CAJ | 학술논문

在目前多专家故障诊断系统采用的模糊综合评判方法基础上，根据某特定故障的理论诊断结果及实际检测、维修结果，利用贝叶斯理论，对特定故障下各专家的初始权重集进行实时更新，实现了模糊综合评判的自学习，提高了诊断的准确性。最后给出了方法在“康明斯柴油机故障诊断专家系统”中的应用实例，验证了方法的有效性。
재목전다전가고장진단계통채용적모호종합평판방법기출상，근거모특정고장적이론진단결과급실제검측、유수결과，이용패협사이론，대특정고장하각전가적초시권중집진행실시경신，실현료모호종합평판적자학습，제고료진단적준학성。최후급출료방법재“강명사시유궤고장진단전가계통”중적응용실례，험증료방법적유효성。
Based on fuzzy synthetic evaluation method of multi-expert fault diagnosis system and according to the theoretical diagnosis result,the actual detection and maintenance result of a cer-tain fault,the weights of multi-expert are updated in real time using Bayesian theory.So the self-learning fuzzy synthetic evaluation is carried out and the diagnostic accuracy is improved.Finally an applicable case is given in the Cummings Diesel Engine Fault Diagnosis Expert System,which demonstrates the feasibility of the method.