Modern Greenhouse Humidity System Modeling And Controlling Based On Mixed Logical Dynamical

Facility agriculture refers to a new mode of agricultural production, which uses engineering technology methods to effectively regulate the greenhouse microclimate environment, so that the crop can grow in an optimal condition. Recently, after years of development, Chinese facility agriculture has made considerable progress and the horticultural area has been ranked first in the world. But there is still a large gap compared with developed countries. As a typical representative of facility agriculture, the modern greenhouse integrates the information technology and biological technology to agricultural production. According to the characteristics of the planted crops, the best growth environment can be controlled. The modern greenhouse is safety with high-yield and good quality. Therefore, further study on the modern greenhouse facility agriculture has important strategic significance on promoting national economic development, enhancing the level of agricultural science and technology, speeding up the adjustment of agricultural industrial structure upgrade.Greenhouse microclimate modeling and controlling is an important part of greenhouse environment control. Reasonable model can simulate the environment inside the greenhouse effectively and provide a rational gist for environment control. Nowadays, most studies on greenhouse temperature and humidity system are using the theory of continuous system, without considering the characteristics of the equipment in China, which have the only on and off state. This dissertation proposes to use the hybrid system theory to solve the problem. Based on the system identification theory, the mixed logical dynamical method is introduced into greenhouse humidity system modeling and the system is controlled using the predictive control based on MLD.The main contents of this dissertation is as follows:firstly, it studies the software and hardware of the greenhouse, with the main function of the measurement and control system being described. Then the humidity which poses a significant effect on greenhouse plant diseases and production is taken as the main research object. Based on the water vapor balance, combining with some physiological processes, like transpiration, taking the greenhouse structure, material, inside temperature, outside temperature, radiation, inside air flow and wind speed as inputs, the dissertation introduces the mechanism model of humidity system, which is also analyzed in the last part.Based on the mechanism model, the sensors are used to collect the environmental data for system identification. Based on the data of different seasons, the outdoor relative humidity and outdoor temperature, solar radiation intensity and outdoor wind speed are taking as inputs, the indoor relative humidity is as outputs, the Akaike’s Information Criterion is used to identify the model structure, by means of the fading memory recursive extended least square algorithm, the model parameters is identified. Finally through the experiment, the high order model and the low order model are compared and the error indicator is analyzed.The mixed logical dynamical modeling framework is used for greenhouse humidity system modeling. By means of the conversion relationship between propositional logic and the linear inequality, the mixed logical dynamical model is got, and the simulation and control which uses the predictive control based on MLD model will be proceeded. The optimal setup value and minimize the equipment switch times are taken as index and the Branch&Bound algorithm is used to solve the switch sequence. The experimental results show the applicability of the model and the rationality of the control method.Finally the dissertation summarizes the full text, concludes the completed work and proposes the further research in the modeling and controlling of greenhouse environment field.