基于BQGA-ELM网络在滚动轴承故障诊断中的应用研究

Application of BQGA-ELM network in the fault diagnosis of rolling bearings

提出一种基于Bloch球面量子遗传算法(BQGA)优化极限学习机(ELM)网络的诊断方法(BQGA-ELM),并将BQGA-ELM运用于滚动轴承故障诊断中。基于UCI标准数据集,通过仿真实验比较Bloch量子遗传算法与其它算法优化ELM的性能,仿真实验表明BQGA的优化效果强于其它优化算法。从实验室采集滚动轴承正常、内环故障、外环故障和滚珠故障四种工况的振动信号,并利用时域分析法提取振动信号的相关特征参量。将提取的特征参量经过数据预处理,再输入到诊断模型中进行滚动轴承故障诊断。结果表明:BQGA-ELM能够准确有效的对滚动轴承故障进行诊断,且其误差收敛与故障诊断时间均优于文中其它诊断模型。

A novel fault diagnosis model for rolling bearings,by the name of BQGA-ELM,was proposed based on the optimized extreme learning machine(ELM)combined with the Bloch spherical quantum genetic algorithm(BQGA).Comparing with other optimization algorithms including the genetic algorithm(GA),particle swarm optimization(PSO)and quantum genetic algorithm(QGA)by numerical simulations using the standard example data in the UCI machine learning repository:data sets,it is shown that the optimization method based on BQGA is superior to other optimization methods.The vibration signals of a rolling bearing in the following 4 cases,namely,normal,running,inner ring failure,outer ring fault and ball fault were collected in the lab,and the related characteristic parameters of the experimental data sets were extracted by time-domain analysis and then input into the diagnostic models.The diagnostic results show that the BQGA-ELM is a more reliable and suitable method than other methods for the defect diagnosis of rolling bearings,and its error convergency and fault diagnosis time are better than other diagnosis models in the paper.