基于置信规则库专家系统的司控器开关量健康状态评估

微动开关是轨道车辆司控器常用的开关设备,对其健康状态评估是保证轨道车辆运行安全的前提.针对司控器微动开关数据样本少、诊断信号具有波动性和非线性、健康状态评估困难等问题,提出一种基于置信规则库专家系统(BRB)的司控器开关量健康状态评估方法.首先,分析微动开关失效机理与故障特征的关系;然后,采用置信规则库将定性知识与定量信息有效结合,采用证据推理(ER)算法进行知识推理,并对所建立的模型初始参数进行优化,得到最优的参数集合,从而提高轨道车辆微动开关健康状态评估的准确性.通过对模型训练及测试,所得结果表明,所提出的方法能准确地评估微动开关状态,便于早期发现故障、跟踪故障发展趋势和及时更换失效部件.

A state estimation method for driver controller's microswitch based on belief rule base

The microswitch is a common switchgear of the railway vehicle driver controller, and the evaluation of its health condition is the prerequisite to ensure the safety of railway vehicle operation. Aiming at the problem that the data of the controller microswitch is few, the signal is nonlinear, and the health status is difficult to assess, a method for evaluating the health status of microswitch based on the belief rule base(BRB) is proposed. Firstly, the relation between the failure mechanism and the fault characteristics of the microswitch is analyzed, and the qualitative knowledge and quantitative information are effectively combined with the BRB, the knowledge is reasoned by using the evidential reasoning(ER) algorithm, the initial parameters of the model are optimized, and the optimal set of parameters is obtained, which improves the accuracy of evaluating the health status of the microswitch. Through the training and testing of the model, the experimental results show that the method can accurately evaluate the fault of the microswitch in railway vehicle, which is convenient to detect faults early, track the development trend of failure and replace the failed components in a timely manner.