光伏发电功率区间概率预测

The Interval Probability Prediction of Photovoltaic Generation

随着光伏发电技术的大量应用,光伏电站出力的波动性和不确定性对电网的影响也日益突出,光伏发电功率预测对预防这种不利影响具有指导意义。概率预测在通常的点预测基础上给出预测结果的概率特征,更有助于电能分析和电网调度。在对光伏发电功率影响因子进行独立分量分析的基础上建立了条件概率预测模型,预测光伏发电功率的区间概率。仿真结果表明,条件概率预测比无条件概率预测更能客观地反映出影响因子对预测结果的影响,为电力系统能量调度提供可靠的决策依据。

Along with the large number of application of photovoltaic power generation technology, photovoltaic power station out-put has put an increasingly prominent influence on the grid because of its volatility and uncertainty. The prediction of photovoltaic power generation has a guiding significance to prevent the adverse influence. Probability prediction can describe probability characteris-tics of the prediction results based on the generally point prediction, which presents more help for electricity analysis grid scheduling. In the paper, the conditional probability prediction model is established to predict the interval probability of photovoltaic generation based on independent component analysis of the impact factors. Simulation results show that the conditional probability prediction is more objective than unconditional probability prediction in revealing the influence of impact factors on prediction results, and the pro-posed model can provide reliable decision basis for power system dispatching system.