Adaptive Change Detection for Long-Term Machinery Monitoring Using Incremental Sliding-Window

作     者：Teng Wang Guo-Liang Lu Jie Liu Peng Yan 

作者机构：Key Laboratory of High-Efficiency and Clean Mechanical Manufacture of MOE School of Mechanical EngineeringShandong University Department of Mechanical and Aerospace EngineeringCarleton University 

出 版 物：《Chinese Journal of Mechanical Engineering》 (中国机械工程学报（英文版）)

年 卷 期：2017年第30卷第6期

页      面：1338-1346页

核心收录：

学科分类：0817[工学-化学工程与技术] 08[工学] 0807[工学-动力工程及工程热物理] 0802[工学-机械工程] 0811[工学-控制科学与工程] 0801[工学-力学（可授工学、理学学位）] 080201[工学-机械制造及其自动化] 

基　　金：Supported by National Natural Science Foundation of China(Grant Nos.61403232,61327003) Shandong Provincial Natural Science Foundation of China(Grant No.ZR2014FQ025) Young Scholars Program of Shandong University,China(YSPSDU,2015WLJH30) 

主　　题：Machine monitoring Change detection Long-term monitoring Adaptive threshold 

摘      要：Detection of structural changes from an opera- tional process is a major goal in machine condition moni- toring. Existing methods for this purpose are mainly based on retrospective analysis, resulting in a large detection delay that limits their usages in real applications. This paper presents a new adaptive real-time change detection algorithm, an extension of the recent research by combin- ing with an incremental sliding-window strategy, to handle the multi-change detection in long-term monitoring of machine operations. In particular, in the framework, Hil- bert space embedding of distribution is used to map the original data into the Re-producing Kernel Hilbert Space （RK_HS） for change detection; then, a new adaptive threshold strategy can be developed when making change decision, in which a global factor （used to control the coarse-to-fine level of detection） is introduced to replace the fixed value of threshold. Through experiments on a range of real testing data which was collected from an experimental rotating machinery system, the excellent detection performances of the algorithm for engineering applications were demonstrated. Compared with state-of- the-art methods, the proposed algorithm can be more suitable for long-term machinery condition monitoring without any manual re-calibration, thus is promising in modern industries.