摘要
为了更好地研究变压器油纸绝缘老化状态与拓展德拜等值电路参数之间的相关性,根据实测回复电压特征量参数等值电路模型,建立求解等值电路参数的数学模型,并将该模型转化为非线性优化问题,利用混合蛙跳算法的全局信息交换和局部深度搜索特性对拓展德拜等值电路参数进行参数辨识。通过两台变压器的辨识结果表明,与粒子群算法相比,由混合蛙跳算法辨识的电路参数计算获得的回复电压值与测量值具有更高的重合度,并能够准确地反映变压器油纸绝缘状态。
In order to study the correlation between the condition of oil-paper insulation of transformer and the extended Debye equivalent circuit parameters,parametric equivalent circuit model was established based on observed recovery voltage characteristics.And then the model was converted into nonlinear optimization problem.The global information exchange and local depth search features of shuffled frog leaping algorithm were applied to identify the equivalent circuit parameters.Compared with particle swarm algorithm,the identified results of the two transformers by shuffled frog leaping algorithm show that it does has a higher coincidence degree and can accurately reflects the insulation state of the oil-paper transformer.
出处
《水电能源科学》
北大核心
2016年第10期186-188,206,共4页
Water Resources and Power
基金
国家自然科学基金项目(51307097)
关键词
油纸绝缘变压器
拓展德拜等值电路
回复电压极化谱
峰值电压
混合蛙跳算法
oil-paper insulation transformer
extended Debye equivalent circuit
recovery voltage spectrum
peak voltage
shuffled frog leaping algorithm