摘要
利用运行数据识别产品设计缺陷或薄弱环节是产品开发和更新换代的主要模式,也是保证产品服役性能稳定的重要途径。提出了一种性能数据驱动的产品关键设计参数(薄弱环节)识别方法。首先,利用极限学习机算法划分运行工况,提出了基于核主成分分析和高斯混合模型的多工况性能退化评估方法,消除了工况变化对性能退化评估的影响,得到性能退化严重的关键功能模块;其次,对性能监测数据进行聚类分析,识别出与模块性能退化密切相关的关键性能监测参数;最后,建立了"性能监测参数—性能参数—设计参数"三者之间的关联关系,识别出导致性能严重退化的关键设计参数。以某大吨位履带起重机作业机构为例,验证了方法的有效性。
Identifying the weaknesses or defects of the current design based on the operating data is the main mode of product development. It is also an important way to ensure the stable performance of products. An approach of critical design parameter identification based on performance data is proposed. Firstly, the operating conditions are identified by the algorithm of extreme learning machine. A performance degradation assessment method based on kernel principal component analysis and gaussian mixture model is conducted, which eliminates the influence of operating conditions on performance degradation assessment and obtains key functional modules with severe performance degradation. Secondly, cluster analysis is carried out on operation monitoring data to identify critical monitoring parameters closely related to module performance degradation. Finally, the correlation between "performance monitoring parameter-product performance parameter-design parameter" is established to identify the critical design parameters. The effectiveness of the proposed method is verified by a case study of a large tonnage crawler crane.
作者
褚学宁
陈汉斯
马红占
CHU Xuening;CHEN Hansi;MA Hongzhan(School of Mechanical Engineering,Shanghai Jiao Tong University,Shanghai 200240)
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2021年第3期185-196,共12页
Journal of Mechanical Engineering
基金
国家自然科学基金资助项目(51875345,51475290)。
关键词
性能监测数据
性能退化评估
关键设计参数识别
高斯混合模型
履带起重机
performance monitoring data
performance degradation assessment
critical design parameter identification
Gaussian mixture model
crawler crane