基于贝叶斯理论的自学习模糊综合评判研究  

Self-Learning Fuzzy Synthetic Evaluation Method Based on Bayesian Theory

作　　者：杨智[1] 李金寿[1] 董华玉[1] 

机构地区：[1]镇江船艇学院船艇装备保障系,江苏镇江212003

出　　处：《军械工程学院学报》2014年第1期47-50,共4页Journal of Ordnance Engineering College

基　　金：军队科研计划项目

摘　　要：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.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.

关 键 词：自学习模糊综合评判 贝叶斯理论 多专家系统 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]