Sub-optimal control of the greenhouse environment: Crop models with and without an assimilates buffer

A simplified crop model with an assimilates buffer (labelled 2S) is developed and compared to a more traditional model without a buffer (1S). The study explores the effects that the different model formulations have on the optimal control of the greenhouse environment. Values of the parameters specific to the 2S model are estimated from information in the literature. The size of the buffer is estimated in two different ways. Based on my previous modelling experience, a constant buffer co-state (virtual value of assimilates) is chosen to guide the control decisions. Its value is also estimated in two different ways: Calculated optimal values of the assimilates co-state, for various model parameter sets, turn out to be similar to values based on the proportion between assimilates and fresh fruit. Results for Dutch conditions and typical control equipment, show that the 2S optimal control strives to achieve balanced (sink = source) growth in summer, by increasing temperature and restricting CO2 enrichment relative to the 1S control, where growth is always balanced (by definition). Utilizing 2S also improves the positive correlation between mean daily tem-perature and mean daily global radiation, a correlation which could be used to guide the control in practice.(c) 2022 IAgrE. Published by Elsevier Ltd. All rights reserved.

Automated summary

A simplified crop model with an assimilates buffer is developed and compared to a traditional model without a buffer. The study investigates the effects of different model formulations on optimal control of the greenhouse environment. The results show that the model with a buffer improves growth balance and enhances the positive correlation between temperature and radiation.