基于组合模型的轨道电路复杂故障诊断方法研究

Research on Diagnosis Method of Complex Fault Diagnosis of Track Circuit Based on Combined Model

由于轨道电路的性能受环境因素的影响较大,其故障的发生具有较大的随机性和模糊性,而单一诊断方法因考虑角度的片面性以及知识库的不完善性,往往很难满足轨道电路故障诊断的实际需求。鉴于单一诊断方法存在的不足,本文根据模糊故障诊断法、遗传算法和灰色系统理论3种诊断方法各自的诊断结果,构造最优组合模型,对25Hz相敏轨道电路进行故障诊断。诊断结果表明,利用组合模型对轨道电路故障进行诊断可以克服单项诊断方法信息单一、诊断片面等不足,具有更高的故障诊断准确度。

Because track circuit performance is strongly influenced by environmental factors, the occurrence of faults is of great randomness and fuzziness. In view of one sided the approach to problems and imperfection of the knowledge base, the single diagnosis method usually cannot meet the actual track circuit fault diagnosis re quirements. Aiming at the insufficient of single diagnosis method, according to the respective diagnosis results of the fuzzy fault diagnosis method, genetic algorithm and grey system theory, the optimal combination model for fault diagnosis of track circuits was made and applied to the 25 Hz phase sensitive track circuit. The diag nostic results indicate that the combination model can overcome the insufficiency of singularity of information and one sided diagnosis of the single diagnosis method and achieve higher diagnosis accuracy.