Power Fault Recognition Based on Multiwavelet Packet with Different Pre-processing Methods and Wavelet Neural Network

摘要：

Multiwavelets own better properties than those of traditional wavelets.The pre-processing method of multiwavelet is one of key factors afiecting its applications.A noveI power fault recognilion method based on multiwavelet packet coefficient entropy(MPCE)and wavelet neural network(WNN) is proposed.Firstly,the appropriate multiwavelet packet decomposition of the sampled transient current is performed and each MPCE of transient current is calculated.Then eigenvector of the current signal is constructed,and by taking the eigenvector as training samples the WNN is trained to implement the fault recognilion,The proposed method with different pre-processing methods is compared with the means based on traditional wavelet packet and WNN.Simulation results show that the proposed method is effective and feasible.

关键词： Multiwavelet packet Multiwavelet packet coefficient entropy Pre-processing method Traditional wavelet packet Wavelet neural network

作者: LI Dongmin LIU Zhigang SU Yuxiang CAI Jun

会议类型: 国际会议

会议名称: The International Conference Information Computing and Automation(2007国际信息计算与自动化会议)

会议地点: 成都

会议语种:英文

页码: 547-550

在线出版日期: 2007-12-19（万方平台首次上网日期，不代表论文的发表时间）