Electric heating control of agricultural greenhouse in winter using an embedded technology based chaotic particle swarm optimization PID controller

Abstract

The agricultural greenhouse is a complicated system with changeable multi-factors. To eliminate the uncertainty of systems, intelligent algorithms are widely used to optimize the framework of the proportional, integral, and derivative (PID) controller. A particle swarm optimization with chaotic logistic mapping (CPSO) is proposed to calibrate PID parameters. The CPSO-PID is implanted into the heating control system by embedded technology (ET) to improve the energy savings and system performance of the greenhouse in winter. Computational fluid dynamic (CFD) calculates the heat and mass transfers to describe the temperature distribution. It also serves as an offline energy demand predictor to cooperate with a three-stage fan coil unit (FCU) loops online response strategy to control the heating system. The determination coefficient R² of 0.874 of the fitting results verifies that the CFD simulation reached the application level. An interference case shows the robustness of this method. In the full-scale experiments, compared with the GA-PID and PSO-PID controllers, its energy savings are 1.65 % and 8.20 % with a lower mean temperature deviation of 0.63 °C and 0.53 °C, respectively. These results show that the proposed control method can improve heating system performance with more suitable temperature, stronger adaptive capacity, faster response time, and lower energy consumption.