协同进化PSO算法优化RBF网络在齿轮箱故障诊断中的应用

发布时间：2018-06-03 20:18

本文选题：协同进化PSO算法 + RBF神经网络　；参考：《中北大学》2011年硕士论文

【摘要】：故障诊断技术是现代化生产发展的产物,齿轮箱是工程机械中的重要部件,齿轮和滚动轴承是齿轮箱中的易损元件。据统计,传动机械中80%的故障是由齿轮箱故障引起的,因此,对齿轮箱的运行状态监测和故障模式识别一直是机械故障诊断技术中的重点。本论文在潜心研究协同进化PSO算法相关理论的基础上,结合RBF神经网络优化的具体问题,提出一种基于协同进化PSO算法的RBF神经网络优化模型,并将优化后的RBF神经网络应用于齿轮箱故障诊断技术中,以期对齿轮箱系统的各种异常状态或故障工况做出及时、准确而有效的判断,指导齿轮箱系统的运行,提高齿轮箱系统的可靠性、安全性和有效性,最终把由齿轮箱故障带来的经济损失降低到最低水平。实验结果表明,本论文所提出的基于协同进化PSO算法的RBF神经网络优化模型具有可行性,且优化后RBF神经网络的测试结果与传统RBF神经网络的测试结果相比具有较高的训练精度和较快的收敛速度。因此,通过本论文的研究,不仅为RBF神经网络提供了一种新的优化途径,同时也大大提高了RBF神经网络在齿轮箱故障诊断技术中的诊断效率,进而丰富和发展了粒子群优化算法和神经网络在齿轮箱故障诊断中的应用。
[Abstract]:Fault diagnosis technology is the product of modern production, gearbox is an important part of construction machinery, gear and rolling bearing are vulnerable components in gearbox. According to statistics, 80% of the faults in transmission machinery are caused by gearbox faults. Therefore, the monitoring of gearbox operation status and fault pattern recognition are always the key points in mechanical fault diagnosis technology. Based on the theory of coevolutionary PSO algorithm and the specific problem of RBF neural network optimization, a RBF neural network optimization model based on coevolutionary PSO algorithm is proposed in this paper. And the optimized RBF neural network is applied to the gearbox fault diagnosis technology, in order to make timely, accurate and effective judgment on the abnormal state or malfunction condition of the gearbox system, and guide the operation of the gearbox system. Improve the reliability, safety and effectiveness of the gearbox system, and finally reduce the economic loss caused by the gearbox failure to the lowest level. Experimental results show that the proposed RBF neural network optimization model based on co-evolutionary PSO algorithm is feasible. Compared with the traditional RBF neural network, the test results of the optimized RBF neural network have higher training precision and faster convergence speed. Therefore, the research in this paper not only provides a new way to optimize the RBF neural network, but also improves the diagnosis efficiency of the RBF neural network in the gearbox fault diagnosis technology. The application of particle swarm optimization algorithm and neural network in gearbox fault diagnosis is enriched and developed.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3;TP183

【引证文献】