Research On The Modelling And Intelligent Control Of The Greenhouse Microclimate

Modern greenhouse is to create the most suitable environment for crops cultivated inside such that the goal of high-quality, high-quantity, low-cost, best-benefit at harvest can be reached. The control of the greenhouse microclimate is one of the key techniques in modern greenhouse and becoming increasingly important.The paper proposed the method of the modelling taking into account the intelligent control of the greenhouse environment, based on the balance of the energy and mass. The energy and mass exchange physical processes resulted from radiative, ventilation, convection and crop transpiration in the greenhouse is discussed. The model of the temperature and humidity is built respectively. Through experiments in an intelligent greenhouse in the Zhejiang Province Hi-Tech demonstration garden of agriculture, the relevant data that the model needs was gathered. Through simulation and analyse of the error between the simulated value and the measured value according to the root mean squares error (RMSE) standard, it was shown that the dynamic model can provide reliable estimates of both temperature and humidity in the greenhouse. As an application of the model, the environment condition in greenhouse and the capacity of the control equipment can be analysised by the model simulation in the computer, instead of the inconvenient measure method in the experiments.Interactions between the internal and external variables, the complexity of the phenomena (multivariable, nonlinear, nonstationary), and the complexity of the process model are such that it is often difficult to implement the conventional techniques of regulation. The fuzzy control of greenhouse microclimate is proposed. According to the interdependence between inputs and outputs and the couplings in the multivariable system, a decentralized control structure including two robust fuzzy controllers is developed. Simulation results show the control algorithms can drive to set point valuesand obtain the good performance, except the little disturbance around the set point values.But one fuzzy controller need a lot of parameters, the expert's experience only play a guide role and is difficult to make sure every parameter accurately. The fuzzy controller optimization through genetic algorithms (GA) is developed. During the GA optimization of Gaussian input membership functions for the error and the change-in-error of the temperature, the good performance control curve line of the greenhouse temperature is obtained. Compared with the basic fuzzy control, it can obtain the better performance and improve robustness.