Adaptive Fuzzy Control Of Solar Energy Heat Utilization System

Solar Energy Heat Utilization System (SEHUS) is a time-variant, nonlinear , complex system. How to implement the intelligent control of Solar Water Heater System (SWHS) (a subsystem of SEHUS) is an important and difficult research subject. So far, several methods such as classic fuzzy control , PID control , general forecast control and so on have been adopted, some progresses of theory research and practical application of SWHS have already achieved. However, some limits still exist in these methods. Because parameters of the control systems are preset, when the exterior environment has greatly changed, they will not change adaptively, as a result the control effect will obviously not be satisfied.Fuzzy control technology , neural network and genetic algorithm are important bases of intelligent control, the trend of modern automatic control. They have their respective advantages. Fuzzy control technology is particularly useful for complicated nonlinear systems which are difficult to set up accurate mathematic models. Neural network has the abilities of self-learning and parallel-processing, and its BP algorithm is a powerful method to establish and tune fuzzy inference systems. Genetic algorithm (GA) is a random search method. It decreases the effect of original values greatly through crossover and mutation operations, and it can easily find out the global optimal results.In this paper, in order to achieve the intelligent control of water's temperature of solar water heater, we combine the advantages of fuzzy control technology , neural network and genetic algorithm to set up a fuzzy neural network controller based on genetic algorithm. That is: at first, we use sample data and adopt genetic algorithm to tune the parameters of antecedent variables' membership functions and the output values; then we apply BP algorithm to tune these parameters again online to optimize controller.The simulation of solar water heater system is carried out and compared with the general fuzzy control method and PID control method. The result shows that the controller designed in this paper are better. The controller's steady-error is smaller, tracing capacity is better and response speed is faster without surge.