基于变异系数和支持向量机的含DG台区线损智能诊断研究
Research on Intelligent Diagnosis of Line Loss in DG Substation Areas Based on Coefficient of Variation and Support Vector Machine
付慧 1史明明 1李双伟 1王靓 1费骏韬 1王浩羽2
作者信息
- 1. 国网江苏省电力有限公司电力科学研究院,江苏南京 211103
- 2. 南京工程学院,江苏南京 211167
- 折叠
摘要
分布式电源的接入改变了台区潮流,增加了台区线路高损概率.为找出引起台区线路高损成因,提出一种基于变异系数法和支持向量机的分布式电源接入台区线损智能诊断方法.首先利用变异系数法对分布式电源接入台区特征指标进行关联性分析;其次,采用支持向量机算法衡量线损预测数据与真实数据相对误差,进而筛选出线损异常台区;然后,基于变异系数法得到的各指标权重更新异常台区数据,诊断线路高损成因;最后,采用25节点台区案例,进行方法验证.试验结果验证了所提方法的有效性.
Abstract
The access of distributed power sources changed the power flow in the substation areas and increased the probability of high losses in the substation lines.To identify the causes of high line losses in substation areas,an intelligent diagnosis method for line losses in distributed generation(DG)access substation areas based on coefficient of variation method and support vector machine was proposed.Firstly,the coefficient of variation method was used to perform correlation analysis on the characteristic indicators of DG access stations;secondly,the support vector machine algorithm was used to measure the relative error between line loss prediction data and real data,and the abnormal line loss areas are then screened out;after that,based on the weight of each indicator obtained by the coefficient of variation method,the abnormal station area data was updated to diagnose the cause of high line loss;finally,a case about 25-node station area was used for method validation.The experimental results validate the effectiveness of the proposed method.
关键词
分布式电源/线路损耗/台区特征指标体系/变异系数法/支持向量机Key words
distributed generation(DG)/line loss/characteristic index system of station area/coefficient of variation method/support vector ma-chine引用本文复制引用
基金项目
国家电网江苏省电力公司科技项目(J2021190)
出版年
2024
NETL
NSTL
万方数据