基于相空间重构的电磁继电器电性能参数预测研究

摘要

提出了基于相空间重构的电磁继电器电性能参数预测的GA神经网络方法。借助相空间重构技术，重构了电性能参数序列的高维空间轨迹；通过采用基于GA算法的神经网络拟合重构轨迹，建立了电性能参数序列预测模型。以电磁继电器接触电阻和燃弧能量性能参数为例，分析了重构参数对重构轨迹、预测结果的影响。该方法能从高维空间揭示电性能参数退化过程，充分利用神经网络泛化性能，因而预测精度高。实验结果表明，所提方法可行、有效。%The GA neural network method based on the phase space reconstruction is proposed for electrical perform-ance parameters prediction of the electromagnetic relays.The locus of electrical performance parameter series in high dimension space are reconstructed by using the phase space technique.The prediction model of the electrical perform-ance parameter series is built by means of fitting the reconstructed locus with the neural network trained with GA algo-rithm.The effects of reconstructed parameters on the locus and the prediction results are analyzed with the contact re-sistance parameter and arc energy of the electromagnetic relays.The proposed method has the ability to discover the degradation process of electrical performances for the relay in the high dimensional phase space and make full use of generalization of neural network, and therefore has the better precision of prediction.The experimental results show that the proposed method is feasible and effective.