基于遗传算法的电静压伺服系统模型参数辨识

Model Parameter Identification of an Electro-Hydrostatic Actuator Base on Genetic Algorithm

为了优化模型仿真的控制系统参数,针对电静压伺服系统(Electro-Hydrostatic Actuator,EHA),建立控制器、电机、液压作动器模块化模型,根据各模块输入输出量的试验数据,采用最小二乘法对系统模型参数进行辨识,并将辨识结果导入遗传算法的初始种群,进一步对系统模型参数进行优化,得到系统的精确模型。结果表明,优化的辨识结果在时域和频域都更为精确的逼近实际系统,匹配程度在90%以上,证明该方法的可行性。

An Electro-Hydrostatic Actuator(EHA)model,including a controller,a motor and a hydraulic actuator is constructed.The inputs and outputs of each block are recorded when testing the EHA prototype,based on which the model parameters are identified.The Least Square Identification is performed at first,whose parameters identification result is introduced into the initial group of a Genetic Algorithm Identification to get optimized.It is shown that this method is practicable,for the model with parameters from the optimized identification result matched the real system better,with over 90%similarity,both in time domain and frequency domain.