Deficit irrigation-based growth control of maize plants using hybrid model predictive / trellis decoding algorithm

Abstract

Deficit irrigation approaches have been applied in mitigation of pressure on global freshwater availability. Data from growth of plants under different water stress conditions allows the development of growth models taking into account irrigation scheduling. In this work, a model predictive control approach combined with a modified trellis decoding algorithm is applied to control growth of maize plants under deficit irrigation conditions by specifying the irrigation schedule required to achieve specific total leaf length at the end of a specified growth period. The brute force-based algorithm allows the control of growth to within 8.5 % of the desired final total leaf length over a growth period of seven days. The designed control algorithm is an adaptable solution that can be applied to different optimization goals related to plant growth and water consumption, and allows the user to evaluate best-case and worst-case scenarios in advance, which supports decision making with regard to irrigation decisions.