基于遗传优化支持向量机的变压器绕组热点温度预测模型

发布时间：2018-09-17 18:34

【摘要】：油浸式电力变压器的运行寿命及负载能力与绕组热点温度密切相关。精确预测变压器绕组的热点温度,是有效预防变压器热故障、准确预测变压器运行寿命和优化变压器设计的关键技术之一。论文研究了绕组热点温度支持向量机建模。为提高模型预测的精确度,选用径向基核函数优化模型结构;利用遗传算法对参数进行寻优。结合实验室模拟温升变压器绕组温度实测数据,提取输入和输出的特征量,并划分训练集和预测集,建立了基于遗传优化支持向量机的变压器绕组热点温度预测模型。实验表明:应用本文模型预测结果与实测值基本一致,优于BP神经网络以及Elman神经网络的预测结果。
[Abstract]:The operation life and load capacity of oil-immersed power transformer are closely related to the winding hot spot temperature. Accurate prediction of the hot spot temperature of transformer windings is one of the key techniques to prevent transformer thermal faults accurately predict transformer operation life and optimize transformer design. In this paper, support vector machine (SVM) modeling of winding hot spot temperature is studied. In order to improve the accuracy of the model prediction, the radial basis function is used to optimize the model structure, and the genetic algorithm is used to optimize the parameters. Based on the measured temperature data of simulated temperature rise transformer windings in laboratory, the characteristic values of input and output are extracted, and the training set and prediction set are divided, and a prediction model of hot spot temperature of transformer windings based on genetic optimization support vector machine is established. The experimental results show that the predicted results of the proposed model are in good agreement with the measured values, and are superior to those of the BP neural network and the Elman neural network.
【作者单位】：重庆大学输配电装备及系统安全与新技术国家重点实验室;国家电网重庆市电力公司市区供电分公司;
【基金】：国家创新研究群体基金(51021005) 国家重点基础研究发展计划(973)(2012CB215205)资助项目
【分类号】：TM411

【参考文献】