(1.成都理工大学信息管理学院.数学地质四川省重点实验室.成都610059; 2.新疆财经学院,乌鲁木齐830012;3.西吕卫星发射中心技术部.四川两昌615000) [摘要]将粒子群优化的BP神经网络作为模型,参考自适应控制系统的控制器,把参考模型输 出与系统实际输出的均方误差作为PSO-BP神经网络的适应函数,通过PSO算法强大的搜索 性能使自适应控制系统的均方误差最小化。仿真实例结果表明,基于粒子群优化算法的BP 神经网络自适应控制系统收敛快、精度高,有较好的网络的泛化和适应能力,能够很好地控制 系统的输出跟随参考模型的输出。
[关键词]模型参考自适应控制系统;粒子群优化算法;BP神经网络;参考模型 [分类号]O231.2 [文献标识码]A
BP neural network model reference adaptive control system based on the particle swarm optimizer
CHEN Ling1, YAN Hai-bo2 , MAO Wan-biao3
1. College of Information Management , (Chengdu Universit3, of Technology, Chengdu 610059, China; 2, Xinjiang- Institute of Finance & Economic.s, Urumqi 830012, China; 3, Department of Techno]ogy of XSLC. Xichang 615000, China
Abstract: In this paper, PSO-BP NN is used as the controller of model reference adaptive controlsystem, the mean square error of the reference model output and the model reference adaptive controlsystem factual output as the adaptive function, and then, the mean square error is minimized by thesearch capability of PSO. The simulation results show astringency and precision and adaptability ofPSO-BP NN model reference adaptive control system are fine. The output of PSO-BP NN modelreference adaptive control system can follow up the scent of the reference model output very well.Key words: model reference adaptive control system; particle swarm optimizer(PSO) algorithm; BP neural network; reference model
因篇幅问题不能全部显示,请点此查看更多更全内容