基于混合智能算法的电力变压器故障诊断研究

最小二乘双支持向量机算法(LSTSVM)具有训练样本速度快、二分类效率高的特点,在电力系统电力变压器故障诊断中占有独特优势。单一的智能算法诊断自己片面,很难全面进行故障诊断,因此多采用复合智能算法。在LS-TSVM模型基础上引入蚁群算法,利用蚁群算法强大的搜索能力进行寻优计算,结合二叉树构建的LS-TSVM模型可对变压器故障进行全面诊断。通过实际的算例进行仿真,结果表明,混合智能故障诊断方法不仅准确率高,准确度也比传统ANN模型有所提高,证明了该算法模型的有效性和实用性。

Research on Fault Diagnosis of Power Transformer Based on Hybrid Intelligent Algorithm

LS-TSVM(Least Square-Double Support Vector Machine)has the characteristics of fast training sample and the high efficiency of classification.It has unique advantages in fault diagnosis method for power transformer of power system.Because a single intelligent algorithm can only diagnose from one aspect,it is difficult to diagnose comprehensively.So compound intelligent algorithm is often used.In this paper,ant colony algorithm is introduced on the basis of LS-TSVM model.According to the strong search ability of ant colony algorithm,it is easy to optimize the calculation.LS-TSVM model constructed by combining the optimization calculation of ant colony algorithm with binary tree can be very good for transformers.The simulation results show that the proposed method is not only more accurate,but also more accurate than the traditional ANN model,which proves the validity and practicability of the algorithm model.