Evaluation of Experiments for Estimation of Dynamical Crop Model Parameters

Planned experiments are usually expected to provide maximal benefitswithin limited costs. However there are known difficulties inoptimal design of experiments. They are related to the case whenonly limited number of parameters could be estimated, becauseavailable experiments are the non-informative. The useful methodfor this case is considered based on the dominant parametersselection procedure (DPS). The methodology is illustrated here withdata from five planned experiments related to the NICOLET lettucegrowth model. The maximal number and the list of estimatedparameters are determined while the conditional number of theinformation Fisher matrix (modified E-criterion) is kept below agiven upper constraint. \textit{Copyright}$^{\mathit{\copyright}}$\textit{2005 IFAC}.