航空发动机PID控制参数优化的改进遗传算法

提出一种基于参考模型特征指标的P ID控制参数寻优算法。采用遗传算法(GA)优化某型涡扇发动机P ID控制参数,以理想二阶系统作为参考模型,将其与实际闭环系统输出差值平方的时间积分作为系统的适应度函数。设计过程中只需选择二阶系统的自然频率和阻尼比就能准确地实现期望的动态和稳态性能。与传统的基于系统性能指标加权的适应度函数相比,新的适应度函数计算方法避免了加权系数与系统响应形式没有明确对应关系的缺陷。新方法所选参数物理意义明确,算法简单,易于实现。

Improved Algorithm for Calculating GA Fitness Function of Aeroengine PID Controller Parameter Optimazation

A new algorithm based on the characteristic index of the reference model for PID controller parameter optimization is proposed.Using the desired second order system as the reference model,by a genetic algorithm(GA),the PID controller parameter of a certain areoengine can be optimized.The fitness function of GA is composed by the time integral of the square of the output difference between the desired second order and the actual close loop systems.This paper selects the natural frequency and the damping ratio of the desired second order system,and can obtain accuratly desired dynamic and steady state system performances.Compared with the traditional fitness function based on the weighting combination of the system performance,the new fitness function calculation method can avoid the limitation in which the weighting coefficients have no a clear relation with the system response.The parameters selected in the new method have specific physical meanings.The new method is very simple and easy to be realized.