FITTING THE MBM-A MODEL OF PLANT GROWTH TO THE DATA OF TOMGRO: IMPLICATION FOR GREENHOUSE OPTIMAL CONTROL

The MBM-A (Modified Botanical Model-Adjusted) is a two state variable, three growth stage crop model, adjusted for greenhouse optimal control.
The investigated problem is: does the simple MBM-A model mimic and predict the approximately optimal trajectory of a very complicated plant growth model like TOMGRO? While TOMGRO has about 50 parameters and 71 state variables, and can not be directly used for the purpose of e.g. optimal greenhouse control, MBM-A has only 5 parameters and two variables.
Daily data from TOMGRO during the given 8 months long growing season are used to extract parameters for the MBM-A model.
A stage oriented process is presented, where the optimal control problem associated with the MBM-A model is solved, and a corresponding set of co-state variables are used for the optimization of the TOMGRO trajectory.
A permanently oscillating climate was assumed in this study.
The switches between growth stages are determined in terms of effective degree (°C) -days whereby the length of the season in days from planting (DFP) is fixed.
The optimal growth of the greenhouse plant is assumed to be balanced, i.e. the source activity (daily accumulated dry matter) is balanced with the sink demand controlled by the effective temperature.
We show that the MBM-A parameters can be successively extracted from the data of TOMGRO by the proposed stage-oriented process.

International Symposium on High Technology for Greenhouse System Management: Greensys2007

I. Ioslovich, P.-O. Gutman

crop models, crop growth, greenhouse, optimal control, control intensity

https://doi.org/10.17660/ActaHortic.2008.801.57