基于随机森林回归模型的台区负荷不平衡控制方法  

A Control Method for Load Imbalance in Substation Areas Based on Random Forest Regression Model

作　　者：李想 王可 张高山 冯金兰 LI Xiang;WANG Ke;ZHANG Gaoshan;FNEG Jinlan(NARI-TECH Nanjing Control Systems LTD,Nanjing 210000,China;NARI Group Corporation,Nanjing 210000,China)

机构地区：[1]国电南瑞南京控制系统有限公司,南京210000 [2]南瑞集团有限公司,南京210000

出　　处：《自动化与仪器仪表》2025年第3期325-328,334,共5页Automation & Instrumentation

基　　金：国网能源控制器专变研发与应用项目(52460 9230050)。

摘　　要：随着电力系统的不断发展,台区负荷不平衡问题日益凸显,对电力系统的稳定性和经济性造成显著影响。现有方法未充分利用历史负荷数据,难以实现高精度的短期负荷预测,设计基于随机森林回归模型的台区负荷不平衡控制方法。通过智能电表采集台区负荷数据。假设在短时间内数据存在横向相似性,即某一特定时刻的负荷数据与相邻同类日的负荷数据相近。基于此假设,结合统计学原理,利用样本的统计指标与预先设定的阈值来检测是否存在异常数据,实施非正常数据的修正。基于随机森林回归模型设计台区短期负荷预测模型,实施台区短期负荷预测。为优化传统模型预测控制算法的性能,在滚动优化步骤之前,先对这27个基础电压矢量实施初步筛选。通过优化后的模型预测控制算法实现台区负荷不平衡控制。测试结果表明设计方法控制后的波形实现了三相平衡,不平衡控制性能表现良好。With the continuous development of the power system,the problem of load imbalance in substations is becoming increasingly prominent,which has a significant impact on the stability and economy of the power system.The existing methods do not fully utilize historical load data,making it difficult to achieve high-precision short-term load forecasting.Therefore,a substation load imbalance control method based on random forest regression model is designed.Collect substation load data through smart meters.Assuming that there is horizontal similarity in the data within a short period of time,that is,the load data at a specific moment is similar to the load data of adjacent similar days.Based on this assumption,combined with statistical principles,the statistical indicators of the sample and pre-set thresholds are used to detect abnormal data and implement correction for abnormal data.Design a short-term load forecasting model for the substation area based on the random forest regression model,and implement short-term load forecasting for the substation area.To optimize the performance of traditional model predictive control algorithms,preliminary screening is performed on these 27 basic voltage vectors before the rolling optimization step.Realize load imbalance control in the substation area through optimized model predictive control algorithm.The test results show that the waveform controlled by the design method achieves three-phase balance,and the imbalance control performance is good.

关 键 词：随机森林回归模型 台区短期负荷预测 台区负荷不平衡控制 

分 类 号：TN464[电子电信—微电子学与固体电子学]