混合插值的ESMD在电机轴承故障特征提取的应用

针对极点对称模态分解(ESMD)处理电机轴承故障信号存在局部模态混叠的问题,提出了一种基于有理Hermite插值和三次样条插值改进的ESMD电机轴承故障特征提取方法。由于首先有理Hermite插值可通过控制形状参数来调节插值曲线,将固有模态分量(IMF)的瞬时频率带宽作为的优化准则,采用有理Hermite插值与三次样条插值相结合,既避免了插值耗时过长,又考虑了插值曲线的平滑性;采用自适应权重调整的粒子群算法(PSO)确定每阶IMF的最优形状控制参数,避免陷入局部最优,使得IMF最优,从而提高ESMD自适应性和分解精度。试验结果表明,该方法可有效提取电机轴承故障特征,并有效缓解ESMD模态混叠,与其它方法相比分解效果更好。

Application of hybrid interpolated ESMD in motor bearing fault feature extraction

Aiming at the problem of local modal aliasing of motor bearing fault signal for extreme-point symmetric mode decomposition (ESMD), an improved fault condition of ESMD motor bearings based on rational Hermite interpolation and cubic spline interpolation was proposed.Since the first rational Hermite interpolation can adjust the interpolation curve by controlling the shape parameters, the instantaneous frequency bandwidth of the Intrinsic Mode Function (IMF) was used as the optimization criterion, and the combination of rational Hermite interpolation and cubic spline interpolation was avoided.The interpolation takes too long, and the smoothness of the interpolation curve is considered.The particle swarm optimization (PSO) algorithm with adaptive weight adjustment determines the optimal shape control parameters of each order IMF, avoiding falling into local optimum and making the IMF optimal, thereby improving ESMD adaptability and decomposition accuracy.The experimental results show that the proposed method can effectively extract the fault characteristics of motor bearings and effectively alleviate the aliasing of ESMD modes.Compared with other methods, the decomposition effect is better. 