Co-state Variables as Strategic Set-point Generators for Greenhouse Environmental Control

Abstract

Growers need sometimes to adjust the growth course of their crop. This, however, involves many set-points whose interaction is complex. In this study it is suggested that the grower only changes a few strategic-level controls, namely the co-states of the most important system states (biotic and abiotic), while leaving the tactical level, namely the set-point and flux adjustments, to the computer. It is demonstrated that while the crop model may be inaccurate and the weather unexpected, a corrective course may be estimated based on the accumulated observable information. The correction is accomplished by generating pairs of 'equivalent' trajectories, where changing co-states compensates for erroneous information or unexpected situations. The same general approach also applies to short term processes such as day-to-night heat storage.