基于改进logistic模型法的配电网负荷预测

Distribution network load forecasting based on PSO-logistic model method

为了达到“双碳”目标,新开发地区的配电网需要进行目标网架结构规划,目标网架的搭建又离不开准确的负荷预测。常规配电网规划通常以自然增长法和点负荷法求取过渡年负荷增长速度,误差往往较大。本文考虑新开发地区负荷增长规律,先使用空间负荷预测进行饱和负荷预测,然后以空间负荷预测的值作为饱和值,利用logistic法进行近期负荷预测,并运用粒子群算法对logistic的参数进行寻优。最后通过以华东某地区2001年至2020年历史负荷数据与该地区发展规划为分析案例,印证了该预测模型相比于传统模型以及未改进模型的优越性。

In order to achieve the dual carbon goal,the distribution network in the newly developed area needs to carry out the target grid structure planning,and the construction of the target grid is inseparable from accurate load forecasting.Conventional distribution network planning usually uses the natural growth method and the point load method to obtain the transitional annual load growth rate,and the error is often large.This paper considers the load growth law of the newly developed area,first uses the spatial load forecasting to forecast the saturated load,then uses the value of the spatial load forecast as the saturated value,uses the logistic method to forecast the short-term load,and uses the particle swarm algorithm to optimize the logistic parameters.Finally,by taking the historical load data from 2001 to 2020 in a certain region in East China and the development plan of the region as an analysis case,the superiority of the prediction model compared with the traditional model and the unimproved model is confirmed.