The Research Of Intelligent Optimization Tuning For Thermal Control System

This paper analyzed shortages of traditional genetic algorithm and simulate annealingalgorithm and improved them. The improvements on the traditional GA mainly focus on threeaspects: reducing search-range, crossover operator, selection operator; those on thetraditional SA mainly used returning temperature strategy, and added a memory. The twoimproved methods were used to optimizing the controller parameters of thermal system,the imitating result show that the improved GA overcame the shortages of traditional GAin the following aspects: precocious, easily falling into partial optimization and hard tojumping out, and enhanced the optimizing efficiency and the solution precision; theimproved SA overcame the shortages easily falling into partial optimization oftraditional SA, and increased the successful efficiency, and the memory addedguaranteed the last result is the best one found by the improved SA.