基于证据熵多源组合规则的风力发电机组故障诊断方法

针对证据理论在风力发电机故障融合诊断中存在的高冲突问题,提出一种基于证据熵的多源融合组合规则.由风力发电机历史故障特征数据模糊化获取原始证据,根据多类传感器实时数据重要性的不同,采用信息熵原理得到各证据的重要性参数即权重,对加权修改后的证据进行Dempster融合得到最终结果,最后基于决策准则作出故障诊断.风力发电机故障诊断实例表明,本方法在一定程度上降低了证据之间的冲突性,提高了故障诊断的准确率.

Wind Turbine Fault Diagnosis Based on Multi-Source Composition Rules of the Evidence Entropy

In order to solve the high conflict problem in the integration process of traditional combination rules of the evidential theory a multi-source combination rule based on the evidence entropy is proposed. According to the different importance of evidences obtained from multiple sensors the importance parameters i. e. ,the weights of each evidence, are given based on the evidence entropy principle. Combining the evidences on the weighted with Dempster rule to get the final results we can get the final fault diagnosis based on the decision criterion. The diagnosis results of a wind turbine show that the proposed method can reduce the conflicts of evidences and then enhance the diagnostic accuracy.