基于EEMD能量熵和混合算法的电机轴承故障诊断

Motor Bearing Fault Diagnosis Based on EEMD Energy Entropy and Hybrid Algorithm

中文关键词:  轴承  故障诊断  粒子群  模拟退火(SA)  集合经验模态分解（EEMD）

英文关键词:bearing  fault diagnosis  particle swarm  simulated annealing(SA)  ensemble empirical mode decomposition (EEMD)

中文摘要:

      针对提升机电机轴承振动信号的非平稳特性和单一粒子群算法(PSO) 优化径向基函数（RBF）神经网络时存在网络收敛速度慢和适应度值易陷入局部最小的缺点，提出基于集合经验模态分解（EEMD）能量熵和模拟退火粒子群混合算法（SAPSO）优化RBF神经网络的提升机电机轴承故障诊断方法。基于EEMD求取振动信号各固有模态函数分量的能量熵，并使用相关性分析方法剔除虚假的分量，把筛选后的有效数据作为故障识别的特征向量；利用模拟退火（SA）算法具有局部概率突跳的特性，将SA算法和PSO算法相结合，在优化RBF诊断模型隐含层参数时以实现不同算法间的优劣互补。仿真结果表明，使用SAPSO算法优化后的RBF神经网络模型在提升机电机轴承故障诊断中能够加快网络收敛速度和提升故障识别精度。

英文摘要:

      Considering the non-stationary characteristics of the vibration signals of the hoist motor bearing and shortcomings that the network convergence is slow and the fitness value is easy to fall into the local minimum when using RBF neural network based on the single particle swarm optimization (PSO),a fault diagnosis method for hoist motor bearing is proposed which employs ensemble empirical mode decomposition (EEMD) energy entropy and simulated annealing particle swarm optimizing (SAPSO) for optimizing RBF neural network.The energy entropy of each intrinsic mode function component was obtained via EEMD,whose false components were removed by correlation analysis.The selected valid data were served as the feature vector for failure recognition.Because the simulated annealing algorithm (SA) has the characteristics of probability jump,combining SA and PSO can complement the advantages of each other.The results show that the method can speed up the network convergence and improve the accuracy of fault identification.