多信息量融合的电力变压器状态评估模型

Condition Assessment of Power Transformer Using Information Fusion

为提高电力变压器状态评估的准确性,促进检修模式向状态检修技术发展,以220kV等级油浸式电力变压器为例,在实际调研的基础上研究了变压器运行状态与评估指标数值之间的隶属关系和指标初始值差异对评估结论的影响,应用未确知数学理论对专家的评估指标权重量化结果进行处理。建立了基于半哥西分布函数的绝缘油试验评估指标状态函数和考虑指标初始值差异的电气试验评估指标状态函数。实例分析表明,所建立的各试验指标的状态函数和基于未确知有理数的指标权重求解方法可以更加便捷地实现变压器的状态评估,为检修决策提供合理建议。

In order to improve the accuracy of transformer condition assessment,and to promote the maintenance model to transform to the condition based maintenance,a 220 kV oil-immersed power transformer was studied.Based on the field investigation,the membership between index value and transformer operating status and the impact of different initial index value was analyzed.The unascertained mathematics was used to process the assessment indicators’ quantification results given by experts.The state function of insulating oil test based on the half Cauchy distribution function and the state function of electrical test considering the initial value differences were constructed.Case analysis verifies that the state function of each evaluate index and the index weight calculation method based on unascertained rational number can easily evaluate the status and provide a reasonable proposal in condition based maintenance of power transformer.