based Control
Abstract: In this paper, we extend a previously introduced methodology for tube-based robust economic MPC to consider nonlinear average constraints, i.e., constraints on system states and inputs that need to be satisfied on average. A specifically defined integral stage cost takes the disturbance into account when considering the performance. The key idea is to use an appropriately tightened version of the average constraints by using a modified auxiliary output function in their formulation. By means of the tightened constraints, satisfaction of the original average constraints can be guaranteed despite disturbances acting on the system. For some special cases, we provide concepts to simplify the tightening (which might in general be difficult to determine). In addition, we discuss how average constraints can be used in order to enforce convergence of the closed-loop system to an invariant set. Finally, the proposed approach is illustrated with a numerical example.

Keywords: Model-based Control, Optimization and Scheduling, Process Applications
Abstract: Many processes are influenced by uncertain parameters or external disturbances, such as temperature changes. The control of such systems is in general challenging. In this work, we consider robust multi-scenario Model Predictive Control (MPC). Its central idea is to assume a finite number of possible values for the uncertainties and to model their combinations in a scenario tree. We adapt the classical dual mode approach of nominal MPC to establish recursive feasibility and stability for the multi-scenario case, using a common terminal region and common terminal cost function for all uncertainty realizations. For linear systems, the computation of these ingredients can be formulated as a semidefinite program. In a simulation, we apply the suggested approach to building climate control and show that it robustly stabilizes the system while a standard MPC controller violates state constraints and becomes infeasible.

Keywords: Optimization and Scheduling, Process Applications
Abstract: Well placement optimization aims to determine optimal well locations so that the economic benefit from oil production can be maximized. Geological uncertainty has a significant impact on the optimal well placement plan and therefore has to be considered in the well placement optimization problem. A geological realization reduction framework for well placement under geological uncertainty is proposed in this work. The objective is to optimally select a small subset of realizations and incorporate them into the well placement optimization problem, so as to reduce the computational efforts. A reservoir case study demonstrates that the selected smaller subset of realizations is a very good representation of a larger superset of realizations and can significantly decrease the computational time associated with the well placement optimization problem.

Keywords: Energy Processes and Control, Model-based Control, Process Applications
Abstract: This paper presents a successful application of soft sensor based control for froth pipeline water content in oil sands industry. Water content is a key process quality variable for inter-site froth transportation through pipeline, which should be controlled within a specific range by adding hot processing water into the pipeline to maintain a reliable operating condition. A dynamic soft sensor incorporated with on-line bias update is developed to estimate the water content in real-time. Based on the soft sensor, a control strategy is consequently implemented to control the froth pipeline water content. Significant improvement of water content control is obtained from the successful application to a froth pipeline, which demonstrates the effectiveness, reliability and advantage of the soft sensor and control proposed.

Keywords: Modeling and Identification
Abstract: A distributed drift-flux model and a low-order lumped model describing a multiphase (gas-liquid) flow in the well during Under-Balanced Drilling (UBD) has been presented. This paper presents a novel nonlinear adaptive observer to estimate the total mass of gas and liquid in the annulus and production constant of gas and liquid from the reservoir into the well during UBD operations. Furthermore, it describes a joint unscented Kalman filter to estimate parameters and states for both the distributed drift-flux and lumped model by using real-time measurements of the choke and the bottom-hole pressures. The performance of the adaptive observers are evaluated for typical drilling scenarios. The results show that all adaptive observers are capable of identifying the production index, although the adaptive observers based on the low-order lumped model achieves better convergence rate than adaptive observer based on the drift-flux model. The results show that the LOL model is sufficient for the purpose of estimating the production parameters.

Keywords: Process Applications
Abstract: While most of the available Leak Detection Systems (LDS) can detect pipeline leaks, leak localization is still an unresolved problem. The main reason for this problem is the limited number of sensors installed in long pipelines. Because of lack of measurements, precise leak location cannot be easily determined. This study uses particle filter as soft-sensor to estimate the states at intermediate locations with the available end-point measurements. The residuals between the estimated states and the real measurements at the intermediate locations are then used for leak detection and localization. The proposed method can improve the leak localization accuracy by proper use of the intermediate pressure measurements. Simulation results show the efficacy of the proposed method.

Keywords: Process and Control Monitoring, Modeling and Identification, Process Applications
Abstract: Utilizing flow rate and pressure data in and out of the mud circulation loop provides a driller with real-time trends for the early detection of well-control problems that impact the drilling efficiency. This paper presents state estimation for infinite-dimensional systems used in the process monitoring of oil well drilling. The objective is to monitor the key process variables associated with process safety by designing a model-based nonlinear observer that directly utilizes the available information coming from the continuous-time online process output measurements at the topside of the well. The observer consists of a copy of the plant plus output injection terms where the gain is computed analytically in terms of the Bessel function of the first kind. The design is tested using data from a real-field drilling commissioning test in the North Sea by Statoil Oil Company. The results show that the design estimates the flow and pressure accurately.

Keywords: Modeling and Identification, Process Applications
Abstract: Uncertainty in data or in the parameters of models occurs in many real world applications. Quantifying this uncertainty and its effects is required for robust design, control and optimization. In this paper, we attempt to build a proxy model for the stochastic solutions of coupled governing equations describing coalbed methane (CBM) production at different well bottomhole pressures. To achieve this, monthly production from wells (output) is expanded as a linear combination of Legendre orthogonal polynomials in the input (well bottomhole pressure) and the Wiener-Askey polynomial chaos is used to propagate the uncertainty of the model parameters. A Gaussian quadrature technique is then employed to solve for the coefficients of the basis functions in the proxy model. Alternatively, nonlinear least squares curve fitting using the Levenberg-Marquardt algorithm (LMA) is also used with polynomial chaos expansion to generate the stochastic proxy model. The proxy model now enables robust optimization using statistical metrics of CBM production calculated over the entire parameter space. In the case of multiple decision variables, the appropriate proxy model built using these techniques will allow for robust optimization without the use of any search algorithms.

Keywords: Modeling and Identification, Optimization and Scheduling, Energy Processes and Control
Abstract: The computation of enclosures for the reachable set of uncertain dynamic systems is a crucial component in a wide variety of applications, from global and robust dynamic optimization to safety verification and fault detection. Even though many systems in engineering are best modeled as implicit differential equations (IDEs) and differential algebraic equations (DAEs), methods for the construction of enclosures for these are not as well developed as they are for ordinary differential equations (ODEs). In this paper, we propose a continuous-time approach for the guaranteed over approximations of the reachable set for quasilinear IDEs. This approach builds on novel high-order inclusion techniques for the solution set of algebraic equations and state-of-the-art techniques for bounding the solution of nonlinear ODEs. We show how this approach can be used to bound the reachable set of uncertain semi-explicit DAEs by bounding the underlying IDEs. We demonstrate this approach on two case studies, a double pendulum where it proves superior with delayed break-down times compared to other methods, and anaerobic digestion of microalgae which has nine differential and two algebraic states.

Keywords: Modeling and Identification, Optimization and Scheduling, Process Applications
Abstract: Accurate estimation of parameters is paramount in developing high-fidelity models for complex dynamical systems. Model-based optimal experiment design (OED) approaches enable systematic design of experiments to generate input-output data sets with high information content for parameter estimation. Standard OED approaches however face two challenges: (i) experiment design under incomplete system information due to unknown true parameters, which usually requires many iterations of OED; and (ii) incapability of systematically accounting for the inherent uncertainties of complex systems, which can lead to diminished effectiveness of the designed optimal excitation signal as well as violation of system constraints. This paper presents a robust OED approach for nonlinear systems with arbitrarily-shaped time-invariant probabilistic uncertainties. Polynomial chaos is used for efficient uncertainty propagation. The distinct feature of the robust OED approach is the inclusion of chance constraints to ensure constraint satisfaction in a stochastic setting. The presented approach is demonstrated by optimal experimental design for the JAK-STAT5 signaling pathway, which regulates various cellular processes in a biological cell.

Keywords: Optimization and Scheduling, Process Applications, Batch Process Modeling and Control
Abstract: Process models that are affected by uncertainties need a robust mechanism to account for them in the model based design of experiments (DOE). The aim of this study is to design a set of experiments to estimate the parameters of multiscale kinetic models for the catalytic decomposition of ammonia. Along with uncertainties in the model, the problem is challenging due to constraints on experimental conditions. A stochastic D-optimal design is used to find the optimal experimental conditions using maximization of the expectation of properties of the Fisher information matrix (FIM). The expectation of FIM is calculated by sample average approximation (SAA) based on Monte Carlo simulations. Particle swarm optimization (PSO) is used to perform stochastic optimization to find the optimal set of experimental conditions. A novel method based on the rescaling of velocities is proposed for handling of equality and inequality constraints in particle swarm optimization.

Keywords: Modeling and Identification
Abstract: Model migration has been proved to be an effective modeling tool to adopt an existing base model from an old process to a similar, yet non-identical process. However, if the process differences are more complex and differ from sample to sample, then the existing model migration strategies can be non- flexible and inadequate. Based on the concepts laid out in an earlier article (Lu and Gao (2008b)), this paper presents an enhanced Bayesian model migration strategy for statistical models. This is achieved by applying Bayesian adjustments to a base model developed using the Gaussian process (GP). The benefits of the proposed method are demonstrated on a continuously stirred tank reactor.

Keywords: Modeling and Identification
Abstract: Partial least squares(PLS) regression has been widely used to capture the relationship between inputs and outputs in static system modeling. Several dynamic PLS algorithms were proposed to capture the characteristic of dynamic systems. However, none of these algorithms provides an explicit description for dynamic inner model and outer model. In this paper, a dynamic inner PLS is proposed for dynamic system modelling. The proposed algorithm gives explicit dynamic inner model and makes inner model and outer model consistent at the same time. Several examples are given to show the effectiveness of the proposed algorithm.

Keywords: Modeling and Identification, Process Applications
Abstract: In the present work, a new input variable selection method for building linear regression models is proposed. The proposed method is referred to as nearest correlation Louvain method based variable selection (NCLM-VS). NCLM-VS is a correlation-based group-wise method; it constructs an affinity matrix of input variables by the nearest correlation (NC) method, partitions the affinity matrix by the Louvain method (LM), consequently clusters input variables into multiple variable classes, and finally selects variable classes according to their contribution to estimates. LM is very fast and optimizes the number of classes automatically unlike spectral clustering (SC). The advantage of NCLM-VS over conventional methods including NCSC-VS is demonstrated through their applications to soft-sensor design for an industrial chemical process and calibration modeling based on near-infrared (NIR) spectra. In particular, it is confirmed that NCLM-VS is significantly faster than the recently proposed NCSC-VS while NCLM-VS can achieve as good estimation performance as NCSC-VS.

Keywords: Process and Control Monitoring, Process Applications
Abstract: The objective of this paper is to address the problem of state estimation in an Anaerobic Digestion Reactor (ADR) with unknown inputs - typically some influent concentrations, and to compare two different state and input estimation schemes. The first one is the classical Extended Kalman Filter (EKF) based on an augmented system that considers a slowly varying input, whereas the second one is a recently proposed Unknown Input Observer (UIO), formulated in the spirit of a Kalman Filter, for continuous estimation with discrete measurements. The two filters are evaluated in simulation, demonstrating the superiority of the UIO.

Keywords: Modeling and Identification, Optimization and Scheduling, Process Applications
Abstract: This paper presents a new approach to input design for closed-loop identification. The idea is to maximize the trace of the Fisher information matrix associated with the plant model, while enforcing explicit constraints on both inputs and outputs. The result is the richest possible excitation signal for which the operation of a running closed-loop system remains within acceptable bounds. The function to be maximized is a convex quadratic. A Moving Horizon Predictive (MHP) framework is used to solve the input design problem at each sample time. The method can be combined with a fixed model variable regressor technique to estimate time delays. The suggested technique is implemented and used to identify machine-directional processes in an industrial paper machine.

Keywords: Model-based Control, Modeling and Identification, Process and Control Monitoring
Abstract: This paper considers observer design for nonlinear dynamical systems which can be approximated by a dissipative Hamiltonian realization. The design approach decomposes the system associated one-form of a given dynamical system over an indeterminate metric using an homotopy operator to generate exact (potential driven) and anti-exact parts. Then the convexity of the potential system given by the exact part is assessed and we propose a metric equation which yields a Lyapunov function for the potential driven observer system. An application of this method is demonstrated for a two-dimensional van der Pol oscillator.

Keywords: Modeling and Identification
Abstract: A Bayesian algorithm is developed for estimating parameters in nonlinear stochastic differential equation (SDE) models. The proposed algorithm uses prior information about parameters and builds on the approximate expectation maximization (AEM) algorithm (Karimi and McAuley, 2014a). A nonlinear continuous stirred tank reactor (CSTR) model is used to compare the effectiveness of the Bayesian algorithm to that of the AEM algorithm. For the CSTR example studied, the proposed method provides more accurate parameter estimates, especially for small data sets.

Keywords: Model-based Control
Abstract: This paper considers the optimal strategies for constrained linear state estimation. Prior information for estimating state variables is often available in the form of inequality constraints on states. In the latest developments of optimal state estimation theory consideration of state constraints has been often neglected since constraints do not fit easily in the structure of the optimal filter, for example, the issue of state constraints being present has to be addressed adequately, for example nonnegativity of concentration. In order to address this issue and to extend previous developments on the accuracy of state estimation, this work develops the constrained optimal state estimation for finite-dimensional systems. Finally, a numerical example illustrating the proposed method is presented.

Keywords: Model-based Control, Process and Control Monitoring, Process Applications
Abstract: In this work, we propose a distributed adaptive high-gain extended Kalman filtering approach for nonlinear systems. Specifically, we consider a class of nonlinear systems that are composed of several subsystems interacting with each other via their states. In the proposed approach, an adaptive high-gain extended Kalman filter is designed for each subsystem. The distributed Kalman filters communicate with each other to exchange subsystem state estimates. First, an implementation strategy which specifies how the distributed filters should communicate is designed. Second, the detailed design of the subsystem filter is described. Subsequently, the stability of the proposed distributed state estimation is analyzed. Finally, the effectiveness and applicability of the proposed design are illustrated via the application to a chemical process example.

Keywords: Model-based Control, Process Applications
Abstract: Until recently, computational complexities and lack of detailed, nonlinear dynamic models have stood as main obstacles to widespread industrial adoption of model-based on-line optimization for operation of batch and semi-batch reactors. With recent advances in both dynamic modeling techniques and nonlinear programming (NLP) solvers, it is conceivable now to use significantly-sized, nonlinear models directly for on-line state/parameter estimation and optimal control calculations. In this study, we propose a framework for doing this. In the proposed framework, nonlinear first principles dynamic model-based optimizations are performed at several time points throughout a batch run over a fixed horizon, in order to estimate the current state of the model and to refine the target batch time and input variables. Here we combine shrinking horizon nonlinear model predictive control (sh-NMPC) with expanding horizon least squares estimation (eh-LSE). This framework is tested on a large-scale anionic propylene oxide (PO) polymerization process, whose operation considers not only certain end-product specifications but also safety constraints. It is shown that the proposed method is not only computationally feasible (averaging less than 10 CPU seconds at each sampling time) but leads to excellent performance, satisfying the product specification target despite initialization errors and measurement noise.

Keywords: Model-based Control, Optimization and Scheduling, Modeling and Identification
Abstract: It is well known that certain properties of the process dynamics can be deduced from steady-state information about a process only. In this paper we consider the dual problem, that of determining steady-state properties from process dynamics. In particular, we are concerned with the problem of determining extremum points in the steady-state input-output map from dynamic response data. This is a highly relevant problem in cases where the aim is to determine steady-state optimal operating conditions using real time process measurements. For this purpose, we first consider the connection between bifurcations of the zero dynamics and the steady-state input-output map. Based on these results, we show that steady-state optimal conditions can be determined from the process dynamics through consideration of local phase-lag properties of the process only. We demonstrate the usefulness of this result by showing that the optimum of a chemical reactor can be located, without any prior knowledge, using sinusoidal perturbations and a phase-lock loop.

Keywords: Optimization and Scheduling
Abstract: This paper deals with the real-time optimization of uncertain plants and proposes an approach based on surrogate models to reach the plant optimum when the plant cost gradient is imperfectly known. It is shown that, for processes with only box constraints, the optimum is reached upon convergence if the multiplicative gradient uncertainty lies within some bounded interval. For the case of general constraints, conditions are derived that guarantee plant feasibility and, in principle, allow enforcing cost decrease at each iteration.

Keywords: Optimization and Scheduling, Model-based Control
Abstract: In model-based real-time optimization, plant-model mismatch can be handled by applying gradient- and bias-corrections to the cost and constraint functions in an iterative optimization procedure. One of the major challenges in practice is the estimation of the plant gradients from noisy measurement data, in particular for the case of several optimization variables. In this paper we compare four modifier adaptation schemes which were proposed to handle noisy data, iterative gradient-modification optimization, dual modifier adaptation, nested modifier adaptation and modifier adaptation with quadratic approximation. Simulation studies for the optimization of the Otto-Williams reactor with plant-model mismatch are used to illustrate the performance of the different schemes.

Keywords: Optimization and Scheduling, Process Applications, Model-based Control
Abstract: The integration of process design and control (IPDC) has become increasingly important and received considerable academic and practitioner attention over the recent decades. Moreover, some works in the field have utilized advanced controllers into IPDC framework. Considering the fact that the hierarchical control structure which consists of local PID controllers and an advanced process control (APC) as the central controller has been used widely in industrial practice, in this paper, the two-level hierarchical control structure is taken into account within the IPDC framework for the first time. The IPDC problem using hierarchical control structure is formulated as a bi-level optimization problem, which separates the design decisions from the APC decisions and then keeps the problem size manageable. And the APC decisions are incorporated in a superstructure-based dynamic optimization formulation which considers both cost and controllability. A continuous stirred tank reactor is used to test the proposed IPDC methodology.

Keywords: Model-based Control, Process and Control Monitoring, Modeling and Identification
Abstract: Plug-flow crystallization (PFC) is a promising continuous pharmaceutical crystallization process but is prone to fouling due to uncontrolled crystallization on the reactor surface (encrustation). This results in operational issues such as (1) flow blockage, (2) increased thermal resistance, and (3) reduced supersaturation, and in turn leads to limited continuous operation and reduced crystal quality and yield. In this work, we introduce a model-based anti-fouling control (AFC) via spatial and temporal heating and cooling cycle.

Keywords: Energy Processes and Control, Process Applications, Process and Control Monitoring
Abstract: This article presents an overview of the principle, engineering implementation and applications of spectroscopic measurements technology, such as NIR (near infrared), FTIR (Fourier transform infrared), Raman and NMR (nuclear magnetic resonance). The discussions are focused on oil sands industry in which the process majorly consists of the mining, slurry hydro-transport, extraction, froth treatment and upgrading plants.

Keywords: Energy Processes and Control, Optimization and Scheduling
Abstract: In this paper, the problem of optimal crude oil procurement combined with refinery operations is addressed to obtain an epsilon-global optimal solution. Rather than the traditional planning and scheduling methodologies relying on linear programming (LP), a nonlinear model using Geddes fractional index (FI) is employed to describe the behavior of the crude distillation unit (CDU) and integrated with the entire plant-wide model. Although this representation provides more accurate prediction of the real production of the refinery than the conventional fixed yield approach, its global optimization becomes more difficult owing to the existence of many nonlinear, non-convex terms. To overcome this challenge, advanced interval reduction techniques are developed and combined with state-of-the-art global optimization software to obtain an epsilon-global optimal solution more efficiently. The optimization and comparison are conducted to show the effectiveness of the proposed approach.

Keywords: Process and Control Monitoring, Optimization and Scheduling
Abstract: Most of the previous studies on the required critical minimum velocity to move the solid bed inside the slurry pipelines are related to the design step where the dynamics of the process are not thoroughly considered. In this paper, a general framework for on-line estimation of the critical minimum velocity is proposed and applied to the underflow stream of an important Primary Separation Vessel (PSV) in an oil sand industry. Appropriate statistical and probabilistic models are used to improve the on-line measurements and estimations. The proposed method demonstrates a satisfactory performance in detection of different conditions of the PSV operations.

Keywords: Optimization and Scheduling, Energy Processes and Control
Abstract: A key challenge in production optimization is handling of model uncertainty. Traditionally, production optimization is done in a deterministic setting, ignoring the uncertainty. In this work, we formulate the problem as a two-stage stochastic programming problem. The solution to the problem is a strategy for operating the wells, instead of a single setpoint obtained from the deterministic problem. This strategy is easy to follow for the operator. A synthetic case study shows how the proposed approach increases the expected oil production by 1.5 percent.

Keywords: Optimization and Scheduling, Process and Control Monitoring, Process Applications
Abstract: Optimal process operation is not always guaranteed due to the presence of significant uncertainty about the plant models that are used to make decisions, for example in Real time optimization, and also due to the differences between control architecture layers which operate on different time-scales and use different kind of models. Modifier adaptation is a methodology that achieves optimality despite the presence of uncertainty by using plant measurements. This paper presents the Nested modifier-adaptation methodology applied to the operation of distillation columns and the results obtained are compared with the previous modifier adaptation methodology using dual control optimization.

Keywords: Optimization and Scheduling
Abstract: This work focuses on making the best possible decision at the RTO level, when it is not economically viable to have implemented a full Production scheduling and business planning optimization. It attempts to merge some of the longer-term decisions that are done in the production scheduling and inventory management into the RTO, thereby minimizing the total cost of implementations while attempting to get some of the benefits that a full production/inventory scheduling activity would bring. In the current work a decision on inventory levels is done within RTO by solving the optimization problem over a longer horizon and by augmenting the objective function for RTO with inventory cost based on historical average of marginal cost. The objective function in RTO is based on minimization of costs, and minimization of the proposed objective function leads to an overall reduction of long term marginal cost. A case study is presented in which average marginal cost is considered greater and lower than the current cost of production and shows that the long term marginal cost reduces over a period of time.

Keywords: Modeling and Identification
Abstract: Soft sensors based on dynamic PLS (DPLS) have been widely used in industrial applications for predicting hard-to-measure quality variables. However, DPLS is prone to over-fitting due to an increasing number of model inputs. A plethora of approaches have been proposed to improve DPLS-based soft sensors, among which variable selection has been a prevailing one. Recently, a new method termed as DPLS-TS has been proposed to penalize dynamic parameters in DPLS using a temporal smoothness regularization, which helps reduce model complexity and deliver smooth predictions for quality variables. In this work we present a comparative study of temporal smoothness regularization and variable selection in terms of their improvements in prediction performance when a large number of lagged time series data are involved. Comparisons are performed through a simulated case of crude distillation unit.

Keywords: Batch Process Modeling and Control, Model-based Control, Optimization and Scheduling
Abstract: We propose an approximate model in form of a system of ordinary differential equations (ODE) for a class of first-order multivariate linear partial differential equations (PDE) of hyperbolic type. The resulting closed ODE-system utilizes least-square approximations over orthogonal polynomial bases for specific factors of the PDE. The class of examined PDEs is typical in population balance systems with fines removal. The proposed modeling scheme is of interest for control and optimization of multivariate systems with distributed parameters.

Keywords: Modeling and Identification
Abstract: For solving some process control engineering problems which can be treated as a time-series, a fast and accurate self-organization learning strategy is proposed based on the significance evaluation of hidden neurons with respect to the network output. This approach is introduced to optimize the architecture and parameters of span-lateral inhibition neural network (S-LINN) simultaneously. The insignificant neuron(s) will be pruned automated step by step based on the determination of significance index. The proposed self-organizing approach has been tested on one time-series prediction benchmark problem. Simulation results demonstrate that the proposed method has good exploration and exploitation capabilities in terms of searching the optimal structure and parameters for S-LINN.

Keywords: Modeling and Identification, Process Applications
Abstract: A state-space realization approach is presented for the identification of time-delay dynamic systems. It is proposed an experiment with low complexity input signals such as the double pulse. The proposed identification method is a generalization of the Ho-Kalman-Kung technique for pulsed input signals. Due to its state-space formulation, it is essentially suitable to both SISO and MIMO systems.

Keywords: Modeling and Identification, Process and Control Monitoring, Process Applications
Abstract: Multivariate calibration, classification and fault detection models are ubiquitous in QbD (Quality by Design) and PAC and PAT (Process Analytical Chemistry and Technology, respectively) applications. They occur in the both the development of processes and their permissible operating limits, (i.e. models for relating the process design space to product quality), and in manufacturing (i.e. models used in monitoring and control). Model maintenance is the ongoing servicing of these multivariate models in order to preserve their predictive abilities. It is required because of changes to either the sample matrices or the instrument or response. The goal of model maintenance is to sustain or improve models over time and changing conditions with the least amount of cost and effort. This paper presents a roadmap for determining when model maintenance is required, the probable source of the response variations, and the appropriate approaches for achieving it. Methods for evaluating model robustness in order to identify models with lower ongoing maintenance costs are also discussed.

Keywords: Modeling and Identification, Batch Process Modeling and Control, Process Applications
Abstract: A transformation to variant and invariant states, called extents, is used to decouple the dynamic effects of reaction systems and serves as basis for incremental model identification, in which kinetic models are identified individually for each dynamic effect. This contribution introduces a novel transformation to extents for the incremental model identification of two-phase distributed reaction systems. Distributed reaction systems are discussed for two cases, namely, when measurements along the spatial coordinate are available and when there are not. In the second case, several measurements made under appropriate operating conditions are combined to overcome the lack of measurements along the spatial coordinate. This novel method is illustrated via the simulated example of a two-phase tubular reactor.

Keywords: Batch Process Modeling and Control, Process Applications, Optimization and Scheduling
Abstract: Process optimization is an important issue for raising product quality and ensuring safety in batch processes and it is usually conducted at the point only when set-points are reached. In this work, process dynamic shift results from set-point tracking control, termed as transition tracking process, is detailedly analyzed in an integrated optimization framework, which also synthesizes the multi-stage characteristic of batch processes. Establishment of the regression relationship between process variables and quality indexes becomes possible and gradient directions are updated iteratively.The effacy of the proposed scheme is illustrated on the injection molding, which is a typical multi-stage batch process.

Keywords: Batch Process Modeling and Control, Process and Control Monitoring, Process Applications
Abstract: This work proposes a multiscale model-based operation framework for large-area plasma-enhanced chemical vapor deposition (PECVD) of thin film silicon solar cells with uniform thickness and film surface microstructure that optimizes light trapping. The results of a multiscale process model indicate that in order to manufacture a thin film with a diameter of 20 cm and a uniform thickness with surface microstructure that optimizes light trapping: a) a sinusoidally-grated wafer surface should be used in which the grating period and depth should correspond to values that lead to film surface roughness on the order of visible light wavelength range, and b) the substrate temperature should be regulated, along several concentric zones across the substrate, to compensate for a radially non-uniform deposition rate of the film on the wafer owing to gas-phase transport phenomena. Due to the dependence of film growth rate on substrate temperature, the wafer surface is separated into four concentric zones, each with an independent heating element. Simulations demonstrate that the use of appropriate sinusoidal wafer grating and the regulation of substrate temperature provide a viable and effective way for the large-area PECVD of thin film silicon solar cells with uniform thickness and film surface microstructure that optimizes light trapping.

Keywords: Batch Process Modeling and Control, Process Applications, Model-based Control
Abstract: In this contribution, the effects of different degrees of uncertainty description are investigated experimentally using an exothermic chemical reaction with safety constraint on the temperature. For that purpose, two robust trajectories are designed that respect the artificially created uncertainties of the experiments either coarsely using a single multivariate normal distribution (1GMD) or in a more detailed fashion using a Gaussian mixture density (GMD) consisting of 32 multivariate normal densities (32GMD). For the optimization, the uncertainties are propagated using the unscented transformation. Both trajectories were run 71 times in an open-loop manner. The more detailed trajectory (32GMD) leads to a 9% higher yield without increasing the risk of constraint violation. Furthermore, many experimental realizations of two robust closed-loop process control schemes are being compared. They differ again only in the degree of the underlying uncertainty description. Although the frequent corrections of the controller marginalize the advantage of a more detailed stochastic process prediction, the 4GMD-controller still allows for 3% more educt conversion compared to the 1GMD-controller.

Keywords: Batch Process Modeling and Control, Modeling and Identification
Abstract: A novel data-driven control performance assessment (CPA) method is proposed for batch processes controlled by iterative learning control (ILC) based on two-dimensional linear quadratic Gaussian (LQG) benchmark. Previous studies on CPA for ILC are based on an assumption that the model of the controlled batch process is known, whereas this study proposes a model-free CPA method. Based on the two-dimensional system theory, the closed-loop batch process under ILC can be converted into a two-dimensional Roesser model. This study proposes a novel closed-loop two-dimensional subspace identification method for the converted parameters unknown two-dimensional Roesser model. Using the identified model, the two-dimensional LQG tradeoff performance assessment surface can be obtained. The proposed method is verified by performing some simulations.

Keywords: Batch Process Modeling and Control, Model-based Control
Abstract: A two-stage cooling protocol for batch cooling crystallization was optimized, to maximize mean crystal size while minimizing the variance in size. A population balance model represents the crystallization process, incorporating crystal growth and nucleation, for use in a multi-objective optimization. At all points along the Pareto front, the optimal strategy has the nucleation event occurring on the temperature plateau between the two cooling stages. This shortens the nucleation period-supersaturation is rapidly depleted, ultimately generating fewer crystals of larger size. In comparison to a simple linear cooling profile, the two-step strategy produces crystals of larger size with smaller variance.

Keywords: Batch Process Modeling and Control, Model-based Control, Process Applications
Abstract: A process optimization method based on partial least squares (PLS) has been used in pharmaceutical processes. However, its applicability and performance are limited because PLS cannot cope with nonlinearity and changes in process characteristics. In this research, a new process optimization method based on locally weighted PLS (LW-PLS) is proposed. To solve a nonlinear optimization problem based on LW-PLS, in which any global model is not constructed, self-adaptive differential evolution (jDE) is adopted. The validity of the proposed method is demonstrated through a numerical example and an industrial case study of a pharmaceutical granulation process.

Keywords: Modeling and Identification
Abstract: In this paper, a closer look at the underlying computer code of the well-known Tennessee Eastman Process model is taken. Since its introduction in the 1990s typical simulation software, e.g. MATLAB®, which is used to simulate the process model, has changed. Thus the original program flow devised by Downs & Vogel no longer holds. This results in problems regarding the repeatability of simulation results. This problem and its cause are considered in the following and a solution in terms of a modified code is presented. Furthermore, some additional changes are discussed, widening the useability of the simulation model (e.g. lower simulation runtime, additional process measurements).

Keywords: Modeling and Identification
Abstract: A detailed pseudo-bulk model has been used for prediction of conversion and particle size distribution (PSD) of vinyl acetate in a continuous emulsion polymerization reactor. Finite volume (FV) and moment techniques are applied for solving population balance equation under continuous operation. It is found that both methods can predict sustained oscillations in the monomer conversion, however the FV method matches the experimental data better than the moment method. The monomer conversion and free surfactant concentration are controlled via two single control loops. In this work, a new control strategy for controlling monomer conversion has been proposed. It has been shown that monomer conversion and free surfactant concentration can be controlled simultaneously by manipulating initiator and surfactant feed rates.

Keywords: Modeling and Identification, Process Applications, Batch Process Modeling and Control
Abstract: Taylor-Couette reactor (TCR) consists of two concentric cylinders with small gap and typically the inner cylinder is rotating. In the concentric gap of the reactor, special patterns of flow are generated. Taylor-Couette flow has a Taylor vortex that is characterized by axisymmetric toroidal vortices. This phenomenon was first discovered by Taylor in 1923. Recently, the Taylor-Couette reactor has been used as a crystallizer and phase-transformation device. For example, the phase transformation of Guanosine 5-monophosphate (GMP) investigated by using the drowning-out crystallization method. Also, the Taylor-Couette reactor is used as a crystallization device experimentally and numerically, in which the fine particles are classified. In our previous work, the dynamic characteristics of Taylor-Couette flow and the possibility of particle classification were investigated numerically and experimentally for the case of vertical TCR in a continuous system. However, none of these studies involves the crystallization modeling due to the difficulty of simultaneous modeling of fluid dynamics and PBE.

In the present study, a CFD-PBE model is developed for verifying the performance of continuous TCR as crystallizer and classifier. A saturated ammonium sulfate solution is used as the sample substance. This CFD-PBE model is based on the Eulerian multi-phase model to describe the liquid-solid two phase flow and the discrete method PBE model including nucleation and growth kinetics. Since the PSD of TCR is much narrower than CSTR, TCR could be adjusted as crystallizer for seed manufacturing.

Keywords: Model-based Control, Process Applications
Abstract: This work aims to improve corporate functional departments' confidence in adopting modern control approaches in new scenarios and thus presents control structure solutions based on MPC for two control problems facing existing upstream oil & gas production plants; these are the disturbance growth in the series connected process and the control system dependency on operators. The proposed approach integrates distributed MPC (DMPC) as a master controller for the existing classical control of each subsystem, with a focus on those with high interaction phenomena. The proposed DMPC also considers safeguarding, constraints and the enhancement of plant-wide optimal performance. The suggested control solution reduces the role of control room operators which is shown to reduce the growth in the impact of process disturbances. Compared with some alternative control structures (centralised MPC, decentralised MPC, distributed MPC (DMPC), and hierarchical DMPC) this proposal is simple, inexpensive to implement, and critically, builds on the local team operational experience and maintenance skills.

Keywords: Optimization and Scheduling, Batch Process Modeling and Control, Model-based Control
Abstract: To enhance convergence property of Genetic Algorithm (GA), we in this work propose modification in GA by combining the global search property of GA with a convergence property of Box-Complex method. Using the current population of GA, new members are created using Box-Complex concept, which replaces equal number of worst population members. A comparative study of the proposed GA with the conventional GA and widely accepted Jumping Gene GA (JG GA) is presented in this work. We have considered two mathematical and a batch reactor optimal control applications for evaluating the efficacy of the proposed GA. There are two user defined parameters in the proposed algorithm, namely extent of Box-Complex Assistance(BCA), and expansion/contraction factor alpha. Effect of both these parameters on convergence is presented in this work for the proposed GA.

Keywords: Optimization and Scheduling, Energy Processes and Control
Abstract: A new stochastic programming model is proposed for integrated design and operation of natural gas production networks under uncertainty. The model not only addresses the uncertainties explicitly, but also considers the gas compositions and pressure-flow relationships for the whole production network. Due to the many nonconvex constraints, the new model leads to a very difficult large-scale nonconvex mixed-integer nonlinear programming (MINLP) problem. A modified nonconvex generalized Benders decomposition (NGBD) method is developed to efficiently solve this problem. This method integrates in NGBD several bound contraction strategies, which render tighter convex relaxations for constructing tighter lower bounding problems and accelerating the solution of nonconvex subproblems. Case study of an industrial natural gas production system shows the benefits of the proposed stochastic model and the bound contraction integrated NGBD method.

Keywords: Modeling and Identification, Process Applications
Abstract: In this paper, Alberta electricity spot market or Power Pool pricing is studied and the pool price is modeled through a hidden Markov model and multiple local ARX models. By selecting and preprocessing the exogenous factors (e.g. the price forecast from Alberta Electric System Operator (AESO), demand forecast and so forth), a one-hour ahead prediction model for pool price is formulated with parameters being estimated from the real data. Validation results show that this approach can improve the price forecasting and in particular, for high pool prices.

Keywords: Optimization and Scheduling, Process Applications, Modeling and Identification
Abstract: Typical production objectives in the lift gas assisted oil wells include stable and optimal production that meets demands on oil and gas. The attainment of these objectives in a typical production facility involving depletion of oil and gas, from a diversity of reservoirs that have different pressures, compositions, gas to oil ratios and natural flows. The objectives are also bound by constraints on compression and its costs, specifications on the export gas for pressure, volume and composition as per different customer end demands. While optimal allocation of the limited lift gas is desirable, an enterprise-wide, coordinated view of the production from the diverse gas or oil reservoirs would facilitate greater flexibility towards meeting the customer demands in an optimal manner. In this paper, we consider a smaller prototype of a production facility involving a low pressure, medium pressure and high pressure gas reservoirs, which also exhibit varying compositions, gas to oil ratios and natural flows. We propose a formulation that promotes co-ordination between these diverse reservoirs so as to honor the constraints on export gas or liquid pressures, compositions or natural flows. The results have been verified on a prototype of a three well production facility using a simulation model adapted from Jahanshahi et al. (2012).

Keywords: Optimization and Scheduling, Energy Processes and Control
Abstract: A supply network consisting of multiple air separation units and compressors supplies oxygen gas at two pressures to a major steelmaker in South Wales. Each machine has a different efficiency curve, power requirement, and capacity. The aim was to minimise the cost of power and liquid usage of the supply network whilst meeting customer gas flow demands and preventing product spill.

Mass balances of the site were produced and integrated into a Microsoft Excel spreadsheet representation of the supply network. A mixed integer nonlinear programming approach was adopted to allow machines to operate within flow limits and turn on or off based on demand. The GRG nonlinear solver method was used to minimise the cost of running the network arrangement, determined by the sum of all estimated machine powers and the cost of liquid back up usage. Constraints were programmed to maintain steady state, meet demand, and keep machine flows within bounds.

This work demonstrates that extra power requirement, liquid vapourisation and product spill caused by inefficient compression arrangements result in annual site losses of £0.6M. It is shown that through real time optimisation of the gas network a significant reduction in these financial losses can be achieved.

Keywords: Modeling and Identification, Process Applications
Abstract: An identification methodology is proposed in order to model chemical processes using the anisochronic paradigm. This methodology requires only having the reaction curve of the plant to obtain all the parameters of the model for both over-damped and underdamped processes. On the other hand an optimal tuning for PID controllers considering the closed-loop robustness is also proposed for a given anisochronic model. Both the identification and the control are tested in a pH neutralization process, a well known PID benchmark and in a nonlinear CSTR system, to verify the usefulness of the proposed methodology in chemical processes.

Keywords: Model-based Control, Process Applications
Abstract: Urea selective catalytic reduction (urea-SCR) process control is very important for exhaust gas aftertreatment in diesel engines. In this paper, a ammonia coverage ratio tracking control system is designed to achieve high NOx conversion efficiency and low ammonia slip for urea-SCR process. The control system consists of a nonlinear feedforward controller based on flatness and a feedback controller to deal with the uncertainties and disturbance. Finally, the simulation results are given to demonstrate the effectiveness of the proposed control scheme.

Keywords: Optimization and Scheduling, Modeling and Identification, Process Applications
Abstract: Catalytic naphtha reforming is one of the most important processes for high octane gasoline manufacture and aromatic hydrocarbons production. In this article, a modified differential evolution (DE) algorithm is proposed to optimize an actual continuous catalytic naphtha reforming (CCR) process. The optimization problem considers to minimize the energy consumption and maximize the aromatics yield. The CCR process model is established by adopting the 27-lumped kinetics reaction network, and all parameters are adjusted based on the actual process data. The DE algorithm is modified to maintain the diversity of the population. In this mechanism, individuals further from the best individual have larger possibilities to be selected in the mutation operator. The modified DE is evaluated by solving 6 benchmark functions, and the performance is compared with classic DEs. The results demonstrate that the modified DE has better global search ability and higher computation efficiency. Furthermore, the optimization results of catalytic naphtha reforming process indicate that the proposed algorithm has the ability of locating the optimal operating points, in which the aromatics yield is improved, while energy consumption is reduced. Meanwhile, the optimal operating points and results are discussed at the end of the article.

Keywords: Optimization and Scheduling, Modeling and Identification, Model-based Control
Abstract: Optimal scheduling of maintenance and improvement on the water main system is a substantial problem since it can greatly reduces the potential threat of failure and the expenses. The Markov decision process (MDP) based methodology to find the optimal schedule which minimizes the cost is proposed. Since it needs the information about the current state of pipe, cost, and deterioration model, the definition and the usage of auxiliary information are also presented. The objective function and detailed algorithm of dynamic programming (DP) are modified due to the difficulty of implementing the conventional DP approaches. The result is compared to two simple policies via Monte Carlo simulation. Validity of the solution and improvement in computational time are proved. Proposed decision framework provides an easy way to obtain the comprehensive and local-specialized policy.

Keywords: Energy Processes and Control, Optimization and Scheduling
Abstract: Demand response has become a topic of significant research, development, and deployment over the last few years. The energy demand management is a critical task in industrial process systems for the potential benefits to be realized by promoting the interaction and responsiveness of process operation. However, the dynamic behavior, especially transition trajectories, of the underlying process is seldom taken into account during this task. Furthermore, the incorporation of energy constraints related to electricity pricing and availability is one of the key challenges in this process. The purpose of this study is thus to present a novel optimization formulation for energy demand management in dynamic process systems that takes transition behavior and cost explicitly into account, while simultaneously handling time-sensitive electricity prices. This is accomplished by bringing together production scheduling and transition control through a real-time optimization framework. The dynamic formulation is cast as a mixed-integer nonlinear programming problem and demonstrated using a continuous stirred tank reactor example where the energy required is assumed to be roughly proportional to the material flow.

Keywords: Process and Control Monitoring, Model-based Control, Modeling and Identification
Abstract: In this paper a methodology is presented to assess the performance of IMC PI controllers from closed-loop response data for a chosen setpoint signal. Time and frequency domains tools are used to assess the closed-loop performance. Simulation examples illustrate the methodology.

Keywords: Energy Processes and Control, Optimization and Scheduling
Abstract: Due to increasing penetration of wind energy in the recent times, wind farm owners tend to generate increasing amount of energy out of wind farms. In order to meet targets, many wind farms are operated with a layout of numerous turbines placed close to each other in a limited area leading to greater energy losses due to ‘wake effects’ instead of generating more power. To solve the problem in the most optimal way, these turbines need to satisfy many other constraints such as topological constraints, minimum allowable capacity factors, inter-turbine distances etc. Existing methods to solve this complex turbine placement problem typically assume knowledge about the total number of turbines to be placed in the farm, which might be unrealistic. This study proposes a novel hybrid optimization methodology, a combination of evolutionary and classical optimization approaches, to simultaneously determine the optimum number of turbines to be placed in a wind farm along with their optimal locations. Application of the proposed method on a representative case study yields 43% higher Annual Energy Production (AEP) than the results found by one of the existing methods

Keywords: Process and Control Monitoring, Energy Processes and Control, Modeling and Identification
Abstract: Distillation columns are the major energy consumers in petrochemical and chemical industry and their efficient operation is essential for energy saving and product quality enhancement. This paper presents an inferential active disturbance rejection control (ADRC) method for product composition control in distillation columns. The proposed control strategy integrates ADRC with inferential feedback control. Tray temperatures are used to estimate the top and bottom product compositions which are difficult to measure on-line without time delay. In order to overcome the colinearity in the tray temperature data, principal component regression (PCR) is used to build the soft sensors, which are then integrated with ADRC. In order to overcome static control offsets caused by the discrepancy between soft sensor estimations and the true compositions, intermittent mean updating is used to correct PCR model predictions. The proposed technique is applied to a simulated methanol-water separation column.

Keywords: Process Applications
Abstract: Active Disturbance Rejection Control (ADRC) has received considerable attention in recent years as an effective tool for advanced control practitioners to solve control problems for nonlinear uncertain systems. This paper presents results of a simulation study for the control of an evaporator system benchmark, when a multiple linear ADRC control structure is used, and the controllers are designed based on first and second order linear models approximating the process dynamics. In addition to the control performance under nominal conditions, the robustness with respect to plant-model mismatch is studied. The major advantage of ADRC compared to MPC and PI control approaches is the simple and transparent tuning, and that only a coarse process model is sufficient for the design. For a broader industrial application of ADRC in the process industries, user-friendly off-line design tools as well as real-time ADRC control software for PLCs and DCS still have to be developed.

Keywords: Process and Control Monitoring, Process Applications, Modeling and Identification
Abstract: Control of sulfur content in diesel streams demands for on-line measurement of this component, with low pure time delay and easy operation. In this way, 2D fluorescence Spectroscopy becomes a reasonable choice for soft-sensor development. This work reports a different strategy to overcome the aforementioned lack of selectivity or signal overlapping using the selection of individual fluorescence spectral components combined into multilinear models to predict sulfur content in diesel streams from Ultra Low Sulfur Diesel and Diesel S100. The results obtained using Pure Spectral Chemometrical Modeling and Ant Colony Optimization showed good results as a source of information to determine the sulfur content in diesel.

Keywords: Process and Control Monitoring, Modeling and Identification
Abstract: Stiction is a persistent control valve problem in the process industry responsible for oscillations and, consequently, losses of productivity. Its early detection and separation from other oscillation causes is an important issue in the industrial context. One of simple and effective approaches to detect stiction has been proposed by Yamashita that employed a pattern recognition principle. While its performance is good in flow control loops, it fails to properly diagnose other types of processes.

The present work details a new approach that enables the application of the Yamashita pattern recognition principle to level and other integrating process control loops. A simulation study demonstrates its capabilities in clean and noisy environments and analyzes the impact of the noise on the diagnostic performance.

Keywords: Process and Control Monitoring, Modeling and Identification
Abstract: One of the challenges of utilizing soft sensors is that their prediction accuracy deteriorates with time due to multiple factors, including changes in operating conditions. Once soft sensors are designed, a mechanism to maintain or update these models is highly desirable in industry. This paper proposes an index that can monitor the prediction performance of soft sensor models and provide guidance about when to update these models. In the proposed approach, a Kalman filter based model mismatch index is developed to monitor the prediction performance of soft sensors with the support of traditional process monitoring indexes, T² and SPE. Then, the soft sensor model can be updated through partial least squares (PLS) regression by using samples from the online training set and new process conditions. The proposed online update method is applied to an industrial process case study and the effectiveness of the proposed approach is demonstrated by comparing with traditional recursive partial least squares (RPLS).

Keywords: Process Applications, Process and Control Monitoring, Optimization and Scheduling
Abstract: The open-loop two-move method is an effective compensation to remove oscillations caused by control valve stiction, but takes too much time for implementation. This paper proposes a new implementation of open-loop two-move compensation method based on the estimation of controller output associated with the desired valve position, but without any assumption of the valve position in oscillation. The proposed method can reduce the time cost and lower the amplitude of open-loop step responses, leading to a quicker response of the process and less negative effect during the compensation. The effectiveness of the proposed method is demonstrated by several examples.

Keywords: Modeling and Identification
Abstract: A dynamic model that considers heat transfer laws, internal heat distribution, and different cooking strategies was developed to simulate dumpling cooking. The cooking strategies that were considered include refilling water, changing heating power, and adding dumplings. The concept of degree of doneness (DoD) was used to gauge the appropriate level of cooking dumplings. Experiments were designed to determine the acceptable range of DoD. The developed model enables the prediction of water and dumpling temperatures, water evaporation, and dumpling quality. The model, which reflected the evolution of dumpling cooking, may be used as a basis to optimize the process for energy saving and full automation purpose.

Keywords: Energy Processes and Control, Process and Control Monitoring, Process Applications
Abstract: Fuel cells in domestic micro combined heat and power (µCHP) systems are subject to several degradation phenomena that can be counteracted with appropriate control strategies. Cost is an important factor, and additional equipment must be minimised. Controllers are proposed to counteract stack drying, flooding, poisoning by CO, and to estimate anodic hydrogen excess ratio to prevent hydrogen starvation. Tests on an emulator in LabVIEW indicate that the controllers are able to perform their function, and will be tested on a real system in 2015.

Keywords: Process and Control Monitoring, Process Applications, Energy Processes and Control
Abstract: An industrial workflow represents a sequence of tasks or actions that describes an operational procedure. In this paper, workflow strategies are developed that capture operational knowledge by analyzing event logs of how an operator has executed a procedure while controlling or monitoring a process. In this work, our prime focus is on infrequent process operations such as plant start-up and shutdown procedures. We propose a workflow conformance method to continuously monitor and compare operator actions with standard operating procedures (SOPs) and identify procedural violations that could compromise process safety and efficiency. An industrial case study is presented to illustrate applications of workflow conformance monitoring to identify operational problems associated with human factors, process and instrumentation.

Keywords: Modeling and Identification, Model-based Control
Abstract: This paper uses the key-term separation principle and develops a multi-innovation stochastic gradient algorithm for Hammerstein MIMO output error systems. The basic idea is to decompose a Hammerstein MIMO system into two subsystems and to identify the parameters of each subsystem interactively. The parameter estimation accuracy can be improved with the innovation length increasing. The simulation example verifies the effectiveness of the proposed algorithm.

Keywords: Process Applications, Batch Process Modeling and Control
Abstract: Gas-metal arc welding is a widely used welding process. The testing of such welds is done offline and in most cases after the welding operation is over. To monitor the progress of the welding run, it is essential to develop multivariate data analysis techniques that can classify the welds into good or bad runs and also be able to predict the quality variables. In this work, popular multivariate data analysis methods such as hierarchical clustering analysis, principal component analysis and partial least squares are used to develop classification and regression models to predict the weld quality based on various parameters. The results indicate that models obtained using these methods are effective in classification and prediction of weld quality and can be further developed for online and industrial uses in weld run monitoring.

Keywords: Process and Control Monitoring, Modeling and Identification, Model-based Control
Abstract: Although many PID tuning approaches are available, it is not easy to find a method that does not require any engineer/operator interference. In this work, we present a fully automated approach for PID tuning based on relay feedback. This method involves sending the relay feedback test data from PLCs (Programmable Logic Controller) into a historian, analyzing the test data using a tuning application to generate a tuning report that contains PID parameters and sending the report back to the operator station to retune the controllers in PLCs.

This paper is focused on the following three keys steps: 1) A method to identify persistent steady-state conditions in a control loop using routine operating data because any tuning test is performed when the process is operating at steady state, 2) A novel procedure to implement relay based tuning test, 3) A new model identification method which is a combination of frequency-domain and time-domain analysis. Subsequently, the identified plant model is used to obtain PID tuning parameters based on IMC design.

The approach has been tested on an industrial test setup in which all the control loops of the Tennessee Eastman process are controlled by a Siemens PLC. The necessary relay parameters, the hysteresis and relay amplitude, for the test are estimated automatically where interference by an engineer or an operator is not required. The new method for model identification is robust against measurement noises. The proposed method is able to tune the important control loops in the Tennessee Eastman process successfully.

Keywords: Process Applications, Modeling and Identification, Model-based Control
Abstract: In this work, we propose a control-relevant multiple linear modeling approach for simulated moving bed chromatography (SMBC) by linearizing the first principles model at carefully chosen equilibrium points. Subsequently, sub-models to account for port switching for each of the linear model are obtained. Model aggregation is done using Bayesian weighting to generate multiple model predictions for the nonlinear dynamics of SMBC. The multiple model approach is validated using simulations for cyclic steady state (CSS) of SMB as well as for a transition between two optimal CSS points for separation of a glucose-fructose mixture.

Keywords: Energy Processes and Control
Abstract: To investigate potential improvement in wind turbine control employing LIDAR measurement, pseudo-LIDAR wind speed data is produced with Bladed using a designed sampling strategy, and assessed with preliminary frequency-domain analysis. A model-inverse feed-forward controller is adapted to combine with feedback control so as to enhance pitch control performance at high wind speed. This controller is applied to an industrial-scale 5MW wind turbine model and the control performance is compared with a baseline feedback controller. Simulation study demonstrates that the combined feed-forward/feedback control scheme has improvements in reducing pitch angle variation and reduction of load relevant metrics.

Keywords: Optimization and Scheduling, Model-based Control, Process Applications
Abstract: Control of reactive distillation (RD) systems is a challenging problem due to nonlinear dynamic and steady state behavior arising from complex interactions between reaction kinetics and vapor liquid equilibria. The focus of work reported in the literature on control of RD systems is on solving servo and regulatory control problems. However, when the unmeasured disturbances / system parameters drift from their nominal values, the operating performance of a RD system can potentially be improved if real time optimization (RTO) techniques are employed for deciding the optimum operating point on-line. In this work, an integrated RTO and adaptive NMPC approach has been proposed for operating an RD system in a economically optimal manner. At the core of the integrated scheme is a nonlinear Bayesian state and parameter estimator, which is used as a common link between the RTO and the NMPC components. Estimates of the drifting unmeasured disturbances / parameters generated by the state estimator are used to update the steady state model used for RTO and the dynamic model used for predictions. This facilitates relatively frequent application of RTO without having to wait for the system to reach steady state and makes the NMPC formulation adaptive. Efficacy of the proposed integrated optimizing control scheme is demonstrated by conducting simulation studies on an ideal RD column. The control problem under investigation is optimal inferential control of product concentrations in the face of drifting reactant flow disturbance. Analysis of the simulation results reveals that the proposed integrated approach is able to satisfactorily identify and track economically beneficial optimum operating point of the system.

Keywords: Model-based Control, Optimization and Scheduling, Energy Processes and Control
Abstract: In this work, the authors have proposed a multiple-linear model based control scheme for the boiler-turbine unit. The proposed control scheme makes use of local linear state space models determined at different load conditions, by linearizing the nonlinear boiler-turbine model reported by Bell and Astrom, 1987. Each local linear state space model is stabilized via state feedback approach and is used to design a multiple-linear model based controller for the boiler-turbine unit. Genetic Algorithm is employed to optimally tune the parameters of each local linear controller. The global controller output is determined as a linear combination of all local controller outputs. The extensive simulation studies show that the proposed control scheme effectively handles the input constraints of the boiler-turbine unit and meets the required electrical demand.

Keywords: Model-based Control
Abstract: Based on the direct synthesis approach, a simple and effective design method of centralized proportional integral (PI) controller for non-square processes is investigated in this paper. With the help of desired closed-loop diagonal transfer function to reduce interaction between individual loops, analytical expressions for PI controller, through the steady and dynamic information of the open-loop transfer function, are derived for the first order plus time delays model which often arises in the chemical production process. Compared with the existing direct synthesis approaches, the proposed controller design method requires no approximation of the pseudo-inverses of process. Example is introduced to show the effectiveness and simplicity of the proposed technique.

Keywords: Process Applications
Abstract:

Keywords: Model-based Control, Modeling and Identification, Process Applications
Abstract: In this paper, we develop an economically optimizing Nonlinear Model Predictive Controller (E-NMPC) for a complete spray drying plant with multiple stages. In the E-NMPC the initial state is estimated by an extended Kalman Filter (EKF) with noise covariances estimated by an autocovariance least squares method (ALS). We present a model for the spray drying plant and use this model for simulation as well as for prediction in the E-NMPC. The open-loop optimal control problem in the E-NMPC is solved using the single-shooting method combined with a quasi-Newton Sequential Quadratic Programming (SQP) algorithm and the adjoint method for computation of gradients. We evaluate the economic performance when unmeasured disturbances are present. By simulation, we demonstrate that the E-NMPC improves the profit of spray drying by 17% compared to conventional PI control.

Keywords: Model-based Control, Batch Process Modeling and Control
Abstract: Batch processes are of great importance in process industry. However, the control algorithm design is difficult for those with constraints. This is because stability and recursive feasibility along directions of time and batch should be guaranteed simultaneously. In this paper, a stable model predictive control strategy with zero terminal state constraints is proposed. Stability and recursive feasibility along two directions are guaranteed and proved. Simulation results are given to show the effectiveness of the algorithm.