基于条件风险的风电场随机优化调度模型及其快速求解方法

摘要

风力发电具有不确定性,且短期/超短期预测精度较低,其大规模并网给电力系统调度带来了新挑战.综合考虑风电场预测出力分布在小概率情形下是不准确的,且风电场实际发电功率应受系统消纳能力上限限制,建立了基于条件风险的风电场随机优化调度模型.模型认为风电场预测出力应服从某一截断的分布,且截断部分的概率值应直接累加到截断处.该模型为混合整数随机优化调度问题,为高效求解此问题,提出蒙特卡洛条件风险方法.最后以江苏某实际电网为例进行计算分析,分析结果表明本文提出的基于条件风险的风电场随机优化调度模型可以较好地解决风电场短期/超短期随机优化调度问题,提出的蒙特卡洛条件风险方法可以高效求解该类混合整数随机优化调度问题.%Wind power is of uncertainty and the prediction accuracy in short-term or ultra short-term is low,so its large-scale connection brings new challenges to power system scheduling. Considering the inaccurate forecasts with a small probability of power output distribution for the wind farm and its actual power should be limited by the upper limit of the system accommodation capacity, this article establishes conditional value at risk based energy stochastic optimal scheduling model for the wind farm.In the model,the distribution function of the wind power output is a truncated distribution,and the probability value of the truncated part should be added directly to the truncation. The model is solved efficiently by the Monte Carlo conditional value at risk(MCCVaR)method for mixed integer stochastic optimal scheduling problem.Finally,the simulation results of a real power system in Jiangsu province show that the wind farm energy stochastic optimal scheduling model based on the conditional value at risk can be applied in short-term or ultra short-term stochastic dispatching,and can be solved efficiently by MCCVaR.