基于CSO-SVR的低压架空线路谐波损耗评估

针对架空线路物理解析模型的谐波损耗计算精度不高的问题,文章提出基于纵横交叉算法(Crisscross Optimization,CSO)优化的支持向量回归(Support Vector Regression,SVR)模型对架空线路谐波损耗进行评估。首先采用无需复杂网络结构设计的SVR模型,拟合线路特征与线路损耗之间的关系；然后利用CSO优化算法对SVR超参数进行全局搜索,以动态优化获取最优超参数组,对线路损耗做出评估。文章依托国内某大型电能质量综合试验平台进行低压架空线路谐波试验,获得线路损耗实测数据；利用这一数据对所提模型进行验证。试验结果表明,采用CSO算法对SVR超参数进行优化,可有效提升SVR模型的线损评估性能；对比其他模型,所提模型的谐波线损评估精度更高,评估值更为接近实测值。

Harmonic loss evaluation of low voltage overhead lines based on CSO-SVR model

In view of the low calculation accuracy of the physical analytical model of harmonic loss, the support vector regression (SVR) model optimized by crisscross optimization (CSO) is proposed to evaluate the harmonic loss of overhead lines. Firstly, the SVR model without complex network structure design is used to fit the relationship between line characteristics and line loss; Then the CSO optimization algorithm is applied to search the SVR super parameters globally and obtain the optimal super parameter group dynamically to estimate the line loss. Based on a large power quality test platform in China, the harmonic test of low voltage overhead line is carried out, and the proposed model is verified by the measured data of this test. The experimental results show that using CSO algorithm to optimize the super parameters of SVR can effectively improve the line loss evaluation performance of SVR model; Compared with the other models, the proposed model present the higher accuracy and the harmonic loss evaluation value is closer to the measured value.