摘要
针对滚动轴承的故障诊断问题,提出了一种基于遗传算法的BP神经网络滚动轴承故障诊断方法。以BP神经网络的误差为目标函数,利用遗传算法进行BP神经网络的权值和阈值优化,并用优化后的BP神经网络进行故障诊断。通过MATLAB仿真,结果表明遗传算法优化的BP神经网络相比传统的BP神经网络具有更好的诊断效率和准确度。
It establishes an BP neural network of genetic algorithm method to achieve fault diagnosis for rolling bearings. Taking the error as objective function,it optimizes the weights and biases of BP neural network with genetic algorithm,and accomplishes the fault diagnosis via the optimized BP neural network. This genetic algorithm optimization of BP network has better diagnostic efficiency and accuracy compared to traditional BP network in simulation results by MATLAB.
出处
《机械设计与制造工程》
2015年第3期65-68,共4页
Machine Design and Manufacturing Engineering
基金
重庆工商大学融智学院培育项目(20140205)
关键词
滚动轴承
故障诊断
遗传算法
BP神经网络
rolling bearing
fault diagnosis
genetic algorithm
BP neural network