The Application Of Parameter Optimization Strategy Of Fuzzy Control In Exhausted Steam Recovery System

Resources are not inexhaustible, and people are more and more attention to environmental protection. So energy conservation and environmental protection has become a hot issue around the world, much of concern. Fuzzy control's basic theorem and design methods have been made breakthrough constantly, but it is not perfect. There are still many hard nuts to crack, one of which is the design and optimization of the structure and parameters of fuzzy controller. From this starting point, this paper makes a study on the optimization of the fuzzy controller parameters using in exhausted steam recovery system.Firstly, this paper analyzes the impact of the membership functions upon the completeness and the interference of control rules theoretically. Then, for further study, it makes a research on the shape, distribution, similarity of membership function and quantification factor, scaling factor through simulation, and has got some discoveries. The first one is that the more "steep" of the membership function shape, the higher sensitivity of the system; otherwise, if the function is "flat", the response will change to"smooth". The second one is that the system will get higher resolution and regulating precision in the domain where the membership functions distribute densely. Otherwise, the regulating will be "rough". The third one is that the degree of similarity between the membership functions is unsuitable neither too high nor too low. When the error tends to "zero", high degree of similarity will make the output of the controller too "fuzzy" to regulate fast. On the other hand, low degree of similarity will make the regulating precision much lower. The last one is that if the quantification factor of error increases, the system response rate will be accelerated, but it's prone to over-regulation and oscillation. With different effect, when quantification factor augments, the over-regulation will be effectively restrained. Scaling factor affects the size of the steady-state error. These discoveries may be valued of reference in designing and optimizing membership functions. As to the optimization of fuzzy control rules, this article presents a strategy in which the quantification factors, scaling factor and control rules will be simultaneously optimized. The satisfied results obtained from the simulation experiments explain that this method is correct and practical.Furthermore, this paper makes a study of the application of control algorithms. It analyzes the feasible programs of using these control algorithms in configuration softwares. A fuzzy ActiveX control is embedded in iFIX during the simulation experiment. And the simulation has achieved a satisfactory result.