Identification of a nonlinear dynamic biological model using the dominant parameter selection method

The identification of nonlinear models sometimes encounters problems because of the limited amount of available measurements in combination with a large number of uncertain model parameters to be identified. E.g., the determination of the chemical composition of a lettuce crop is a rather expensive procedure; thus the number of experimental measurements is limited. As a result, the number of parameters of the dynamic model that can be successively identified is also limited, and the subset of the parameters to be identified must be chosen in a reasonable way. ter estimation for an extended nonlinear three-state model for lettuce growth in greenhouses is presented in this paper. The varying structural nitrogen concentration and water contents are the new elements included in the model. The dominant parameter selection (DPS) method was used to select a suitable set of identifiable parameters. The resulting calibrated model predicts quite well the experimental data which also include observations with severe nitrogen stress.