Optimal Control-Based Neurocontroller for Crop Growth in Greenhouse

In this paper, a neuro-dynamic programming-based optimal controller is proposed to drive the crop development in greenhouse systems. The neurocontroller drives the crop growth development by minimizing a predefined cost function, which considers the greenhouse operational costs and the final state errors under physical constraints on process variables and actuator signals. In particular, the neurocontroller is applied to manage the tomato seedling development through control of the greenhouse microclimate. In the neurocontroller design process, the nonlinear dynamic behavior of the crop-greenhouse system and data from the climate in Tulum Valley, Argentina, in July 1999, are considered. The control law obtained is suboptimal due to the use of neural networks to approximate the optimal cost-to-go function. In order to show the practical feasibility and performance of the proposed neurocontroller, simulation studies were carried out for the tomato seedling development