针对HXD2型机车法维莱制动系统存在多种工作模式、结构复杂、故障种类多和系统不确定性大等特点,使用基于解析冗余关系的诊断技术对制动机系统进行故障检测和准确隔离。利用功率键合图对复杂动态系统建模的优势获得制动机的键合图模型,根据键合图模型的因果路径关系推导系统的解析冗余关系,并获得系统故障特征矩阵,再通过系统观测特征与故障特征的比较实现系统的故障检测和隔离。根据这一原理设计的检测与诊断系统已应用于太原铁路局湖东机务段,其仿真实验和现场实际应用说明该诊断系统在保障制动机安全运行方面是有效的。
In view of the inherent characteristics of the multiple working modes, system uncertainty, complicated structure and a large of fault types of the Faiveley braking system of Locomotive HXD2. Fault diagnosis and accurate fault isolation of the Faiveley braking system were carried out by adopting the technique based on the analytical redundancy relations. The bond graph model of the braking system was established by utilizing the advantage of the power bond graph in modeling complex and dynamic systems. The analytical redundancy relations of the braking system were derived according to the causality of the power bond graph and the fault signature matrix of the braking system was obtained thereupon. Fault detection and isolation were realized by comparing observed signatures and fault signatures. The proposed detection and diagnosis system based on the analytical redundancy relations has already been put into practice in the Hudong Locomotive Depot of the Taiyuan Railway Administration. Simulation tests and practical applications show that the proposed detection system is effective in protecting safe operation of the braking system.