基于多电参量采集的分布式园区光伏发电预测系统设计  

Design of Distributed Park Photovoltaic Power Generation Prediction System Based on Multi-electrical Parameter Acquisition

作　　者：刘静利 刘传良 刘忠辉 温国强 LIU Jingli;LIU Chuanliang;LIU Zhonghui;WEN Guoqiang(Weifang Power Supply Company,Weifang 261000)

机构地区：[1]国网山东省电力公司潍坊供电公司,山东潍坊261000

出　　处：《电力安全技术》2025年第3期53-57,共5页Electric Safety Technology

基　　金：国网山东省电力公司潍坊供电公司群众性创新项目(2024-YF010)。

摘　　要：针对分布式园区光伏发电功率预测准确率难以提升进而影响调度控制的问题,提出了基于多电参量采集的分布式园区光伏发电预测系统,并设计了基于多电参量采集的硬件装置和基于预测修正的区域分布式光伏发电功率预测算法;构建预测日和相似日光伏出力偏差和气象影响因素偏差的预测模型,结合历史经验值和深度学习技术,提升了光伏预测的准确率。现场应用表明,多电参量采集装置可实时准确采集光伏发电并网数据,预测算法可较准确地预测光伏发电功率。In view of the difficulty to improve the accuracy of photovoltaic power generation prediction which affects the scheduling control,a distributed park photovoltaic power generation prediction system is proposed based on multi-electrical parameter acquisition,as well as designing a hardware device based on multi-electrical parameter acquisition and the distributed photovoltaic power generation prediction algorithm based on prediction correction.A prediction model for the photovoltaic output deviations of prediction days and the similar selected days and for the deviations due to meteorological factors is established as well as combining with the historical experience and deep learning techniques,which improves the accuracy of photovoltaic prediction.The field application shows that the multi-power parameter acquisition device is available to accurately collect the grid-connected data of photovoltaic power generation in real time,and the prediction algorithm enables to more accurately predict the capacity of photovoltaic power generation.

关 键 词：多电参量 分布式园区 光伏发电 电网调度 发电预测 

分 类 号：TM615[电气工程—电力系统及自动化]