基于振动信号谱形状熵特征的高压断路器操动状态辨识方法

Identification Method for Operation State of High Voltage Circuit Breakers Based on Spectral Shape Entropy Characteristics of Vibration Signals

高压断路器储能弹簧势能瞬间释放转换为机械能促使机构部件运动,并传递到动触头完成分、合闸操作,其伴随的振动信号表现出非平稳、非线性特点。该文提出非平稳信号的谱形状熵特征分析方法,首先利用自适应噪声集合经验模态分解(CEEMDAN)将频率结构繁杂的非平稳信号逐步拆分,得到代表不同频段特征的分量;然后将筛选后分量的功率谱转换到划分子区域的极坐标下,由发散因子提高主峰区域敏感度,根据波形散布在子区域的特点,计算谱形状熵特征;最后由分群粒子群算法优化支持向量机模型对断路器操动状态进行辨识。实验结果表明,谱形状熵能够刻画振动信号波形变异和功率主峰分布特征,对高压断路器典型机械故障识别准确率高,且辨识过程时间大大缩短。

The potential energy of the energy storage spring of the high voltage circuit breaker is instantaneously released and converted into mechanical energy,inducing the movement of the mechanism components.This enables the static and dynamic contacts to complete the opening and closing operations.The accompanying vibration signal exhibits non-stationary and non-linear characteristics.In this paper,a feature analysis method called spectral shape entropy is propose.Firstly,the non-stationary signal with complicated frequency structure is gradually split by using the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)to obtain the components representing different frequency band features.Then the power spectrum of the filtered component signal is transformed to polar coordinates,and the sensitivity of the main peak region is improved by the divergence factor.Calculating the spectral shape entropy characteristics according to the waveform scattered in sub-regions.Finally,the support vector machine optimized by grouped particle swarm algorithm model is used to identify the circuit breaker operating states.The experiments show that the spectral shape entropy can characterize the waveform variation of vibration signals and the distribution of the main peaks of the power spectrum.It has high accuracy in identifying typical mechanical faults of circuit breakers and takes less computing time.