基于改进支持向量机的多能源电力系统中储能容量规划及运行模型

研究多能源电力系统中储能装置的定容及运行,有利于减小功率波动,降低对电网的冲击,提高电能质量。以青海省海西千万瓦级可再生能源基地为例,首先根据光伏电站和风电场的历史数据分析了两种新能源发电系统的出力特性,在此基础上建立了支持向量机模型,对新能源电站的输出功率进行了短期预测。根据光伏电站和风电场的出力预测误差,建立了ARMA误差预测模型,进一步修正了光伏电站和风电场的预测曲线,最后根据出力预测曲线的功率谱确定了储能系统的容量及出力曲线。研究成果可为新能源并网提供技术支持。

Energy Storage Capacity Planning and Operation Model of Hybrid Power System Based on Improved SVM

Research on energy storage capacity planning and operation mode of hybrid power system can improve power quality, reduce power fluctuation and the impact on power grid. This paper takes Haixi million kilowatt level renewable energy base in Qinghai Province as an example. The output characteristics of two renewable energy systems were analyzed based on the historical data of photovoltaic and wind power station. Then support vector machine (SVM) model was built to predict short-term output power of renewable energy power station. According o the prediction errors of output power, autoregressive moving average (ARMA) error forecasting model was established to modify the prediction output power curve. Finally, the capacity and output curve of storage energy system was given by power spectrum of output power forecasting curve. The results provide technical support for renewable energy grid connection.