报告人：徐安察
报告地点：人民大街校区惟真楼523报告厅
报告时间：2025年11月20日星期四15:30-16:30
邀请人：葛磊
报告摘要：
徐安察，浙江工商大学统计与数据科学学院教授，博导，长期从事可靠性建模与管理相关领域的研究，以第一作者/通讯作者在NRL, EJOR,IISE Trans.,JQT,SJS等期刊上发表论文50余篇。主持国家自然科学基金面上和青年项目4项，浙江省自然科学基金重点项目1项。获浙江省自然科学奖、福建省自然科学奖、第一届全国统计科学技术进步奖等。目前担任中国运筹学会可靠性分会副理事长、国内统计英文期刊《Statistical Theory and Related Fields》Associate Editor。
主讲人简介：
Remaining useful life (RUL) prediction is essential for prognostics and health management, enabling reliable and cost-effective operation of engineering systems. Stochastic degradation models offer a principled probabilistic framework to capture the evolution of system health and quantify uncertainties in failure progression. This report investigates online RUL prediction based on three stochastic processes: Wiener process, gamma process, and inverse Gaussian process. For each model, we present an online estimation and prediction algorithm capable of updating RUL distributions as new degradation data become available. The proposed methodologies provide a practical and robust foundation for real-time RUL prediction under uncertainty, supporting data-driven maintenance decision-making in industrial applications.
学校主页