Keywords: Process and Control Monitoring
Abstract: The recent interest in big data has shown up in almost all aspects of knowledge discovery; including engineering, medicine, business, commerce, finance, and even science to benefit from the power of big data. The Internet of Things, smart and wireless sensors, wireless communications, mobile devices, smart devices, and smart manufacturing make data an abundant source of information from which to derive knowledge and make decisions.

For process engineering systems where processes, units, and equipment are designed with well-specified purposes under well-controlled operations, mechanistic models and principles are dependable. However, for the operation of emerging or abnormal situations that are not expected in the design, data become indispensable assets for the decision-making in safe and efficient operations. In this plenary we offer a perspective on the essence of process data analytics, how data have been effectively used in process operations and control, and new perspectives on how process systems operations might evolve to a paradigm of data-enhanced operations and control. In particular the focus is on the latent structure modeling of high dimensional and massive amount of data from which to explore interesting latent data structures for the purpose of process data analytics, including abnormal condition monitoring, inferential estimation, and predictions based on principal auto- and cross- correlations. The concept of principal time series modeling will be introduced. To conclude the talk, we give a future perspective in exploring the power of new machine learning techniques that have enjoyed tremendous development in two decades.

Keywords: Batch Process Modeling and Control, Process and Control Monitoring, Process Applications
Abstract: In the process industries Big Data has been around since the introduction of computer control systems, advanced sensors, and databases. Although process data may not really be BIG in comparison to other areas such as communications, they are often complex in structure, and the information that we wish to extract from them is often subtle. Multivariate latent variable regression models offer many unique properties that make them well suited for the analysis of historical industrial data. These properties and use of these models are illustrated with applications to the analysis, monitoring. optimization and control of batch processes, and to the extraction of information from on-line multi-spectral images.

Keywords: Model-based Control, Optimization and Scheduling
Abstract: We discuss the design of sampled-data economic nonlinear model predictive control schemes for continuous-time systems. We present novel sufficient convergence conditions that do not require any kind of terminal constraints nor terminal penalties. Instead, the proposed convergence conditions are based on an exact turnpike property of the underlying optimal control problem. We prove that, in the presence of state constraints, the existence of an exact turnpike implies recursive feasibility of the optimization. We draw upon the example of optimal fish harvest to illustrate our findings.

Keywords: Model-based Control, Process Applications, Process and Control Monitoring
Abstract: With decades of successful application of model predictive control (MPC) to industrial processes, practitioners are now focused on ease of commissioning, monitoring, and automation of maintenance. Many industries do not necessarily need better algorithms, but rather improved usability of existing technologies to allow a limited workforce of varying expertise to easily commission, use, and maintain these valued applications. Continuous performance monitoring, and automated model re-identification are being used as vendors work to deliver automated adaptive MPC. This paper examines industrial practice and emerging research trends towards providing sustained MPC performance.

Keywords: Process and Control Monitoring, Modeling and Identification, Model-based Control
Abstract: For thermal technologies in heavy oil and oil sand reservoir extraction, such as cyclic steam stimulation and steam-assisted gravity drainage (SAGD), suboptimal steam conformance leads to recovery factors between 25-50%. Although preliminary research using Proportional-Integral-Derivative (PID) control in SAGD operations has proved beneficial towards steam conformance, PID control is responsive only to deviation from set-points and lacks constraint-handling capabilities. This results in suboptimal actuation signals that are sometimes unattainable. This paper summarizes research on a Model-Predictive-Controller (MPC) with proactive adjustments of steam injection rate. The steam injection rate was determined based on recursive parameter updates of a suitable time varying dynamic model describing the implicit relationship between the subcool temperature difference and the input heat rate, to achieve optimal steam conformance. Furthermore, the steam injection rate was constrained such that the pressure with which the steam impinged on the formation, called well bottom hole pressure (BHP), was below the formation fracture pressure of 4500 kPa at all times.The real time control study was made possible by establishing a bidirectional communication link between the Computer Modeling Group (CMG) STARS TM, and MATLAB/Simulink software. The three- dimensional heterogeneous reservoir model, developed in STARS TM acted as a virtual plant and the MPC, developed in MATLAB/Simulink, acted as an onsite controller. Results show 35.7% improvement in oil recovery and a more efficient cumulative steam-to-oil ratio (cSOR) profile in comparison to the base case of steam injection at a constant BHP of 4000 kPa.

Keywords: Model-based Control, Modeling and Identification, Process and Control Monitoring
Abstract: Maintaining uniformly satisfactory control performance of an MPC scheme in the face of changing operating conditions is a difficult task. An adaptive MPC scheme that directs the output towards a reference and simultaneously injects a probing signal to get more information about the system for better model identification appears to be ideally suited for achieving this goal. In this work, taking motivation from the dual control problem originally developed by Feldbaum (1960), a MIMO adaptive dual MPC (adaptive DMPC) formulation has been proposed, which does not require external probing signals to improve the model parameter estimates. The objective function of the MPC is modified to include terms that ensure that sufficient excitation is injected into the system while performing the control tasks. The efficacy of the proposed adaptive DMPC formulation is evaluated by conducting experimental studies on the benchmark heater mixer system. The experimental results demonstrate that the proposed formulation is able to inject probing inputs of small magnitude while meeting the desired servo and regulatory control objectives.

Keywords: Model-based Control, Process Applications, Optimization and Scheduling
Abstract: Irrigation system is a large scale system, consisting of many interacting channels, and spanning vast geographical areas. In practice, the water level is expected to the desired reference values for safety and efficiency. In this paper, the model of irrigation system is modeled by the principle on conservation of mass. Because of the strong coupling between neighboring pools in irrigation system and large disturbance from the environment, a non-cooperative distributed model predictive control (NDMPC) algorithm based on Nash optimality is proposed for the large scale irrigation system to design the controller for setting points regulation. Finally, a simulation example is given to demonstrate the effectiveness of the proposed algorithm.

Keywords: Model-based Control, Process Applications, Optimization and Scheduling
Abstract: The goal of this contribution is to study numeric solution techniques for implementing optimizing control of polymerization processes in tubular reactors with multiple side injections of monomer along the reactor. The onfiguration of the reactor causes long delays between the inputs and the measurements at the reactor outlet and sharp moving fronts when the inflows are changed. Moreover, the polymerization kinetics are strongly nonlinear. The process is described by 1D partial differential equations along the reactor length. This makes the application of optimizing control based on a rigorous process model challenging. The so called weighted essentially non-oscillatory scheme (WENO) is used to discretize the spatial dimension of the plant model. This method avoids the need for a very large number of discretization points and still the model can be simulated sufficiently accurately. The resulting ode model contains 1600 states and comprises five manipulated variables. We implement the optimizing controller using two different approaches: At first single shooting with control vector parametrization is used which is simple to implement and has fewer decision variables. This is compared to full discretization scheme using orthogonal collocation on finite elements which results in a very large but very sparse and structured nonlinear programing problem. The simulation results show that both approaches have a similar performance and drive the system to a significantly more productive steady state.

Keywords: Model-based Control
Abstract: A discrete model predictive controller for inverse response control systems is presented. The controller has two degrees of freedom (DOF): The feedback path contains current and past errors, and the feedforward path contains future setpoint values for tracking control. The controller can eliminate the inverse response completely due to its characteristic of having an infinite number of predictive values of the setpoint disturbance.

Keywords: Energy Processes and Control, Model-based Control
Abstract: This paper presents a control oriented model development for waste heat recovery Rankine based control systems in heavy duty trucks. Waste heat recovery systems, such as Rankine cycle, are promising solutions to improve the fuel efficiency of heavy duty engines. Due to the highly transient operating conditions, improving the control strategy of those systems is an important step to their integration into a vehicle. The system considered here is recovering heat from both EGR and exhaust in a serial arrangement and use a mixture of water and ethanol as working fluid. The paper focuses on a comparison of a classical PID controller which is the state of the art in the automotive industry and a nonlinear model based controller in a simulation environment. The nonlinear model based controller shows better performance than the PID one and ensures safe operation of the system.

Keywords: Modeling and Identification, Energy Processes and Control, Process Applications
Abstract: Hydrogen rich gas produced by hydrocarbon reforming contains CO of several thousand ppm level which acts as a poison for the proton exchange membrane fuel cells. Preferential oxidation (PROX) is one of the promising methods that reduce its concentration below 10ppm. The PROX reactor with the CuO-CeO2 catalyst shows optimal reactor temperature and a control system that maintains the PROX reactor temperature at its optimal one under changing environments and catalyst deactivations. Experimental studies to obtain a dynamic model for the CO PROX reactor have been done and the Wiener-type nonlinear model is obtained. Then the extremum seeking control that can track the optimal temperature for the Wiener-type nonlinear system is applied and its performance is verified experimentally. The extremum seeking control is a non-parametric real-time optimization method and can be applied effectively to other CO PROX reactors with different operating conditions.

Keywords: Energy Processes and Control, Process Applications, Model-based Control
Abstract: Dynamic and flexible operation of a carbon capture plant is important as thermal power plants must be operated very flexibly to accommodate large shares of intermittent energy sources such as wind and solar energy. To facilitate such operation, dynamic models for simulation, optimization and control system design are crucial. In this paper, we present a dynamic mathematical model for the absorption and desorption columns in a carbon capture plant. Moreover, we implement a decentralized proportional-integral (PI) based control scheme and we evaluate the performance of the control structure for various operational procedures, e.g. start-up, load changes, noise on the flue gas flow rate and composition. Note that the carbon capture plant is based on the solvent storage configuration. To the authors knowledge, this is the first paper addressing the issue of start-up operation and control of carbon capture. The study demonstrates that the implemented control structure keeps the carbon capture process at 90% CO2 removal rate with a deviation up to 8% during load variations. In addition, it reveals that the control structure brings the process to the desired set point in approximately 10 min during process start-up.

Keywords: Energy Processes and Control, Process Applications
Abstract: In this paper, a scalable framework is developed to analyze complex process networks consisting of multiple material integration loops. Using information on the orders of magnitude of different material flows, the framework identifies the time scales where different process units evolve, the manipulated inputs available in each time scale, and the canonical forms of the reduced order models for each time scale, in an automated fashion. The application of the framework is illustrated via a case study on an example process network.

Keywords: Energy Processes and Control, Optimization and Scheduling
Abstract: This paper presents an integrated approach for the optimization and control of heat-exchanger networks (HENs) that are managed proactively, where both renewable energy generation and time-varying electricity pricing are considered in calculating the utility cost and revenues. A receding horizon or real-time optimization strategy is carried out to deal with the resulting variability in energy resource availability and costs. The heat-exchanger super-structures are first defined where transitions between the included HEN sub-structures are assumed to be instantaneous. The optimal results are sent to the local controllers as structural set-points to be implemented. The explicit nonlinear models of heat exchangers built in Aspen Plus® are adopted to describe the dynamic behaviors of the HENs. Furthermore, the dynamic models are incorporated not only for simulating the local control scheme, but also the supervisory level to account for the structural transitions that incur operational and possibly capital costs.

Keywords: Optimization and Scheduling, Modeling and Identification, Process Applications
Abstract: In this work, the operational optimization of seawater reverse osmosis (SWRO) process is studied to reduce the energy cost. Effects of load fluctuation and electricity price change were considered, and storage tanks were used to adjust the water production and balance the freshwater demand. Based on comprehensive first-principles, a detailed and accurate mathematical model with spiral-wound membranes, water storage tanks and water demands is developed and validated by plant data and the ROSA9.0 simulation package. The operational optimization problem with the form of DAEs was solved with simultaneous approach and IPOPT solver. Computing results of real SWRO plant were then investigated. It is shown that, compared with conventional operation and operation with constant water level, about 27.88% and 24.17% of energy cost, respectively, can be saved by optimizing both feed flowrate and feed pressure within normal range. Optimal profiles of key performance states were analyzed and future work for operational cost saving are described.

Keywords: Process and Control Monitoring, Process Applications, Modeling and Identification
Abstract: Effective monitoring of industrial processes provides many benefits. However, for dynamic processes with strong nonlinearity many existing techniques still cannot give satisfactory monitoring performance. This is evidenced by the well known Tennessee Eastman (TE) benchmark process, where some faults, {em e.g.} Faults 3 and 9, have not been comfortably detected by almost all data-driven approaches published in the literature. This is because most data driven approaches, such as the principal component analysis (PCA) are linear. In recent years, powerful nonlinear analysis tools using kernel principles have been proposed. However, these tools have not been successfully applied to dynamic systems due to enormous dimensionality and complexity issues. This paper proposes nonlinear dynamic process monitoring based on kernel canonical variate analysis (KCVA). The proposed technique performs the traditional canonical variate analysis with KDE (CVA-KDE) in the kernel space generated from kernel PCA. The kernel PCA accounts for the nonlinearity in the process data while the CVA captures the process dynamics. The approach was tested on the TE benchmark problem for fault detection. The results obtained showed that KCVA detected faults at a higher rate and much earlier than CVA especially in the more difficult faults such as Faults 3 and 9 in the TE process which cause very little variation in the measured variables.

Keywords: Process and Control Monitoring
Abstract: Traditional process monitoring methods based on kernel canonical variate analysis do not extract variances. They cannot judge whether a process fault that is detected affects product quality. A nonlinear quality-relevant process monitoring method based on kernel input-output canonical variate analysis (KIOCVA) is proposed. Firstly, Process variables and quality variables are mapped into higher-dimensional linear feature spaces via unknown nonlinear mappings respectively. The higher-dimensional linear feature spaces are projected to three subspaces, an input-output correlated subspace that captures correlations between process data and quality data, an uncorrelated input subspace and an uncorrelated output subspace. To monitoring the variances of the uncorrelated input subspace and the uncorrelated output subspace, principal component analysis is performed. Correlations and variances in the higher-dimensional linear feature spaces are extracted by means of nonlinear kernel functions. The proposed KIOCVA method can judge the process fault that is detected affects product quality or not. The effectiveness of the proposed method is demonstrated by case studies of Tennessee Eastman process.

Keywords: Process and Control Monitoring
Abstract: For enhancing product quality and operation safety, statistical process monitoring has become an important technique in process industries, where principal component analysis (PCA) is a commonly used method. However, PCA assumes that the training data matrix only contains an underlying low-rank structure corrupted by dense noise. When gross sparse errors, i.e. outliers, exist, PCA often fails. In this paper, a robust matrix recovery method called stable principal component pursuit (SPCP) is utilized to solve this problem. A process modeling and monitoring procedure is developed based on SPCP, the effectiveness of which is illustrated using the benchmark Tennessee Eastman process.

Keywords: Process and Control Monitoring, Optimization and Scheduling
Abstract: This paper reviews the problems associated to the presence of gross errors in data reconciliation problems as well as the main approaches to avoid their undesirable effects: detection and removal of the faulty measurements and minimization of their effect on the reconciled results by using robust estimators that modify the cost function. In the first case, Principal Component Analysis is used to detect the gross errors while, in the second, the Fair function is used. A combined approach is also presented that improves the PCA results. The methods are evaluated in a realistic large-scale problem with plant data corresponding to the hydrogen network of a petrol refinery

Keywords: Process and Control Monitoring, Modeling and Identification, Process Applications
Abstract: The paper presents a detailed comparison of different identification techniques applied to valve stiction quantification, possibly in the presence of nonstationary unknown disturbances. The control loop with sticky valve is modeled as a Hammerstein system, in which the nonlinearity is identified using enumeration of the parameters’ space. Five different techniques for identification of the linear model are compared in terms of achievable performance. In particular, the capability to cope with the presence of nonstationary disturbances is analyzed. The techniques allow one to estimate the unknown actual valve position (MV), without requiring any process knowledge, being based only on data which are usually recorded in industrial plants: controller output (OP) and controlled variable (PV). Simulations show that external perturbations can be tolerated, thus ensuring a reliable evaluation of stiction in practical situations where external disturbances are usually present. Models which incorporate a time varying additive nonstationary disturbance grant a better process identification and a more accurate stiction estimation in the case of disturbance acting simultaneously with valve stiction. However, simpler models are the best choice when stiction happens to be the only source of loop oscillation. Results are confirmed by application to real data: pilot plant data are used to corroborate the effectiveness of the techniques.

Keywords: Process and Control Monitoring, Model-based Control
Abstract: Valve stiction is one of the most common causes for poor performance in industrial control loops. Therefore, a non-invasive method which can detect and quantify stiction is urgently needed in the process industry. Most of the current stiction estimation methods use time domain criterion, e.g. Mean Square Error, to jointly identify the stiction and process model parameters. However, stiction induced oscillation is a phenomenon which has some specific characteristics in the frequency domain. Thus, extracting frequency domain information in the routine operation data will provide a more reliable and accurate stiction estimation. In this work, under the framework of Hammerstein model identification and global optimization, a new stiction quantification method based on time and frequency domain criterions is proposed. Several simulation case studies are demonstrated to validate the proposed method.

Keywords: Process Applications, Process and Control Monitoring
Abstract: For better performance analysis and design improvement of industrial alarm systems, a variety of advanced alarm management tools have been developed recently. These tools are quite comprehensive and handy in various applications, such as the assessment of alarm systems, detection of nuisance alarms, alarm flood analysis, and recommendation of better configurations. To demonstrate the effectiveness of these tools, this paper presents some industrial case studies based on the alarm data collected from an oil sand extraction plant operated by Suncor Energy in northern Alberta, Canada. Application results show the practicality and utility of the advanced alarm management tools for alarm rationalization and routine alarm management.

Keywords: Process Applications, Process and Control Monitoring
Abstract: An alarm flood is a commonly encountered problem in industrial alarm systems. It may distract operators from critical alarms and lead to plant failures. Thereby, addressing of alarm floods becomes an important task to achieve reliable alarm management. Some sequence alignment algorithms have been used to discover useful patterns of alarm floods from historical alarm data. In order to improve the computational efficiency of the existing algorithms, a fast sequence alignment approach is developed based on a basic local alignment search tool. It accelerates the computation using a seed-and-extend strategy. The priority information is incorporated so that the algorithm is more sensitive to alarms of higher priorities. Besides, the set based pre-matching is proposed to avoid unnecessary computation by excluding irrelevant alarm floods and alarm tags. Industrial case studies demonstrate that the proposed method significantly outperforms the existing algorithms.

Keywords: Process and Control Monitoring, Process Applications, Optimization and Scheduling
Abstract:

Keywords: Process and Control Monitoring
Abstract: Alarm floods have always been serious hazards in industrial process monitoring since they overwhelm operators with large amount of alarm messages raised within a short period of time. In this paper, we propose an algorithm to find an optimal alignment of multiple alarm flood sequences so that based on this alignment an alarm sequence pattern can be easily found. The pattern could reveal the correlation between alarm messages in the alarm floods, which cannot be obtained by applying other alarm management techniques such as delay timers and dead-bands. The pattern could also help find the root cause, locate badly designed part in alarm systems, and predict incoming alarm floods. A dataset from an actual chemical plant has been used to test the effectiveness of the proposed algorithm.

Keywords: Process and Control Monitoring, Modeling and Identification, Model-based Control
Abstract: Complex chemical processes usually operate at multiple operating modes, resulting from various factors, such as changes in market demand, set point modifications, and feedstock changes. It is difficult to monitor a multimode process without generating significant number of false alarms. In this paper, a risk-based alarm system design methodology is proposed to monitor multimode processes. The methodology comprises of three main steps: i) analysis of operating data using Gaussian mixture model, ii) identification of independent operating modes (e.g. set points, virtual transient state), and iii) probabilistic model to assess risk and activation of appropriate warning. A continuous stirred tank reactor with model predictive control system is used to demonstrate the effectiveness of the proposed method.

Keywords: Process and Control Monitoring, Process Applications
Abstract: Alarm systems play an important role in industry to notify operators of abnormal or fault situations. In real industrial plants, however, a majority of nuisance alarm signals, including false alarms and missed alarms, interfere with operators' judgment. The technique of delay-timers, a common means to reduce both false alarm rate (FAR) and missed alarm rate (MAR), is widely performed, yet it suffers from a delayed response. In this paper we propose an improved delay-timer annunciation and clearance method to enhance the performance of conventional delay-timers via bypassing some states in state switch under some conditions. Also the improved delay-timer performance indices are calculated by using Markov chains.

Keywords: Model-based Control, Optimization and Scheduling, Process and Control Monitoring
Abstract: This paper proposes a proportional-integral extremum-seeking control technique. The technique is a generalization of the standard perturbation based techniques that provides fast transient performance of the closed-loop system to the optimum equilibrium of a measured objective function. The main contribution is that the formal development of this technique does not require the need for a time-scale separation. It is assumed that the equations describing the dynamics of the nonlinear system and the cost function to be minimized are unknown. The cost function and its first time derivative are assumed to be measured. The equilibrium of the unknown dynamics are assumed to be asymptotically stable and the cost function dynamics are assumed to be of relative order one. It is shown that the closed-loop ESC can also stabilize the steady-state optimum for an unknown unstable nonlinear control system. The stabilization result is quite general and provides a new approach to output feedback control of nonlinear systems.The effectiveness of the proposed approach is demonstrated using several simulation examples.

Keywords: Model-based Control
Abstract: The adaptive output feedback control problem of chemical distributed parameter systems is investigated while the process parameters are unknown. Such systems can be usually modeled by semi-linear partial differential equations (PDEs). A combination of Galerkin's method and proper orthogonal decomposition is applied to generate a reduced order model which captures the dominant dynamic behavior of the system and can be used as the basis for Lyapunov-based adaptive controller design. The proposed control method is illustrated on thermal dynamics regulation in a tubular chemical reactor where the temperature spatiotemporal dynamic behavior is modeled in the form of a semi-linear PDE.

Keywords: Model-based Control
Abstract: Dissipativity theory is an effective tool for system analysis and control design for network dynamics. This paper develops a nonlinear version of dissipativity in quadratic differential forms (QDF) by using the notion of differential dissipativity that lifts the storage functions and supply rates to the tangent bundle. This paper extends differential dissipativity to the high order derivatives of the displacement of external variables. It allows for a more detailed description of nonlinear systems in terms of dissipativity properties which can results in less conservative dissipativity based stability and performance conditions. The determination of the supply rate and storage function, the stability condition for open-loop systems and interconnection for networked dynamics in the framework of high-order differential dissipativity are investigated.

Keywords: Model-based Control, Optimization and Scheduling, Process and Control Monitoring
Abstract: In this paper, a distributed extremum seeking control technique is proposed to solve a class of real-time optimization problems over a network of dynamic agents with unknown unstable dynamics. Each dynamic agent measures a cost that is shared over a network. A dynamic average consensus approach is used to provide each agent with an estimate of the total network cost. The extremum seeking controller uses the local estimate of the total cost to adjust the value of the local decision variables. The contribution of the proposed technique is the simultaneous stabilization of the network dynamics and the distributed optimization of the total network cost. A dynamic network simulation example is presented to demonstrate the effectiveness of the technique.

Keywords: Optimization and Scheduling, Process and Control Monitoring, Model-based Control
Abstract: Optimal operation of chemical plants is usually accomplished by first finding the optimal steady state using the nominal set of disturbances and model parameters. The optimization is in most cases model based and therefore subject to uncertainties. This may lead to sub optimal control actions with significant economical losses. One idea to tackle this problem is to use the available measurements to adapts the nominal inputs during operation in feedback control scheme. This can be achieved by a neighbouring extremal controller that updates the inputs based on the deviation of the measured outputs from their nominal value. In this paper we generalize the neighbouring extremal control design method that has been presented in the literature to explicitly handle measurement noise and implementation errors. The benefits of the proposed method are illustrated in a case study where we show that the sensitivity of the controller performance to measurement noise is considerably reduced.

Keywords: Model-based Control
Abstract: Distributed control is a computationally efficient and high performance approach for plantwide process systems, where the interactions between process units can be dealt with using the information exchanged among controllers. In this work we develop an approach to analyze distributed control systems where the controllers communicate at a rate lower than the sampling rate for the local process units. This is motivated by the observation that many chemical processes have plant-wide dynamics on a time-scale slower than that of the individual processes' dynamics. Based on the dissipative systems theory, this approach can be used to determine the impact of the reduced communication rate on the stability and control performance of the plant-wide system.

Keywords: Model-based Control
Abstract: We derive a nonlinear model of the steam cycle of a solar power plant, estimate its unknown parameters with measured data, and design a linear model predictive controller based on the resulting model. Simulations show that the linear model predictive controller is able to regulate the electrical power of the nonlinear steam process of the plant to a given reference trajectory very well. Moreover, it results in very good disturbance rejection.

Keywords: Process Applications, Model-based Control, Modeling and Identification
Abstract: Steam quality is a critical process variable in the once through steam generator (OTSG) operation. However, the lack of online measurement and closed loop control of steam quality limits the efficient operation of OTSGs. To resolve this problem, this paper presents a model predictive control (MPC) solution based on soft sensor measurement for OTSG steam quality. Bad status handling strategy is outlined to ensure reliable estimation and control results. Instrumentation reliability issue is also considered in the MPC design. Successful application results have demonstrated effectiveness of the developed control strategy.

Keywords: Energy Processes and Control, Process and Control Monitoring, Modeling and Identification
Abstract: Two modifications to a reference direct methanol fuel-cell (DMFC) system are considered to improve methanol fuel utilisation. Fuel utilisation can be optimised by minimising the parasitic methanol crossover by manipulating methanol concentration in the fuel cells. Simulations of the two systems indicate that there is considerable potential to increase fuel utilisation for dynamic concentration control in DMFC systems subject to variable load.

Keywords: Energy Processes and Control, Process Applications
Abstract: This paper presents the first operability studies for the computation of catalytic membrane reactor feasible regions for process design and control. The equilibrium-limited direct methane aromatization conversion process to hydrogen and benzene is addressed as the application. For the operability analysis, a nonlinear model that describes the membrane reactor system is derived. This analysis is performed for different reactor designs, membrane characteristics and flow configurations. The obtained results demonstrate the capabilities of the nonlinear operability approach to provide novel findings on the design of emerging energy systems.

Keywords: Energy Processes and Control, Modeling and Identification
Abstract: We present a detailed multiphysics model capable of simulating the dynamic behavior of a solid oxide fuel cell (SOFC). This model includes a description of all the important physical and chemical processes in a fuel cell: fluid flow, mass and heat transfer, electronic and ionic potential fields, as well as the chemical and electrochemical reactions. The resulting highly nonlinear, coupled system of differential equations is solved using a finite volume discretization. Our interest lies in simulating realistic operating conditions with the objective of high efficiency operation at high fuel utilization. While there are a number of studies in the literature that present multiphysics models for SOFCs, few have focused on simulating operating conditions that are necessary if SOFC systems are to realize their promise of high efficiency conversion of chemical energy to electrical energy. In this report we present simulation results at operating conditions that approach the required ranges of power density and overall efficiency. Our results include a) the temperature and composition profiles along a typical fuel cell in a SOFC stack, b) the dynamic response of the cell to step changes in the available input variables. Since models such as the one presented here are fairly expensive computationally and cannot be directly used for online model predictive control, one generally looks to use simplified reduced order models for control. We show that degradation modes such as Ni oxidation in the anode can be avoided by defining a metric that signals the onset of degradation as a function of operating conditions. We briefly discuss the implications of our model results on the validity of using reduced models for the control of SOFC stacks to show that avoiding operating regions where well-known degradation modes are activated is non-trivial without using detailed multiphysics models.

Keywords: Optimization and Scheduling, Model-based Control, Batch Process Modeling and Control
Abstract: It is quite important to improve the energy saving of air source heat pump heating system through optimal operation and system engineering method. In this work, the nonlinear dynamic process of the heat pump heating process was established and the integral optimal objective function was then formulated to balance the control accuracy and energy saving. To solve the DAOPs problem efficiently, simultaneous method was used to discretize the problem by collocation of finite elements. Computing results show that potential energy saving of more than 16% can be achieved under permitted accuracy error of zone temperature control, and the ambient air temperature and electricity price have significant effect on the optimal operation and performance of the heat pump heating system. Base on the computing results and analysis of operational parameters, a strategy to dynamically adjust the weight of objective function to balance the control accuracy and energy cost was proposed for the propose of further energy saving, computing results show that: if we dynamically adjusted the weight of objective function with this method, about 4.7% of further energy cost saving can be obtained. Our research is of significant meaning for the optimal operation of heat pump heating system under dynamical conditions

Keywords: Modeling and Identification, Model-based Control, Process Applications
Abstract: For Model Predictive Controlled (MPC) applications, the quality of the plant model determines the quality of performance of the controller. Model Plant Mismatch (MPM), the discrepancies between the plant model and actual plant transfer matrix, can both improve or degrade performance, depending on the context in which performance is measured. In this paper, we do not use performance metrics or ``yes-no"-type tests to merely diagnose the presence or absence of MPM in the plant matrix. Rather, we achieve the further goal of locating the exact MPM-affected elements within the plant matrix. Our proposed detection algorithm consists of two system identification experiments: the first experiment diagnoses the presence of MPM, and the second experiment pinpoints the exact MPM-affected elements. We then exercise the algorithm on artificial 3x3 and 5x5 plants suffering from sparse MPM, and demonstrate the algorithm's capability of correctly locating the MPM-affected entries.

Keywords: Model-based Control, Process and Control Monitoring
Abstract: Model Predictive Control (MPC) is a class of control systems which use a dynamic process model to predict the best future control actions based on past information. Thus, a representative process model is a key factor for its correct performance. Therefore, the investigation of model-plant-mismatch effect is very important issue for MPC performance assessment, monitoring, and diagnosis. This paper present a method for model quality evaluation based on the investigation of closed-loop data and the nominal complementary sensitivity function. The proposed approach ensures that the MPC tuning is taken into account in the evaluation of the model quality. A SISO case study is analyzed and the results show the effectiveness of the method.

Keywords: Process and Control Monitoring
Abstract: Model maintenance is the most time-consuming and cost-intensive in industrial model predictive control. In this paper, a drill-down diagnosis algorithm for deficient models of industrial MPC via a model quality index (MQI) is proposed. The CVs with poor models can be detected first by MQI values with all controlled variables. Then, a leave-one-out algorithm is proposed to further diagnose which sub-models are deficient for the CVs with poor model performance. Thus, the effort and cost of model maintenance can be reduced. The application result to the Wood-Berry distillation column process indicates the effectiveness of the proposed assessment method.

Keywords: Process and Control Monitoring
Abstract: A modern complex industrial process, such as petroleum refinery, petrochemical plant, pulp & paper process and power plant, usually include hundreds or thousands of control loops. It is a well-known fact that many controllers in a process plant are not tuned properly (Desborough and Miller, 2001). Badly tuned controllers would lead to loss in production as well as quality (Bialkowski, 1993; Ender, 1993). Therefore, it is necessary to detect the controllers that are poorly tuned and diagnose their behavior as sluggish or aggressive (oscillatory) and make the operators aware of it so that appropriate retuning actions can be initiated. There are methods in the literature for diagnosing aggressive (Thornhill and Horch, 2007) or sluggish (Hugglund, 1995; Kuehl and Horch, 2005) controller behavior. However, each of these methods has its own limitations and none of them addresses the improper controller tuning issues (sluggish and aggressive behavior) in a unified framework. Here, we have proposed a non-invasive method to automatically detect the badly tuned controller and identify its tuning issue as sluggish or aggressive directly from the routine plant operation data. The effectiveness of the proposed method is demonstrated on both simulated as well as industrial control loop examples.

Keywords: Process and Control Monitoring
Abstract: In most chemical processes, variables are sampled at different rates which brings great challenges to traditional process monitoring methods that are built upon single sampling rate. In this paper, a multi-rate partial least squares algorithm is proposed. Compared to the traditional PLS method, the proposed algorithm takes use of the incomplete data samples through a modification of both of the covariance matrix of the input dataset and the covariance matrix between the input and output datasets. Iteration is used in the model training step to avoid the calculation of same parameters which requires complete training datasets. Then the fault detection and online prediction strategy is proposed based on this algorithm. A case study on TE process shows that the proposed method had an enhanced performance on both monitoring and online prediction, compared to the traditional PLS method.

Keywords: Process and Control Monitoring
Abstract: Many industrial manufacturing processes have multiple operation modes because of different strategy and varying feedstock. The traditional statistical process monitoring tools such as PCA and PLS cannot be applied since they assume that the process must have single mode operation region only. In this paper, all the factors that will affect the change of the mode are considered, a similarity factor including the similarity factor of PLS models and the mean shift of the external variables is introduced to measure the similarity of two sets of data. On basis of this similarity factor, a moving window is used and a mode identification approach for multimode process monitoring is proposed. The proposed approach is demonstrated on the benchmark Tennessee Eastman process.

Keywords: Process and Control Monitoring, Modeling and Identification, Batch Process Modeling and Control
Abstract: This study proposes a new robust parametric identifier for systems that describe the dynamical behavior of metabolic networks. This identifier implements a robust parallel differentiator that recovers the time derivative for all the metabolites involved in the metabolic network. The differentiator is based on the well-known Super-Twisting Algorithm which is applied over the variation of each metabolites that is included within the metabolic network. The derivatives are fed into a parallel nonlinear least mean square scheme that is successful in recovering the parameters that characterize the metabolic network. This identifier is applied to a simplified 22-reactions metabolic network of hydrogen production in Escherichia coli using glucose as substrate. The metabolic network is simulated with parameters obtained from previous studies and they are recovered using the parametric identifier proposed in this study. All the parameter are recovered with less than 5% error.

Keywords: Modeling and Identification
Abstract: In this paper, a novel identification algorithm for the non-uniformly Wiener systems with dead-zone nonlinearities is proposed. Firstly, a uniform model of the Wiener system is reformulated with the help of the lift technique and a switching function. Then, a least squares based iterative recursive algorithm with variable forgetting factor is presented using the auxiliary model and iterative method. Simulation results indicate the effectiveness of the proposed method.

Keywords: Modeling and Identification
Abstract: System identification (SI), especially from small samples, is a challenging problem and of interest in several applications. Standard prediction-error minimization methods (PEM), under these conditions, generally result in estimates with higher variance. Moreover, in the identification of parametric models, one often needs prior knowledge of the input-output delay, obtaining estimates of which, is not possible using classical methods when the delay is either comparable or greater than the sample size. In this work, we develop a compressed sensing (CS)-based method for identifying emph{sparse} equation-error models that includes both emph{auto-regressive eXogenous} (ARX) and emph{AR moving average eXogenous} (ARMAX) structures with large delays, small orders and small delays with large orders, but with missing coefficients. The outcome is an iterative basis pursuit de noising (IBPDN) algorithm for solving non-linear CS problems. In addition, we propose a semi-rigorous method to lower the mutual coherence of the regressor matrix so as to obtain lower variance parameter estimates with the CS techniques. Errors in parameter estimates are computed using the bootstrapping method. Simulation studies on three diverse examples are presented to demonstrate the efficacy of the proposed methodology.

Keywords: Modeling and Identification, Process Applications
Abstract: In actual chemical processes, the fact that some essential variables cannot be directly measured makes the production quality out-of-control and even results in large economic losses. In this study, a novel sample clustering extreme learning machine (SC-ELM) model is developed to achieve timely and accurate measurement. SC-ELM is a fast training algorithm with an excellent generalization performance, and the combined sample clustering approach solves the non-optimal input weights of ELM. The network structure is designed by a fast leave-one-out cross-validation (FLOO-CV) method. Meanwhile, the validity of SC-ELM model is firstly tested by two classical regression datasets. With the comparison of other ELM models, SC-ELM is proved to be an effective model in both modeling accuracy and network structure. Then, SC-ELM is applied in measuring the quality index of a high-density polyethylene (HDPE) process running in a chemical plant, and the experiment results demonstrate that SC-ELM model can achieve quality estimation with higher measuring accuracy and less training time.

Keywords: Optimization and Scheduling, Model-based Control, Energy Processes and Control
Abstract: The schedule dispatch of batteries and operation of photovoltaic(PV) systems can create significant economic benefits for residents. This paper presents a model predictive control method to optimize the charge and discharge of batteries and operation of a PV system. The objective is to minimize the bill cost for customers with the prediction information of solar power generation, load consumption and unit electricity price from grid. Experiment results taken in Nansha, Guangzhou, China are analyzed and prove the creditability of the model and effectiveness of the optimal operation method.

Keywords: Model-based Control, Optimization and Scheduling, Modeling and Identification
Abstract: This paper gives a concise review of work on ensuring feasibility and stability within linear MPC during target changes. A summary of similarities and weaknesses of the various proposals in the literature is presented and this forms a useful baseline for establishing where improvements can be made.

Keywords: Process and Control Monitoring, Model-based Control
Abstract: By combining nonlinear control theory and constructive analysis, we address the control of dynamical systems with a cascade structure where the first subsystem is connected in series with the second subsystem, while the input acts in parallel on both subsystems, which may exhibit multiple steady states with bistable behavior, typically undesired in practice due to safety or operational conditions. The resulting regulation scheme is a cascade control structure which stabilizes the second subsystem within a desired region of the state space while guaranteeing the output regulation in the first subsystem. The performance of the proposed control is evaluated and compared to nonlinear single-input-single-output controllers in a simulation example of a chemical reaction taking place in two reactors connected in series.

Keywords: Batch Process Modeling and Control, Model-based Control, Modeling and Identification
Abstract: The Bisphenol A concentration at the reactor outlet is a direct response to the production quality index, and online soft measurement of Bisphenol A concentration is very necessary. Based on the mechanism of Bisphenol A condensation process, the mechanism equations describing the dynamic behavior of reactor are established, and the state equation and observation equation of Bisphenol A are derived through simplification and derivation of the soft measurement model, and then an Unscented Kalman Filtering is adopted to estimate the Bisphenol A concentration. Simulation results confirm that the method is feasible and effective.

Keywords: Model-based Control, Optimization and Scheduling
Abstract: The Distributed Model Predictive Control (DMPC) has been more and more popular in the control of distributed systems which are composed by many interacted subsystems. The range of subsystems that each local Model Predictive Control (MPC) optimized, called coordination degree, plays an important role in improving the optimization performance of entire closed-loop system. In this paper, the N-step adjacent structure matrix based decomposition method was proposed, where the coordination degree of each subsystem is determined by the union of the all the adjacent matrices over the predictive horizon. Based on this decomposition, each local MPC considers the cost of all the subsystems it impacted on during the predictive horizon, and then improves the optimization performance of entire system with reduced communication burdens. The simulation results show the effectiveness of the proposed method.

Keywords: Modeling and Identification
Abstract: In the fiber production process, the stretching process plays a key role in the quality of the final fiber product. Due to the fiber stretching process with inborn nonlinearity, the performance of a single controller and an optimizer may be compromised or even unsatisfactory. Thus, we consider a multi-model identification method for the fiber stretching process. The dynamic transitions among different operating points are achieved by the change of the operating conditions in the fiber stretching process. To excite all of the nonlinearity character in the fiber stretching process, the transitions among different operating conditions is achieved. The structure of each sub-models, operating points, operating range are assumed. Based on the input output data of the process, a linear parameter varying (LPV) model is built by applying a probability identification method. To achieve the smoothly connected among the different operating conditions, an exponential function is used. Then a global LPV model is constructed by synthesizing the local models. Simulated results show that the LPV method has the effectiveness in solving the inherent nonlinearity of the fiber stretching process.

Keywords: Batch Process Modeling and Control, Optimization and Scheduling
Abstract: Optimal control has wide applications for the control of batch and semi-batch processes to develop an optimum control input policy by extremizing a performance measure. The deployment of optimal control relies heavily on the accuracy of the process models being used for computation of the optimal profile. Often, the process models do not replicate the plants due to various shortcomings such as assumptions made during model formulations, poor first principles knowledge and limited range of experimental data due to short process development cycles. Moreover, scale-up of the processes from lab to manufacturing scale renders the developed models obsolete. The estimated model parameters can significantly differ from their nominal values which calls for the development of a strategy that updates process models so as to achieve an improved and tight control of batch processes. In this paper, we propose a novel methodology based on iterative learning to gradually update models using on-line measurement data at the end of each successive batch run by minimizing the error between plant and model data. In the proposed methodology, we further integrate ILE with optimal control to update the optimal control input profile with the advent of measurement after each successive batch run. An important aspect of this integration is to ensure that model updates between batch runs generate feasible optimal control trajectories. Simulations are performed for the temperature control of a batch reactor system to validate the proposed methodology.

Keywords: Modeling and Identification, Model-based Control
Abstract: This paper proposes an auxiliary model based hierarchical least squares algorithm for multivariable Box-Jenkins-like systems using the hierarchical identification principle. To improve the computational efficiency, a multivariable system is decomposed into two subsystems by using the data filtering technique. Furthermore, this paper presents a data filtering based auxiliary model hierarchical least squares algorithm for multivariable Box-Jenkins-like systems. The simulation example shows that the proposed identification algorithms are effective.

Keywords: Modeling and Identification, Model-based Control
Abstract: This paper presents a multi-innovation stochastic gradient algorithm for an input nonlinear state space system by deriving the identification model of the system and by decomposing the model into two sub-models. The basic idea is to design a state observer to estimate the unmeasurable states and to estimate interactively the unknown parameters of two subsystems by using the hierarchical identification principle. The simulation results show that the proposed algorithm is efficient.

Keywords: Modeling and Identification, Process and Control Monitoring
Abstract: Control loop performance assessment procedures are established as key cornerstone of process optimization and monitoring. Unfortunately, many methods do not consider the maximum possible performance which can be achieved by the fixed structure controller installed in the loop (typically PID). In authors' recent work such method was described. It was shown that -- with respect to classical minimum variance method -- the maximum performance is strongly influenced by the process normalized dead time. Consequently, the process model should be re-estimated during the plant operation. This paper describes technique which integrates continuous estimation of process model including normalized dead-time and control loop performance assessment. The process model is considered as fractional in order to fit well to distributed parameter systems appearing in chemical process control industries. All procedures are packed into advanced function blocks which are ready for real-time applications.

Keywords: Model-based Control, Modeling and Identification, Process Applications
Abstract: In the last decades, there is a boom of fractional calculus (FC) applications in many technical areas including process control. The multiple fractional-order pole models (MFPM) were proven as suitable for modeling processes that are essentially monotone from their physical nature (holds true for a big percentage of chemical processes). Due to implementation aspects, the time-domain real-time simulation of MFPM is a challenging problem. For this purpose, often a high order zero/pole transfer function is employed. This paper describes a novel multiple fractional pole simulator based on exact step response discretization. The proposed technique decreases computational burden and the resulting simulator can be deployed even to real-time embedded devices. The advantages of SW tools developed are demonstrated on various examples including predictive control feedback loop simulation.

Keywords: Process and Control Monitoring, Model-based Control
Abstract: This work presents a methodology for the integrated identification, estimation and accommodation of control actuator faults in particulate processes with discretely-sampled measurements and plant-model mismatch. Initially, a stabilizing state feedback controller is designed on the basis of a reduced-order model of the infinite-dimensional system, and the closed-loop stability region is characterized in terms of the model uncertainty, the fault magnitude, the sampling period and the control design parameters. When state measurements are unavailable, the reduced-order inter-sample model predictor generates state estimates which are updated at each sampling time. A moving-horizon optimization problem is then formulated and solved for on-line actuator fault detection, isolation and estimation using past state and input data. The resulting estimates are used to locate the operating point with respect to the closed-loop stability region, which in turn is used to carry out the fault accommodation logic via updating the pot-fault control model and/or adjusting the controller design parameters. The developed methodology is illustrated using a non-isothermal continuous crystallizer example.

Keywords: Model-based Control, Energy Processes and Control
Abstract: In this paper a reliable H_∞ controller for a class of uncertain switched linear systems is designed using the multiple Lyapunov function technique with the property that the solvability of reliable H_∞ control of the individual subsystems is not necessary. First, reliable H_∞ control is defined for this class of systems and the multiple Lyapunov function method is deployed to ensure that the magnitude of the controlled output is upper bounded by suppressing the external disturbance while guaranteeing internal stability. Last, the H_∞ control method above for the switched systems is directly applied to the standard reliable H_∞ control problem of a class of non-switched systems which a continuous control method cannot solve.

Keywords: Model-based Control
Abstract: In this paper, an observer-based model predictive control (MPC) strategy is presented for distributed parameter systems (DPSs). First, principal component analysis (PCA) is used for dimension reduction by transforming the high-dimensional spatio-temporal data into a low-dimensional time domain. Then an observer is builded to estimate the low-dimensional temporal output using the real-time measurable spatiotemporal output. Finally, the MPC strategy is proposed based on the low-dimensional estimation models. Simulations demonstrate the accuracy and efficiency of the proposed methodologies.

Keywords: Modeling and Identification, Batch Process Modeling and Control
Abstract: In this paper a recursive identification method is proposed to simultaneously estimate the parameters of an output error model together with a delay parameter for industrial processes with time delay, to facilitate model based control design. An extended observation vector is constructed to establish a recursive least-squares (RLS) identification algorithm. An auxiliary model is established to estimate the unknown noise-free output of the system for consistent estimation. Moreover, a variable forgetting factor (VFF) is introduced to enhance the convergence rate and identification accuracy against measurement noise. The convergence of the proposed algorithm is analyzed with a strict proof. An illustrative example is shown to demonstrate the effectiveness of the proposed identification method.

Keywords: Optimization and Scheduling, Modeling and Identification, Process Applications
Abstract: A new method for the optimal design and control of chemical processes under process disturbances and parameter uncertainty is presented. The key idea is to back-off from the optimal (dynamically inoperable) steady-state design by solving a series of optimization sub-problems constructed using Power Series Expansion (PSE) functions. The PSE functions estimate the actual constraints and cost functions’ worst-case variability due to disturbances and parameter uncertainty. The proposed method was applied to address the optimal design of a non-isothermal CSTR. The results show that this method has the potential to address the optimal design of dynamic systems at low computational costs.

Keywords: Optimization and Scheduling, Process Applications, Process and Control Monitoring
Abstract: Obtaining reliable models from experimental data is a point of deep interest in all areas of research. Since the quality of the model depends on the number of selected variables, it is important to develop methods that identify the best ones. This work proposes a method of variable selection based on the Ant Colony Optimization (ACO) algorithm. Using data from a Saccharomyces cerevisiae fermentation, several criteria for trail update and model comparison were implemented and the obtained models were compared. The use of the length of the confidence interval produced the best results, finding the optimal model more frequently.

Keywords: Optimization and Scheduling
Abstract: Due to ever-challenging global market conditions, plant economic optimization is becoming more critical. Recent advances have transformed the traditional steady-state real-time optimization (RTO) system of plant economic optimization to dynamic real-time optimization (DRTO) based on a dynamic prediction model. DRTO strategies that have been proposed perform economic optimization in an open-loop fashion without taking into account the presence of the plant control system. In this work, we propose a bilevel programming formulation for DRTO that includes effects of the closed-loop dynamics of an underlying constrained model predictive controller (MPC). The bilevel program is subsequently reformulated and solved as a single-level mathematical program with complementarity constraints (MPCC). We investigate the economics and control performance of the proposed strategy under varying MPC controller design parameters, and compare them to open-loop DRTO and rigorous multilevel closed-loop DRTO approaches.

Keywords: Batch Process Modeling and Control, Model-based Control, Process and Control Monitoring
Abstract: In this work, we develop a run-to-run (R2R) model parameter estimation scheme based on moving horizon estimation (MHE) concepts for the modeling of batch-to-batch process model parameter variation using a polynomial regression scheme in a moving horizon fashion. Then, a model predictive controller (MPC) with the proposed parameter estimation scheme is applied to a kinetic Monte Carlo (kMC) simulation model of a batch crystallization process used to produce hen-egg-white (HEW) lysozyme crystals. The average crystal shape distribution of crystals produced from the closed-loop simulation of the batch crystallizer under the MPC with the proposed R2R model parameter estimation scheme is much closer to a desired set-point value compared to that of MPC based on the nominal process model.

Keywords: Modeling and Identification, Model-based Control, Process Applications
Abstract: The necessity is greater than ever for a methodology to both diagnose at early stage and evaluate the progression of Parkinson Disease. In this paper, we propose an interesting and innovative methodology for pattern recognition based automated individual-level clinical diagnosis of Parkinson Disease. It makes use of a unique combination of machine learning tools and statistical tools. The methodology comprises of three major steps. First, pre-processed brain Magnetic Resonance Images are modelled using Self- Organizing Map for feature generation. Second, Fisher-Discriminant Ratio is used to reveal distinctive feature(s). Third, Least Squares Support Vector Machine is used for Individual-level patient classification. The applicability of the proposed methodology has been demonstrated using 831 T1- weighted Magnetic Resonance Images obtained from Parkinson's Progression Markers Initiative database. We have achieved classification accuracy of up to 97% for differential diagnosis of Parkinson Disease with confidence interval of 99.9%. This method is particularly suited for diagnosing patients in early stages of the disease, i.e., patients in age of 31- 60 years. In the present landscape, Brain Magnetic Resonance Image is routinely performed to assist Parkinson Disease diagnosis in clinical settings. Thus, the induction of the proposed methodology as a decision support system could make a significant impact on treatment strategies especially by aiding early-stage disease diagnosis.

Keywords: Optimization and Scheduling, Model-based Control
Abstract: Minimum entropy control has been proven to be an effective method in control of non-Gaussian stochastic systems. In this case, the entropy is proposed as a generalization of the variance measure to characterize the randomness of the process. Minimum entropy corresponds to small uncertainty (or derivation), but it cannot guarantee the tracking error approaching to zero. Therefore, mean square error also should be added in the criterion. In this paper, by using a simple example, the method of generating a representative approximation of the Pareto optimal control set is investigated in both analytical and numerical ways. And simulation results show the feasibility of the proposed double-objective optimal control method.

Keywords: Batch Process Modeling and Control, Model-based Control
Abstract: A robust economic nonlinear predictive controller is proposed for batch biochemical processes. The algorithm is based on a dynamic metabolic flux modeling (DFBM) approach that has gained increasing acceptance in the pharmaceutical industry. Since DFBM models are formulated as LP (linear programming) problems, the corresponding robust economic optimal control solution results in a bi-level optimization consisting of maximizing an economic objective subject to the LP based model. To address robustness the DFBM model is represented by using a Polynomial Chaos Expansion model which allows quick propagation of the uncertainty in parameters onto the cost function to be optimized by the controller. The performance of the robust algorithm is applied to E.coli batch fermentation and it is found to be computationally efficient and superior to the nominal (non-robust) predictive controller.

Keywords: Optimization and Scheduling, Batch Process Modeling and Control
Abstract: A single parameter based fast mesh-sorting is proposed in this work. The single parameter algorithm as compared to non-dominated sorting eliminates the classification of the population into non-dominated fronts and calculating crowding distance. The proposed one parameter approach also provides flexibility of choosing any probability based selection operator. On the other hand, non-dominated sorting approach can only use tournament selection directly. We have considered Zitzler-Deb-Thiele’s (ZDT) test functions to test computational and convergence capabilities of proposed algorithm. The performance of the proposed algorithm is compared with conventional Non-dominated Sorting Genetic Algorithm-II (NSGA-II) and NSGA-II with a recent fast corner-sort algorithm. We have also considered optimal control of fed-batch reactor as Multi-objective optimization application.

Keywords: Modeling and Identification, Process and Control Monitoring
Abstract: An integral equation approach to estimate the parameters and the delay term of fractional order (FO) continuous-time models is proposed. Data from a single step response is used for the estimation assuming that the fractional orders are known. In the proposed methodology the FO differential equation is transformed into a FO integral equation. A least-squares equation is then formulated by expanding an integral term that allows the implicit time delay term to have an explicit appearance in the parameter vector. This facilitates direct estimation of the time delay simultaneously with other model parameters. The performance of the proposed method is demonstrated using simulation results. Identification results of a simulated distillation column are presented to demonstrate the applicability of the method.

Keywords: Model-based Control, Optimization and Scheduling, Energy Processes and Control
Abstract: This paper presents an asynchronous distributed self-triggered MPC controller design strategy for large-scale constrained linear systems. Based on the so-called relaxed dynamic programming (RDP) inequality, at a triggered time that corresponding to a subspace of the whole system, the synthesis procedure allows us to locally determine both the updated MPC control action and the next triggering time for this subsystem. The subsystems update their inputs asynchronously, while the resulting self-triggered MPC control law can still preserve stability and constraint satisfaction. The derived conditions can be adapted to robust distributed event-triggered MPC implementation with continuous monitoring of the systems, when the system model is uncertain.

Keywords: Modeling and Identification
Abstract: Apoptosis – programmed cell death – is a key component in cell renewal, i.e. for removing damaged cells or those not any more required without harming their environment. Understanding apoptosis is challenging since the involved signalling network is complex and intertwined. We use linear control theory to shade light on the complex signalling mechanisms and interactions. To this end we linearise different models around the steady state corresponding to the unstimulated case and use linear control analysis to uncover and identify a decentralised control scheme with two plant and two controller modules that are sparsely interconnected. While this separation on a first view is artificial, it allows for using the small gain theorem to analyse stability and to give insight into how inhibitors naturally modulate cell death. This sheds light onto the role of positive and negative feedback present in this signalling pathway. Sensitivity analysis reveals these most influential parameters on the time of death. To illustrate the analysis, we examine in simulations to which extent apoptosis can be reversed once initiated.

Keywords: Process and Control Monitoring, Process Applications, Optimization and Scheduling
Abstract: Detecting the onset of overloading in a semi-autogenous grinding (SAG) mill is a challenging task for operators to perform due to the complex and nonlinear nature of an overload. To detect an overload, operators must simultaneously monitor the correlations between several measurements of the SAG process. However, overloading often goes unnoticed at its early stages because the subtle changes in the correlations between measurements are difficult for an operator to observe. In addition, linear process monitoring techniques such as principal component analysis (PCA) provide inconsistent results with overload detection because of the process nonlinearity. Recently, locally linear embedding (LLE) with a linear classifier has been proposed to detect the early onset of an overload in a SAG mill. In this paper, we compare the suitability of LLE to detect the early onset of an overload against kernel PCA and support vector machines.

Keywords: Process and Control Monitoring, Process Applications, Energy Processes and Control
Abstract: In this paper we discuss the negative impact on monitoring algorithms of working with imprecisely dated data. Two examples from the world of the oil & gas industry are presented and serve to illustrate that this problem can be of practical importance. First analytical results show that when signals with significant time variations are monitored, the impact of dating of measurements can be as troublesome (or even worse) than measurement noises.

Keywords: Model-based Control, Optimization and Scheduling
Abstract: We interpret economic MPC as a scheme that trades off economic performance and stability. We use this notion to design an economic MPC controller that exploits the inherent robustness of a stable auxiliary MPC controller to enhance economic performance. Specifically, we incorporate a flexible stabilizing constraint to the economic MPC formulation that preserves stability of the auxiliary controller. We use multiobjective optimization concepts to argue that the dual variable of the stabilizing constraint can be interpreted as a price of stability and we establish an equivalence between the proposed controller and regularized economic MPC controllers. We demonstrate that nontrivial gains in economic performance can be achieved without compromising stability.