当前位置：首页 > 过刊浏览->2023年45卷6期
基于改进支持向量回归的船用机械零部件寿命估计
Life estimation of marine machinery parts based on improved support vector regression
- DOI:
- 作者:
- 张丽娜, 凌付平
ZHANG Li-na, LING Fu-ping
- 作者单位:
- 江苏航运职业技术学院 金工实训中心, 江苏 南通 226010
Metalworking Training Center, Jiangsu Shipping College, Nantong 226010, China
- 关键词:
- 支持向量回归;机械零部件;寿命估计;退化状态;主成分分析
support vector regression; mechanical parts; life estimation; degraded state; principal component analysis
- 摘要:
- 船用机械零部件退化的敏感特征难以提取，导致其寿命估计均方误差增加。为此，设计一种基于改进支持向量回归的船用机械零部件寿命估计方法。采用小波变换法去除全寿命周期数据噪声，提取零部件退化的时域特征，利用集合经验模态分解获取频域特征。经主成分分析法完成特征降维处理后，确定机械零部件退化的敏感特征。采用考虑莱维飞行机制的改进蚁狮优化算法寻求支持向量回归模型最佳参数。将提取到的敏感特征输入至改进支持向量回归模型中，得到船用机械零部件寿命估计值。实验结果表明，当步长为6时，支持向量回归模型的均方误差指标最小、决定系数指标最大，可实现机械零部件寿命精准估计。
It is difficult to extract the sensitive features of the degradation of marine machinery parts, which leads to the increase of the mean square error of their life estimation. Therefore, a life estimation method of marine machinery parts based on improved support vector regression is designed. The wavelet transform method is used to remove the noise of the whole life cycle data, extract the time domain features of the component degradation, and obtain the frequency domain features using the set empirical mode decomposition. After the feature dimension reduction processing is completed by principal component analysis, the sensitive features of mechanical parts degradation are determined. The improved ant lion optimization algorithm considering Levy flight mechanism is used to find the best parameters of support vector regression model. The extracted sensitive features are input into the improved support vector regression model to obtain the estimated life of marine machinery components. The experimental results show that when the step size is 6, the mean square error index of the support vector regression model is the smallest, and the determination coefficient index is the largest, which can achieve accurate estimation of the life of mechanical components.
2023,45(6): 186-189 收稿日期：2022-09-28
DOI：10.3404/j.issn.1672-7649.2023.06.037
分类号：TH17
基金项目：南通市科技计划项目（MSZ21011）
作者简介：张丽娜(1986-)，女，讲师/技师，研究方向为机械制造工艺教育
