The Application Of Fuzzy Neural Networks In Limekiln Control System

In the control of industry process, we can't get the exact model of controlled object or the model will change with the external disturbing, so we can't get satisfying control effect with the conventional control method. The appearance and development of intelligent control technology provide the effective approach for solving this kind of problem. Fuzzy Neural Networks Control (FNNC), which combines the neural networks with the fuzzy logic and the traditional expert system is applied in this paper. When fuzzy logic is used for control, it doesn't need an accurate mathematics model of the controlled system. The neural networks have many advantages such as nonlinear map, self-learning, distributional memory, parallel processing and so on. So the combination of fuzzy control and neural networks can make it possible that the membership function and the fuzzy rules from the traditional expert system can be converted into the fuzzy neural networks. Then the dispersed knowledge system is formed. After adjusting the weight of the neural networks continuously, the membership function and the fuzzy rules are more accurate. The main contents of the research in this paper are as follows: 1. A survey of the origin and the development status of fuzzy logic theory and neural networks theory is summarized. An introduction of the task background and the combination of fuzzy logic and neural networks. 2. An expatiation of the equivalence of fuzzy system and the neural networks. An introduction of the realization of fuzzy neural technology. 3. An introduction of the typical fuzzy reasoning system, especially fuzzy RBFN. An expatiation of the realization of neural networks based on fuzzy rules. 4. An introduction of the control scheme of limekiln. An expatiation of the calcining temperature control of fuzzy neural networks. 5. The analyses of mathematic model and dynamic characters of limekiln control system. In the simulations using PID control, fuzzy control and fuzzy neural networks control. The results show that the fuzzy neural control method is better than the other methods. 6. An introduction of software and hardware constitution and design of limekiln control system. Fuzzy neural networks are used in the control of calcining temperature. The anticipative effect is achieved.