基于卡尔曼滤波和专家系统的混合故障预测研究

基于卡尔曼滤波和专家系统的相关知识,研究了动态系统内部故障的预测问题.首先根据系统的运行情况提取表征故障征兆的状态,建立对应于故障点的状态方程模型并进行分析;然后在定义了故障预测的优先原则的基础上,利用专家系统加权算法得到的故障预测容许值集合,设计了故障预测的流程图;最后根据流程图给出一种混合的故障预测算法.结果表明,这种算法与传统的故障诊断方法和已有的故障预测方法相比有比较明显的优势,并指出了进一步研究的重点.

Mixed Fault Prediction Based on Kalman Filtering and Expert System

Based on Kalman filtering and expert system,the inner fault prediction problem of the dynamic system is researched.First of all,the state expressing the fault symptom is extracted in accordance with the running circumstances of system,then the state equation model is established and analyzed corresponding to the fault point.Secondly,based on the defined priority principle of Fault Prediction and using the Fault Prediction allowance value set got by weighted algorithm,a sequential graph of FP is designed.Finally,a kind of mixed fault prediction algorithm is presented.The presented algorithm has obvious advantage compared with the traditional fault diagnosis and the former fault prediction.And,the future research emphasis is pointed out.