Application Of Fuzzy Control In The Temperature And Humidity Control For Intelligent Greenhouse

For meeting the objective of" Study on Adaptive Control Based on Fuzzy Neural Networks of Greenhouse (No. 2004AA1CG130) ", which is one of the key research projects in Science and Technology in Harbin. by analyzing greenhouse executing equipment cause the influence for environment factor, Applying fuzzy control , neural network, computer technology sensor technology to design and research the control system of greenhouse, and developing a intelligence fuzzy control system of greenhouse. Main research contents and results are listed as follows:Firstly, this paper introduced the current status and direction of the control system of greenhouse in the world and in this country. And analyzed in detail the temperature and humidity's control characteristic in the greenhouse system of the automatic control, On the basis of analyzing the target's complexity in the greenhouse, designed overall the environmental control system of the greenhouse.Secondly, this paper presented a kind of fuzzy neural network method which combined fuzzy knowledge representation with neural network self-learning ability and adopted a fast study algorithm, Fuzzy Neural Networks Controller (FNNC) was designed, which combined the advantages of neural networks and fuzzy logic to improve the teaming and controlling performance of the whole system. Fuzzy Neural Networks Controller not only can process fuzzy information and finish reasoning function, but also the fuzzy neural networks were trained continuously by inputting specimen data;then the membership function parameters and the weights of fuzzy logic rules were optimized by using back propagation algorithm. The parallel processing network made the self-adaptation of the membership functions and the self-organization fuzzy logic rules possible.Thirdly, this paper not only put forward a kind of fuzzy neural network controller, but also realized it from hardware and software. In this system, as master computer, PC can get signal that is send by lower computer, the parameter can also be modifiable at the same time. The lower computer used American Cygnal Corporation the C8051F040 single- chip, around C8051F040 single-chip which is reinforce by data acquisition module;real-time clock module;and communication interface module and so on. Software contained two parts: master computer software and lower computer software. The former adopts VC++6.0, Windows 2000 as operation platform. The latter adopts C language.Finally, the simulation of using FNNC to control the greenhouse environment was carried out. Compared with PID control and fuzzy control, the simulation results showed that FNNC has lighter exceeding suiting well to control, no shaking, fine stationary, short time to reach thestable state, little error of the stable state. So its dynamic characteristic and static characteristic is the most superior. Therefore it has verified that the feasibility of using FNNC.