基于遗传算法的模糊优化控制

Optimization of Fuzzy Controller Based on Genetic Algorithms

针对传统遗传算法有搜索速度慢、容易陷入局部最优的缺点,采用双种群遗传算法,并改进遗传算子,采用分组式的选择算子、不同种群采用不同的交叉算子和变异算子,将改进的算法应用到模糊控制规则的优化上,通过Matlab仿真结果显示,改进的双种群粒子群遗传算法与普通双种群遗传算相比较,能在比较少的进化代数上收敛,优化后的模糊控制器控制效果更好。

Aiming at the defect of lower solution accuracy and tending to fall into local optimal solution of the standard genetic algorithm, a dual population genetic algorithm and improve operators by using the packet mode selection operator is used. Different crossover operators and mutation operators are used on different populations. The improved algorithm is applied to the optimization of fuzzy control rules. Matlab simulation result shows that the improved method has stronger evolutionary capacity compared with the traditional dual population genetic algorithm, the effect of the control for FC optimized is better.