CAJ | 학술논문

针对铁路客运站行车事故致因过程的动态性以及事故原因的模糊性，提出了一种 Petri 网图形建模同模糊控制理论相结合的分析方法。根据具体的事故产生式规则，建立了行车事故的模糊 Petri 网模型，进一步又将神经网络的BP算法引入到模型参数的学习中，通过给出的模糊推理算法，推理计算出事故的主要原因及可信度。最后将该方法应用到某客运站行车事故分析中，结果证明了其在事故分析中的准确性和可行性。
침대철로객운참행차사고치인과정적동태성이급사고원인적모호성，제출료일충 Petri 망도형건모동모호공제이론상결합적분석방법。근거구체적사고산생식규칙，건립료행차사고적모호 Petri 망모형，진일보우장신경망락적BP산법인입도모형삼수적학습중，통과급출적모호추리산법，추리계산출사고적주요원인급가신도。최후장해방법응용도모객운참행차사고분석중，결과증명료기재사고분석중적준학성화가행성。
Based on the dynamic development of train operation accidents of railway passenger sta-tion and the fuzziness of accident causes,a method combining Petri net and fuzzy control theory is proposed.The train operation accident model is built based on fuzzy Petri net according to acci-dent generated rules.BP algorithm of neural networks is introduced into the learning of model pa-rameters.Then the reasons and reliability function of accidents are reasoned by fuzzy reasoning al-gorithm.Finally,the proposed method is applied to the overrunning of signal accident analysis. The results show that the method is accurate and feasible in the accident analysis.