Modeling and predictive control of greenhouse temperature-humidity system based on MLD and time-series

Aiming at the hybrid properties of greenhouse temperature-humidity control system, predictive control based on mixed logical dynamical (MLD) model was researched. Firstly, hybrid system model of greenhouse temperature-humidity system was built based on MLD, and two subsystems of greenhouse temperature-humidity system were identified by forgetting factor recursive least squares (FFRLS) under the conditions of open ventilating window and closed ventilating window, respectively. Secondly, predictive control problem was described as mixed integer quadratic problem (MIQP), in which there were measurable but uncontrollable outside disturbance inputs including outside temperature, humidity, solar radiation and wind speed, etc. Time-series models were adopted to predict the disturbance inputs, then by branch & bound algorithm, MIQP was solved to obtain an optimal switching control sequences, based on which finite-time stability of the MLD system was analyzed and simulation control results were compared between the optimal switching signal and two given ones to verify the effectiveness of the methods achieved in this paper.