Dynamic optimizing method of fuzzy controller based on genetic algorithm

Xiao-lu FENG,Lai-feng QIN,Ke-fa CEN
DOI: https://doi.org/10.3785/j.issn.1008-973X.2007.03.019
2007-01-01
Abstract:To overcome the problems of long time delay, uncertainty and time variance of main-steam temperature control object in power plant, a dynamic optimizing method for the fuzzy controller was proposed based on genetic algorithm (GA). According to the current error and the error change of the modulated variable, the improved GA and the staggered amendment method were adopted, and only a current activated amendment value of the fuzzy control rule table was optimized at each time. The data in the control rule table of fuzzy controller were dynamically optimized at the state of real time and online. The simulation results show that when the model parameter of the main-steam temperature control object changes by 10%, the fuzzy control system optimized by the proposed controller dynamic optimizing method can limit dynamic difference of main-steam temperature within -3-3°C, and can shorten the regulating time by 40% at least. The fuzzy control system can efficiently improve the dynamic character of main-steam temperature control system.
What problem does this paper attempt to address?