Real-time irrigation scheduling using an optimal control methodology

A systematic methodology is presented for making real-time irrigation decisons. The study uses a model to describe the physical processes in the soil–crop–climate system. The variables describing the state of crop and soil are simulated simultaneously with an integrated model. The soil moisture and salinity, which synergistically determine the water uptake by the crop, are calculated using mass balance models for the root zone. The crop yield is predicted with detailed simulation of the plant growth processes (assimilation, respiration and transpiration) which are controlled by climatic conditions. The control model is based on an analytical optimization method for multistage multivariable sequential decision making problems. The method is based on local recursive approximation of the nonlinear problem with a linear quadratic problem. This methodology is tested with examples in which optimal irrigation schedules are computed for each hour during the growth season.