基于最小二乘支持向量机对核电数控轮槽铣床可靠性分析

Reliability Analysis of Nuclear Electricity NC Milling Machine for Processing Wheel Trough Based on LS Support Vector Machine

基于核电数控轮槽铣床可靠性数据样本量复杂、存在可靠性数据模糊的情况,传统可靠性分析方法无法进行可靠性预计或预计误差较大。由于最小二乘支持向量机方法具有出色的小样本性能和良好的泛化性能,针对这种情况提出一种利用最小二乘支持向量机方法对核电轮槽铣床可靠性进行分析,通过建立机床的可靠性模型,将优化后的机床各子系统可靠性数据作为最小二乘支持向量机参数,对机床可靠性进行最小二乘支持向量机建模。实验结果表明利用上述方法可以快速准确确定机床的可靠性,同时为系统可靠性分析和系统维修决策的研究提出了新的思维路径,具有很好的应用前景。

As the sampling of reliability data is complex with fuzzy data with respect to the nuclear electricity NC milling machine for processing wheel trough, the traditional analysis method fails to predict reliability of the machine and quite often with big errors in data. Bul IN support vector machine has excellent performance of small sample and good generalization performance, so it can l）e used to analyze reliahility via creating tile machine＇ s reliability model and considering each optimized machine suhsystem reliability data as its parameters. Experimental results show that this method can fast and accurately determine the reliability of the machine tool, prowiding new vision and idea for system reliability analysis and system maintenance research, with a good application prospect.