扩展卡尔曼滤波方法在转子典型故障诊断中的应用

转子故障的早期诊断与预示是当前转子动力学的一个难点。文献1]提出了一种基于多模型估计得到转子裂纹参数对转子进行故障诊断的方法,但是该方法需要构造大量的多模型估计器,存在着构造复杂,计算量大的问题。针对这个问题,提出基于扩展卡尔曼理论(EKF)的转子典型故障诊断方法。针对Jeffcott转子建立了不对中、裂纹和弯曲故障模型,还分别构建了基于扩展卡尔曼滤波-加权整体迭代(EKF-WGI)的参数估计方程。通过不对中、弯曲和裂纹故障的实验验证,表明该参数估计方法对于转子典型故障有着较高的诊断能力。与传统的诊断方法提取频谱特征相比,该方法不依赖于经验和故障事例,可以较精确地估计故障参数,在转子的故障诊断中更有针对性。

Application of Extended Kalman Filter Method to Typical Rotor Fault Diagnosis

The rotor fault diagnosis is a key problem of rotor dynamics. Based on multiple model adaptive estimation, Literature 1 proposed a fault diagnosis method by getting the rotor crack parameters. For this problem, a method of typical rotor fault diagnosis is proposed based on extended kalman filter (EKF). For the Jeffcott rotor, several fault models (misalignment, bending and crack faults) are built. Fault parameter estimation equations are set up based on extended kalman filter and weighted global iteration (EKF-WGI). Via the experiment verification of misalignment fault, crack fault and bending fault on bently rotor, it is proved that this proposed method is effective in rotor fault diagnosis. Compare with the traditional diagnosis methods to extract spectrum features, it can estimate the fault parameters accurately without experiences and fault cases, which is more targeted in the rotor fault diagnosis.