计及价格型负荷响应不确定性的概率潮流计算

Probabilistic Load Flow Calculation Considering Price-elasticity Load Uncertainties

随着以风电为代表的间歇性电源的快速增长,未来电力系统将从"发电跟踪负荷"的单向调度模式发展到"源—网—荷"智能互动调度,同时电源侧和负荷侧均具有不确定性。针对上述问题,文中综合考虑风电出力波动、负荷随机变化以及价格型负荷参与电网互动的随机响应过程三方面的不确定因素。首先,描述了计及不确定性的价格型负荷响应概率表征形式,通过价格弹性系数,分析了价格型负荷响应的不确定性及响应范围,以反映出响应行为的随机性和主观性。其次,在传统概率潮流计算的数学模型基础上,将价格型负荷不确定响应量视做新的随机注入变量,引入源荷随机互动概率潮流模型中,并应用点估计法求解得到网络潮流的统计特征值,使得价格型负荷参与电网互动机理研究更具全面性和准确性。最后,通过IEEE 73节点系统的算例分析表明,该模型通过价格型负荷响应能够填补风电出力波动引起的系统功率缺额,降低对支路潮流造成的不利影响。另外,点估计法能同时处理离散型和连续型不确定因素,精度高,计算量小,具有实用价值。

With the rapid increase of intermittent energy,typically wind energy,the dispatch mode of power system will develop from “generation tracking load' dispatch to the“source-grid-load' smart interactive dispatch.At the same time,both the power supply side and load side have uncertainties.In order to deal with the problem,this paper considers the uncertain factors in three aspects,the uncertainty of wind power output fluctuation,that of load variation,and that of the interactive process.First, the form of characterization for the price elasticity load response probability considering uncertainty is described.Through the price elasticity coefficients,the uncertainty of the price elasticity load response and the range of response are analyzed to reflect the randomness and subjectivity of the response behavior.Then,based on the traditional mathematical model of probabilistic power flow calculation,the uncertainty demand response quantity is regarded as randomly injected variables and introduced in the “wind power-price elasticity load' interactive probabilistic load flow model to improve the comprehensiveness and accuracy of the demand response mechanism.And the point estimation method is used to calculate the statistical characteristics of power flow value.The digital simulation results of an IEEE 73-bus system show that the model uses the price-elasticity load response to bridge the gap due to the power output fluctuation of wind power.Also,the adverse impact on branch power flow is reduced.The point estimation method used to solve the model is expected to effectively deal with both discrete and continuous uncertainties thanks to the high precision and light workload.