The Research And Applications On The Intelligent Control Algorithm For The Greenhouse

According to the requirements of the requirements of Nanjing Research Institute of Vegetables and Flowers greenhouse intelligent control project, this thesis develop an intelligent greenhouse fuzzy control system by analyzing the characteristics of the greenhouse ecological environment. The system use fuzzy control technology, neural network technology, computer technology and sensor technology by combining traditional experiences of experts. The main research content and research results are as follows:This thesis first introduces the status and development trend of the greenhouse control system at home and abroad. Then it analyzes the characteristics of the temperature and humidity in modern greenhouse environment control system. This thesis designs the greenhouse environment control system on the basis of analyzing the complexity of the greenhouse charged object.Secondly, by introducing the basic concepts and composition of the principle, also as well as the respective merits of fuzzy control and neural network, this thesis studies the fuzzy neural network controller. Fuzzy neural network will not only effectively combine theory of fuzzy control, fuzzy reasoning and learning ability of the neural network, but also optimize the fuzzy membership function and neural network value by the learning algorithm specified, in order to intelligentize the greenhouse control system.Finally, this thesis tests the fuzzy control system in the greenhouse environment. The results show the feasibility of the fuzzy neural network controller in the greenhouse environment control system. The fuzzy neural network control has the following characteristics: convergence, stability, oscillation, short time to meet expectations, error control in the ideal range by learning the algorithm. In summary, fuzzy neural network controller performance the most prominent both in the dynamic and static statement.