基于多目标遗传算法的航空发动机PID控制器参数优化

提出了1种基于多目标遗传算法以及多属性决策的PID参数设计方法，综合考虑了系统超调量、稳定时间和ITAE指标，采用改进非支配解排序的多目标遗传算法（NSGAII）求出了Paret0最优解；用这些Pareto最优解构成决策矩阵，使用客观赋权的信息熵法对最优解的属性进行了权值计算，然后采用逼近理想解的排序方法（TOPSIS）进行了多属性决策（MADM）研究，对Pareto最优解给出了排序；计算了某型航空发动机PID控制的数值算例，结果表明所提出的联合方法通用性好，所设计的PID性能优异，适合工程实际应用。

Parameters Optimization of PID Controller for Aeroengine Based on Multi-Objective Genetic Algorithm

A design method was presented far PID parameters based on the multiobjective genetic algorithm and multi-attributive decision. The system overshoot, stabilization time and ITAE index were considered comprehensively. The optimum solutions of Pareto were obtained by using the multi-objective genetic algorithm Ⅱ（NSGA Ⅱ） of improved non-dominating solution sorting, which formed the decision matrix. The weighted value calculation for the attribute of the optimum was performed by applying the information entropy method of objective weigths solutions. Then the multi-attribute decisions （MADM） were investigated by using the sort method of approaching perfect solution, and the Pareto optimum solutions were sorted. The numerical examples were calculated for an aeroengine PID controller. The results indicate that the commonality of association method is good, and the performance of PID is excellent which is suitable for the practical application of engineering.