考虑风力发电随机性的超短期风电功率区间预测研究

ULTRA-SHORT-TERM WIND POWER INTERVALS PREDICTION CONSIDERING RANDOMNESS OF WIND POWER GENERATION

首先给出一种评估风电功率区间预测效果的新优化准则,基于人工蜂群算法-神经网络构建简易风电功率区间预测模型,将区间预测模型与马尔科夫链预测模型相结合,对区间内数值点进行概率分析,并通过置信区间修正马尔科夫链预测结果。仿真结果表明,该预测方法不仅能准确预测风电功率置信区间,还可从概率的角度对置信区间内数值点进行分析,提高风电功率预测精度,为优化系统提供依据。

Wind power generation has the characteristic of randomness, and it is significant to predict intervals of uhra- short-term wind power precisely for optimizing the grid power operation scheduling and reserve capacity. The paper proposes a new criterion based on modified root-mean-square prediction interval and mean offset. A simple short-term wind power intervals prediction model based on artificial bee colony neural network is developed to get better forecasting. The integration model of intervals prediction and Markov chain is proposed to analyze the probabilities of wind power and amend predicted results. The simulation result demonstrates that the proposed method can accurately predict the wind power intervals and the probabilities of wind power intervals. The conclusion provides a theoretical basis for optimizing system operation and future analyzing.