分布式故障诊断管理研究

Management and Research of Distributed Fault Diagnosis

复杂武器系统故障诊断是武器型号后勤保障的难题。研究复杂武器系统的故障诊断管理,能提高武器系统故障诊断的准确性和效率。采用分布式故障诊断体系结构,按照故障诊断任务流引擎规划诊断流程,使用任务分解与或树进行诊断任务分解,基于改进合同网协议安排任务分配流程,进行诊断任务分配,各诊断专家诊断结果采用模糊信息融合算法汇总诊断结果,用专家系统推导设计故障诊断单元。诊断系统综合了复杂武器系统多个故障诊断专家的判断,有效提高诊断质量。对建造分布式武器故障诊断系统有参考价值。

Fault diagnosis of complicated weapon system is one difficult problem in weapon logistics support. To research the fault diagnosis management of complicated weapon system can improve correctness, reliability and efficiency of weapon system fault diagnosis. By use of distributed fault diagnosis system structure, according to fault diagnosis flow arranged by fault diagnosis task flow engine, the diagnostic mission was decomposed by use of or/and tree. The assignment process and distribution of diagnostic mission were scheduled based on an improved CNP （Contract Net Protocol ）. All of the diagnostic results were collected by fuzzy information fusion algorithm. The fault diagnosis unit was derived by expert system in the procedure. Therefore, the proposal fault diagnosis system can synthesize all of the judgments of fault diagnosis experts in process of complex weapon system fault diagnosis so as to effectively improve the quality of diagnosis. The studied results can provide valuable reference for construction of distributed fault diagnosis system.