项目名称: 基于多源监测数据融合的剩余使用寿命预测关键问题研究
项目编号: No.61304105
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 胡艳艳
作者单位: 北京科技大学
项目金额: 25万元
中文摘要: 随着对系统可靠性要求的进一步提高和传感器技术的发展,用于复杂系统状态监测的传感器的种类和数量日益增加。特别地,工程实践中由于传感器的种类不同、采样速率不同及初始采样时刻不同等因素导致来自这些传感器的多源监测数据往往并不是同步的,而是异步的。因此,如何有效利用来自这些传感器的多源监测数据,特别是异步多源监测数据,对系统的剩余寿命进行预测,从而实现预防性维护,提高系统的可靠性和安全性,是目前预测维护理论研究和工程实践中亟待解决的新问题。本项目拟主要研究:1)含有未知模型参数的一类随机动态系统的异步多源数据估计融合方法;2)基于同步多源数据融合的一类随机退化过程的剩余寿命预测;3)基于异步多源数据融合的一类随机退化过程的剩余寿命预测。本项目的研究成果不仅具有重要的理论意义,同时对预测维护技术在工程实践中的应用具有指导意义。
中文关键词: 多源数据融合;剩余寿命预测;异步多传感器;预测维护;
英文摘要: With the enhanced requirement for system reliability and the development of sensor technology, the type and quantity of the sensors used for state monitoring of complex systems are increasing. Especially, in engineering practice, the multi-source monitoring data from multiple sensors are often asynchronous, other than synchronous, because of different types, different sampling rates and also different initial sampling time instants. As a result, a new problem raises in the theoretical study and engineering practice of predictive maintenance. That is how to use the multi-source monitoring data from these sensors effectively, especially the asynchronous multi-source monitoring data from multiple asynchronous snesors, to predict the remaining useful life of the system, and then perform predictive maintenance to improve the reliability and safety of the system. This project intends to study: 1) Asynchronous multi-source data estimation fusion approach for a kind of stochastic dynamic systems with unknown model parameters; 2) Synchronous multi-source data fusion based remaining useful life prediction for a kind of stochastic degradation process; 3) Asynchronous multi-source data fusion based remaining useful life prediction for a kind of stochastic degradation process. The research results of this project have signif
英文关键词: Multi-Source Data Fusion;Remaining Life Prediction;Asynchronous Multiple Sensors;PredictIve Maintenance;