接触电阻时间序列的混沌特性识别与预测研究  

Research on chaotic characteristics identification and prediction of contact resistance time series

作　　者：马东娟[1] 李玲玲[1] 

机构地区：[1]河北工业大学电磁场与电器可靠性省部共建重点实验室,天津300130

出　　处：《电力系统保护与控制》2015年第10期57-61,共5页Power System Protection and Control

基　　金：国家自然科学基金(51377044;51475136);天津市高等学校科技发展基金计划项目(20130422);河北省自然科学基金(E2014202230)~~

摘　　要：接触电阻是表征触点接触性能的一个重要指标。对监测得到的接触电阻时间序列进行混沌特性识别与预测。运用Cao氏法确定了接触电阻时间序列的最小嵌入维数,通过绘制其相轨线图和计算最大Lyapunov指数,从定性和定量两个角度分别验证了接触电阻时间序列具有混沌特性。最后基于最大Lyapunov指数方法对接触电阻时间序列部分数据进行了短期预测,并给出了混沌预测的尺度。结果表明,在可预测尺度内,预测效果较好且比较稳定,超出可预测尺度范围后,预测误差变大且不稳定。Contact resistance is an important index for characterization of contact performance. The chaotic characteristics of contact resistance time series for the detected contacts are identified and predicted. The minimum embedding dimension of contact resistance time series is determined by Can method. It proves that the contact resistance time series has the chaos characteristics by drawing phase diagram method and calculating largest Lyapunov exponent from qualitative and quantitative view respectively. The chaos prediction and the reliable scope for partial data of contact resistance time series are presented based on the Largest Lyapunov exponent. The result indicates that the prediction accuracy is higher and steadier within the predictable range than beyond of it.

关 键 词：接触电阻 时间序列 Cao氏法 最大LYAPUNOV指数 混沌特性 预测 

分 类 号：TM58[电气工程—电器]