In order to solve the difficult problem to establish an accurate model of hydraulic components of electro-hydraulic servo system because of uncertainties characteristics of non-linear and time-varying, a modeling method of optimized BP neural network based on simulated annealing genetic algorithm (SA-GA) is presented. The sudden-jump probability of simulated annealing algorithm is utilized to overcome prematurity phenomenon of genetic algorithm, and use the global optimization search capability of simulated annealing genetic algorithm to optimize the initial weights and thresholds of BP neural network. Taking the electro-hydraulic servo system of a certain type demolition mine sweeper as example, use the proposed method to carry out system offline identification. The results of simulation showed that the method which based on SA-GA-BP neural network modeling can effectively fit non-linear and time-varying characteristics of the system and the proposed method is effectively.%为解决电液伺服系统的液压元件存在非线性时变性等不确定因素,使得难以对其建立精确模型的问题,提出一种基于模拟退火遗传算法(simulated annealing genetic algorithm,SA-GA)优化BP神经网络的建模方法.利用模拟退火算法(simulated annealing algorithm,SA)的概率跳变能力克服遗传算法(genetic algorithm,GA)存在的早熟现象,在此基础上采用模拟退火遗传算法的全局寻优能力优化BP神经网络的权值和阈值.以某型爆破扫雷器电液伺服系统为例,利用所提方法对系统进行离线辨识.仿真结果表明:基于SA-GA-BP神经网络的建模方法能很好地拟合系统固有的非线性和时变性特性,所提方法是有效的.
展开▼
