Fuzzy Adaptive Control And Its Application Of Water Temperature

Solar Water Heater System (SWHS) is a time-variant,nonlinear and complex system,how to implement the intelligent control of it is an important task.At present,there are some progress have been made in the field of theory research and practical application of SWHS, in these method,generalized forecast control,pid control,classical fuzzy control,uncertainty multi-objective programming were adopted, but there are some limit in them,for the parameters of the system are preseted,so they haven't adaptive activity when the system is running,while the environment has changed,the control effect is not satisfied.In this paper,we did some research in fuzzy adaptive control theory and designed a fuzzy adaptive controller to control the electric heater of SWHS, and realized the water temperature adjusted automatically of SWHS.Solar heat using system has many characters such as time variant,nonlinear and fuzziness,it's very difficult for controller to construct an accurate mathematics nodel to describe it.The effect of the conventional control strategy is not satisfied.To get the nonlinear property of the system,we use the input-output datas to identify the system and the recursive least square method with forgetting factor was adopted.When fuzzy logic is used for control, it doesn't need an accurate mathematics model of the controlled system, but linguistic variables to describe it. The neural network has many advantages such as nonlinear map, self-learning, parallel processing. So the combination of fuzzy control and neural networks can make it possible that the membership functions and the fuzzy rules can be converted into the fuzzy neural networks. Then the dispersed knowledge system is formed .by adjusting the weight of the neural networks continuously, the membership function are more accurate.The simulation of solar water heater system is carried out and compared with the conventional fuzzy control algorithm. The result showed that the fuzzy adaptive controller has many advantages, such as the speed of inference is faster; the capability of tracing is better and the steady-error is smaller.