复杂传动系统的故障特征提取分析的综合方法研究

摘要

This paper tests the gathered acceleration vibration signal (called experimental signal) of wind power generator gear transmission system and calculated its power spectrum. Used zoom spectrum combining with inverse spectrum to analyze the power spectrum to obtain the characteristics of the experimental signal, and judged the fault location and fault type of the gear transmission system. In order to verify the fault, this paper builds dynamical model of faulty gear transmission system, then imports it in CAE software to simulate, and extracts its vibration signal to analyze by using signal processing methods the same as described above with the purpose of getting vibration signal characteristics of the faulty transmission system, uses the result to judge whether the result of experimental signal is right, and uses fractal method to analyze different status power spectrum fractal characteristics of gear transmission system. The result shows that the correlation dimension can reflect its vibration signal components of complex degree and the difference between the fault position and degree, can provide basis for failure characteristic quantities of gear transmission, to judge fault.%对风电机组齿轮传动系统的加速度振动信号（称为实验信号）进行测试，计算获得其功率谱。并用细化谱与解调谱分析相结合的方法对其功率谱进行分析，得到实验信号的故障特征，初步判断齿轮传动系统发生故障的大概位置及故障类型。为进一步准确验证故障，建立其相应故障的齿轮传动系统的动力学模型，进行动力学模拟分析，对所模拟的振动响应信号采用实验信号相同的处理方法对其处理分析，得到模拟振动信号的故障特征。通过模拟信号的故障特征分析验证了对实验信号故障诊断的正确性。并用分形方法分析了齿轮传动系统不同状态信号功率谱的分形特征。结果表明，关联维数能反映出其振动信号成分的复杂程度，区别出故障的位置和程度。所以，可作为齿轮传动系统的故障特征量及故障诊断的依据。