基于改进的支持向量机理论在配电线路等值覆冰厚度预报中的应用研究

Application of Improved Support Vector Machine Theory to the Prediction of Equivalent Icing Thickness of Distribution Lines

针对目前线路防冰抗冰工作重心主要偏重于主网输电线路弱化了配网防冰工作的缺陷。文章基于支持向量机SVM原理,采用交叉验证法对模型参数进行最优配置,并根据RBF核函数建立最小二乘支持向量机(LS-SVM)预测模型,对配电线路三次覆冰过程中等值覆冰厚度进行预测研究。结果表明:三次覆冰过程中LS-SVM模型与实际覆冰厚度测量值平均相对误差分别为5.46%、2.28%、2.64%,可以看出文章构建的LS-SVM模型具有较好的预报效果,预测精度相对较高。文章研究所得结论对电网公司各级系统运行部应发布配电线路融冰计划,并对配电融冰线路停电进行风险分析,制定并落实风险控制措施具有科学的参考意义。

At present, the main focus of anti-icing and de-icing work is mainly on the transnfission lines of the main network, which weakens the defects of the anti-icing work of the distribution network. Based on the principle of support vector machine （SVM）, the cross-validation method is used to optimize the parameters of the model, and the prediction model of least square support vector machine （LS-SVM） is established according to the RBF kernel function. In this paper, the equivalent ice thickness of distribution lines in the process of three times icing is predicted. The results show that the average relative errors between the LS-SVM model and the actual ice thickness measurements are 5.46%, 2.28% and 2.64%, respectively. It can be seen that the LS-SVM model constructed in this paper has a good prediction effect. The prediction accuracy is relatively high. The conclusion of the paper has scientific reference significance for the power grid company at all levels of the system operation department should release the distribution line ice-melting plan and analyze the risk of distribution ice-melting line blackout and formulate and implement risk control measures.