Recursive Total Principle Component Regression Based Fault Detection and Its Application to Vehicular Cyber-Physical Systems
Recursive Total Principle Component Regression Based Fault Detection and Its Application to Vehicular Cyber-Physical Systems
复制标题
基于递归全原理分量回归的故障检测及其在车载信息物理系统中的应用
DOI：
10.1109/tii.2017.2752709
复制
发表时间：
2018-04-01
影响因子：
12.3
通讯作者：
Yin, Shen
中科院分区：
文献类型：
--
作者：
Jiang, Yuchen;Yin, Shen
信息物理系统（CPS）是工业4.0时代的核心研究课题。这类系统在物理实体和抽象信息之间进行着强烈的交互，普遍存在于工业过程和人们的日常生活中。本文分析了车载CPS在线实现的实际困难，在此基础上提出了一种具有模块化和可扩展性的故障诊断与控制体系结构。阐述了在线跟踪技术和集成学习方法都可以提高系统的自适应能力。对于车载CPS的部署，实时能力的需求是关注的焦点。提出了一种新的基于递归全主元回归的数据驱动故障检测方法。在Carsim上进行了仿真测试，将所提出的方法与现有的多种方法进行比较。
The cyber-physical systems (CPSs) are the central research topic in the era of Industrial 4.0. Such systems interact intensively between physical entities and abstract information, and commonly exist in the industrial processes and people's daily lives. This paper investigates the practical difficulties of the vehicular CPSs online implementation, and based on that proposes a fault diagnosis and control architecture with modular units and reserved extendibility. It is elaborated that the systems’ adaptability could be enhanced by either the online tracking techniques or the ensemble learning schemes. For the onboard deployment of automobile CPSs, the requirement of real-time capacity is in focus. A new recursive total principle component regression based design and implementation approach is proposed for efficient data-driven fault detection. Simulation tests were carried out on the Carsim to compare the proposed approach with multiple existing methods.