摘要
工业设备发生故障前往往会表现出多个信号特征异常,为了更准确预测这些设备的使用寿命,防止事故发生,本文针对设备的双变量退化问题,提出非线性双变量Wiener过程模型.该模型通过随机漂移参数服从二元正态分布描述设备退化过程中的相关性和个体差异性,随后利用马尔科夫链蒙特卡罗(MCMC)抽样方法对复杂参数结构进行参数估计,并通过数据仿真实验验证参数估计的准确性.最后,通过轴承实例将本文模型与其他模型进行比较,结果显示本文模型在寿命预测和可靠性评估方面具有更高的准确性.本文的非线性双变量Wiener过程能有效描述多变量设备的退化过程,可为设备寿命预测提供有力支持.
Before industrial equipment fails,it often exhibits multiple abnormal signal features.In order to more accurately predict the service life of these devices and prevent safety accidents,this paper proposes a nonlinear bivariate Wiener process model for device degradation.The model describes the correlation and individual differences in the process of equipment degradation by using random drift parameters following binary normal distribution.Then Markov Chain Monte Carlo(MCMC)sampling method is used to estimate parameters of complex parameter structures,and the accuracy of parameter estimation is verified by data simulation experiments.Finally,the proposed model is compared with other models through a bearing example,and the results show that the proposed model has higher accuracy in lifetime prediction and reliability evaluation.The nonlinear bivariate Wiener process proposed in this paper can effectively describe the degradation process of multi-variable devices and provide strong support for device lifetime prediction.
作者
肖蒙
信明江
单苏苏
洪铭键
杨力臻
XIAO Meng;XIN Mingjiang;SHAN Susu;HONG Mingjian;YANG Lizhen(School of Rail Transportation,Wuyi University,Jiangmen 529020,China)
出处
《五邑大学学报(自然科学版)》
2025年第1期38-45,共8页
Journal of Wuyi University(Natural Science Edition)
