Hierarchical Model-Based Irrigation Control for Vertical Farms

Abstract

As high energy costs and water scarcity remain major problems in the context of vertical farming, we present a hierarchical control system developed for the optimization of irrigation processes in these farms. The main objective is to minimize the energy cost and water consumption associated with the operation of a vertical farm, with special emphasis on irrigation in a soil-based environment. At a higher level, an optimal control problem based on a dynamic crop growth model and the FAO Penman-Monteith equations is solved offline to determine the optimal daily inputs, considering temperature, radiation, and plant available water as decision variables. At a lower level, strategically placed Proportional-Integral-Derivative (PID) controllers are used to track these irrigation set-points online. These controllers are designed using a dynamic drip irrigation and soil model that simulates the movement of water through soil. In a simulation study, this integration of set-points and subordinate PID controllers shows a robust control system that effectively stabilizes the irrigation process. The results also show a uniform moisture distribution in the soil after a short time, which contributes to uniform plant growth in the farm. The proposed hierarchical control system combines computationally intensive offline open-loop optimization for daily input determination with fast and simple online PID control for real-time stabilization.