张小青, 李艳红. 基于灰狼优化算法的神经网络PMSM混沌同步控制[J]. 云南大学学报(自然科学版), 2020, 42(4): 664-672. doi: 10.7540/j.ynu.20190534
引用本文: 张小青, 李艳红. 基于灰狼优化算法的神经网络PMSM混沌同步控制[J]. 云南大学学报(自然科学版), 2020, 42(4): 664-672. doi: 10.7540/j.ynu.20190534
ZHANG Xiao-qing, LI Yan-hong. Neural network control for chaotic synchronization based on GWO[J]. Journal of Yunnan University: Natural Sciences Edition, 2020, 42(4): 664-672. DOI: 10.7540/j.ynu.20190534
Citation: ZHANG Xiao-qing, LI Yan-hong. Neural network control for chaotic synchronization based on GWO[J]. Journal of Yunnan University: Natural Sciences Edition, 2020, 42(4): 664-672. DOI: 10.7540/j.ynu.20190534

基于灰狼优化算法的神经网络PMSM混沌同步控制

Neural network control for chaotic synchronization based on GWO

  • 摘要: 对永磁同步电动机(Permanent Magnet Synchronous Motor,PMSM)的混沌同步控制进行研究,引入灰狼优化算法(Grey Wolf Optimizer,GWO)和它的几个变体算法,提出一种基于灰狼优化算法的RBF-GWO网络混沌同步控制器. 在RBF-GWO网络控制器中,以RBF神经网络结构为基础,其隐层中心矩阵、高斯均方根宽度向量和权值矩阵的组合为灰狼的位置向量,选择输出平均平方误差的一半作为优化目标函数,以网络实际输出与期望输出之间的差值作为更新灰狼位置向量的依据,每次迭代的最优参数值均保存在α灰狼位置向量中,且向网络返回α灰狼位置向量,直至迭代结束条件满足. 通过PMSM混沌同构同步控制与异构同步控制实验,验证了RBF-GWO网络有效性,给出的基于一种变体GWO算法(即WGWO)的RBF-GWO网络控制器具有相对更强的自适应能力.

     

    Abstract: For the chaotic system of Permanent Magnet Synchronous Motor, the RBF-GWO network chaotic synchronization controllers were designed based on Grey Wolf Optimizer (GWO) and its several variant algorithms. In RBF-GWO, RBF neural network structure was adopted, the combination of the hidden center matrix, Gaussian RMS width vector and the hidden-output weight matrix were assumed as the position vector of grey wolf. Half of the average squared error was selected as the optimizing object function. The difference between the actual output and the desired one was considered as the guide of updating the position vector for the grey wolf, and in each iteration the optimal parameters were stored in the position vector of wolf α which would be returned to the RBF-GWO until the end condition of the iteration was satisfied. Through the chaotic homogeneous and heterogeneous synchronization control experiments, it was proved that the proposed RBF-GWO network was effective, and it was found that the one network based on the WGWO had relatively stronger adaptive capacity than others.

     

/

返回文章
返回