汇报题目： RUL prediction methods for lubricating oil deterioration

汇报人：杜映

汇报时间：2017年4月17日(星期一) 20:00

汇报地点：我院轴承所会议室（中三楼1330）

出国交流时间：19个月 ，2015年8月 – 2017年3月

联合培养单位：University of Toronto

汇报内容：

1．多大MIE简介

2. 科研工作

2.1 隐马尔科夫介绍

2.2基于HMM的润滑油剩余寿命预测模型研究

2.3基于HSMM的润滑油剩余寿命预测改进模型的研究

3. 留学生活及其他体会

汇报安排

题目：参加WOM2017国际会议总结报告会

时间：2017年4月17日（星期一） 20：00-20:40

地点：我院轴承所会议室（中三楼1330）

报告人：博1321 杜映

会议安排

会议名称：21st International Conference on Wear of Materials (WOM) 2017

会议时间：26 - 30 March 2017

会议地点：Long Beach, California, USA

会议简介：

The 21st International Conference on Wear of Materials focuses on both the fundamental and applied aspects of wear and friction of materials at the macro-, micro-, and nano-scale. It will address the understanding of tribological phenomena; particularly the progression in recent decades. Special sessions will concentrate on wear of tools and tooling materials, hot and cold erosion (droplet, solid and cavity-based erosion), wear of brake/frictional materials, marine wear systems, role of third bodies during wear, and surface texturing for wear reduction.

In addition, the conference provides a unique international forum for researchers and practicing engineers from different disciplines to interact and exchange their latest understandings.

会议交流工作

Poster Presentation: Remaining useful life prediction of lubricating oil using condition monitoring and HMM. 张贴报告人：杜映

参加论文信息

Title：Remaining useful life prediction of lubricating oil using condition monitoring and HMM

Author：Ying Du, Tonghai Wu, Viliam Makis

Abstract: Lubricating oil plays a vital role in the full life-span performance of the machine. Lubricating oil deterioration, which leads to the attenuation of oil performance and severe wear afterwards, is a slow degrading process, which can be observed by condition monitoring, but the actual degree of the degradation is often very difficult to be examined. The main purpose of degradation prediction is to estimate the failure time when the oil no longer fulfills its function. We suppose that the state process evolution of lubricating oil degradation can be modeled using a hidden Markov model (HMM) with three states: healthy state, unhealthy state, and failure state. Only the failure state is observable. While the lubricating oil is in service, vector data that is stochastically related to the deterioration state is obtained through online condition monitoring by a wear debris sensor at regular sampling epochs. A method of Time Series Analysis (TSA) is applied to the healthy portions of the oil data histories to get the residuals as the partially observable process to fit the hidden Markov model. The unknown parameters of HMM are estimated by the Expectation-Maximization (EM) algorithm. The remaining useful life prediction of lubricating oil is obtained by applying a cost-optimal Bayesian fault prediction scheme.

欢迎各位感兴趣的同学届时前来交流。