基于气象特征量选取与SVM模型参数优化的新能源超短期功率预测

ULTRA-SHORT-TERM POWER FORECAST OF NEW ENERGY BASED ON METEOROLOGICAL FEATURE SELECTION AND SVM MODEL PARAMETER OPTIMIZATION

提出基于海量气象特征量选取与支持向量机(SVM)模型参数优化的新能源发电超短期功率预测方法,以提高新能源发电预测精度。首先研究基于皮尔逊相关系数的气象特征量提取方法,并利用粒子群算法(PSO)对支持向量机(SVM)新能源发电预测模型参数进行优化,进一步提出联合气象特征选取与模型参数优化的新能源发电功率超短期预测模型以得到全局最优解。然后结合历史发电功率数据,研究新能源发电功率超短期滚动预测模型。最后利用国内某风电场数据进行对比验证,证明所提预测模型可有效提高新能源发电预测精度。

Under the major strategic deployment of building a new power system and realizing the"dual carbon"goal,new energy represented by Wind power generation and photovoltaic power generation will usher in a period of rapid development.New energy power generation is random,intermittent and volatile,and the integration of large-scale new energy power generation into the power grid will seriously affect the security,stability and economic operation of the power grid.Therefore,an ultra-short-term power prediction method for new energy power generation based on the selection of massive meteorological feature quantities and parameter optimization of support vector machine(SVM)model is proposed to improve the prediction accuracy of new energy power generation.Firstly,the meteorological feature extraction method based on the Pearson correlation coefficient is studied,and the parameters of the new energy power prediction support vector machine(SVM)model are optimized by using the particle swarm optimization(PSO).The ultra short term prediction model of new energy power generation combined with meteorological feature extraction and model parameter optimization is further proposed to obtain the global optimal solution.The ultra short term rolling prediction model of new energy power generation is studied by using historical power generation data.Finally,the data of a domestic wind farm is used for comparison and verification,which proves that the proposed prediction model can effectively improve the prediction accuracy of new energy power generation.