水电站长期发电优化调度方案风险分析研究

Study on risk analysis of long-term power generation optimal dispatching scheme for hydropower station

长期来水与电价预测的不确定性引起水电站发电量与发电收入的不确定性,定量表述这些不确定性之间的关系,对水电期货市场供需双方均具有重要意义。提出了来水不确定性的描述方法,利用小波周期叠加方法滑动预报,生成预报误差系列;采用优化调度模型生成发电流量和发电收入系列,分析来水预报误差对发电流量和发电收入的影响,并利用一次二阶矩法计算各月的风险率和年度总风险。成果与双风险因子发电调度模型相比较,表明双风险因子发电调度模型中关于发电流量均方差的计算结果在汛期明显偏大,从而导致不同置信度下的发电收入变幅和风险率偏大。

The long-term predicted runoff and electricity price are two key factors that result in uncertainties in the generated energy and hydropower income;a quantitative description of these uncertainties and their relationship is of great importance to supply and demand in the future hydropower market.This paper proposes a quantitative method that uses the wavelet periodic superposition to make a sequential runoff prediction and to generate a series of prediction error.An optimal operation model was adopted to generate a sequence of power generating flows and power income,the effects of inflow prediction error on the sequence and the income was analyzed,and the monthly and annual risk rates were computed by first-order second moment method.Comparison of the present results with that of two-risk-factors model reveals that the mean square deviation of the flood-season generating flows of the two-risk-factors model is significantly larger than the statistical value of these flows.Under different confidence levels,this error will cause a change in the income and a larger risk rate.