摘要
变压器状态评估是变压器运行、检修与维护的重要依据,而其状态参数的类型杂、数量多及参数间关联关系不明确等是该环节工作的瓶颈问题。针对上述问题,文中构建了基于多信息融合的状态综合评估模型。在深入分析变压器状态评估的需求与研究现状的基础上,利用变压器的试验数据、运行数据、历史数据,设计了基于粗糙集、层次分析法、多神经网络、证据理论的多信息融合评估模型,且通过算例进行校验。结果表明:所提模型能有效对变压器状态进行评估,可为当前变压器的运行与维护提供参考。
Condition assessment is important to the operation, repair and maintenance of transfoimer. However, the complex state parameters and the unclear relationship between the parameters are the bottleneck problems. Therefore, this paper constructs the state evaluation model of transformer operation based on multi-information fusion. Base on the further analysis of demand and present researches for the state evaluation, the test data, the operation data and the history data are used to establish a evaluation model based on the multi-information fusion, which includes the rough set, the analytic hierarchy process, multi-neural network theory and evidence theory. And then, the model is checked by the examples. The results show that the proposed model can effectively evaluate the transformer state, which can be the reference for the operation and maintenance of transformers.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2013年第4期140-144,共5页
Proceedings of the CSU-EPSA
关键词
状态评估
变压器
粗糙集
层次分析法
神经网络
证据理论
condition assessment
transformer
rough set
analytic hierarchy process
neural network
evidence theory
