摘要
本文分析了改进的ELMAN网络的结构,并讨论了神经网络的学习算法,针对BP算法的缺陷,提出了用遗传算法修正网络权值的学习算法。另外,将采用遗传算法进行训练的改进ELMAN网络应用于非线性系统的辨识和建模。通过仿真和在汽车磷化加热系统建模中的应用进一步说明了该方法用于高阶次非线性系统建模的可行性。
A modified ELMAN network and its algorithm are studied in this paper. To overcome the slow convergence of the BP algorithm, genetic algorithm (GA) is proposed, which can train the weight of the network. In addition, a given model is identified by using modified ELMAN network trained with GA, and the model of phosphating temperature control system is also established by this method. Both simulation and experiment demonstrate the effectiveness of the proposed algorithm in system modeling.
出处
《计算技术与自动化》
2004年第1期37-39,共3页
Computing Technology and Automation
