变压器油中溶解气体的多智能体故障诊断方法

Multi-Agent fault diagnosis method based on dissolved gases in transformer oil

为解决目前基于油中溶解气体分析(DGA)方法的变压器故障诊断系统中存在的模型单一化、自适应能力差及容易出现误诊和漏诊的问题,在对现有智能故障诊断系统和变压器DGA方法的研究基础上,提出了基于多智能体(Agent)系统的变压DGA智能故障诊断系统协作模型,给出了模型内的主要Agent、黑板、JavaAgent开发环境(JADE)总线的功能,并详细论述了基于多Agent系统的变压器DGA智能故障诊断协作过程,设计了多专家Agent投票诊断算法及基于黑板的资源竞争算法。最后通过应用实例,表明了基于多Agent系统的变压器油中溶解气体智能故障诊断方法的可行性和有效性。

The current transformer fault diagnosis system has problems including simplified model, weak adaptability and prone to error or miss diagnosis. In order to solve these problems, a eollaborative transformer DGA model used for intelligent fault diagnosis system based on muliti-Agent system is proposed. The paper presents the main function of each Agent, blackboard and JADE bus in this model, discusses the collaborative process of intelligent fault diagnosis system for transformer DGA based on multi-Agent system, puts forward an expert voting diagnosis algorithm based on multi-Agent and a competitive algorithm for resource based on Blackboard method. The application examples show the feasibility and effectiveness of the proposed method.