基于特征权值小波包能量分析的异步电动机电气故障特征提取

Electrical fault feature extraction of asynchronous motor based on feature weight-wavelet packet energy analysis

异步电动机是生产生活中应用最广泛的电动机。对异步电动机电气故障特征进行提取,是保证电动机正常运行的重要手段。首先,对异步电动机电气故障进行仿真与模拟后,获得定子电流信号。然后,通过小波包能量分解和重构这些信号,得到故障特征向量。最后,运用特征权值算法得出每个特征的权重。研究表明:借助特征权值算法与小波能量分析的方法能使提取的故障特征更为精确,仿真验证了其有效性和可行性。

Asynchronous motor is the most widely used motor in production and life.It is an important method to extract the electrical fault characteristics of asynchronous motor to ensure the normal operation of motor.First,after the electrical fault of asynchronous motor is simulated,the stator current signal is obtained.Then,through a wavelet packet energy decomposition and reconstruction of these signals,the fault feature vector is obtained.Finally,the weight of each feature is obtained by using the feature weight algorithm.The research shows that the extracted fault features can be more accurate by the feature weight algorithm and wavelet energy analysis,and its effectiveness and feasibility are verified by simulation.