典型分布式光伏出力预测数据集

光伏电站出力数据每5分钟从电站机房监控系统获取；气象实测数据从气象站获取，气象站建于电站30号箱变附近，每5分钟将采集的数据通过光纤传输到机房；数值天气预报数据利用中国电科院新能源气象应用机房的WRF业务系统（包括30TF计算刀片机、250TB并行存储）进行中尺度模式计算后输出预报产品，每日8点前通过反向隔离装置推送到电站内网预测系统。

The output data of the photovoltaic (PV) power station is collected every 5 minutes from the monitoring system in the station's machine room. The meteorological in-situ measurement data is acquired from a meteorological station, which is situated near the No. 30 box-type transformer of the power station. The collected data is transmitted to the machine room via optical fiber every 5 minutes. The numerical weather prediction (NWP) data is processed through mesoscale model calculations using the WRF operational system in the new energy meteorology application machine room of China Electric Power Research Institute (CEPRI), which is equipped with 30TF computing blade servers and 250TB parallel storage, to produce forecast products. These forecast products are pushed to the intranet prediction system of the power station via reverse isolation devices before 8 a.m. every day.

搜集汇总

背景与挑战

背景概述

该数据集专注于分布式光伏出力预测，包含光伏电站出力、气象实测和数值天气预报三类数据，时间分辨率为5分钟，用于支持短期和超短期预测研究。数据来源于实际监控系统和专业气象模型，是国家重点研发计划项目的成果，适用于可再生能源发电技术领域的算法开发和验证。