摘要
针对含风电接入的电力系统,建立了以总燃料耗量和购电费用最小为目标的多目标随机动态经济调度模型。借助场景法将该模型转化为大规模多目标确定性动态经济调度模型。采用法线边界交叉法将这个多目标优化问题转换为一系列单目标优化问题,并采用非线性原对偶内点法求解。在迭代过程中,按照场景顺序将简化修正方程的系数矩阵排列为对角加边形式,方便对其实施解耦,并运用异步块迭代法求解,从而将一组高维修正方程组的求解转化为若干个分别与预测场景和误差场景相对应的低维修正方程组的求解。在某省级50机系统上的计算结果表明,场景解耦方法可大幅度降低内存需求且适于求解大系统、多场景、多目标动态经济调度问题。
A multi-objective stochastic dynamic economical dispatch model considering wind power integration is established to minimize both fuel consumption and power purchase cost.This model can be transformed into a large-scale multi-objective deterministic model based on the scenario method.The normal boundary intersection method is introduced to convert this multiobjective problem into a series of single objective optimization problems,which can be solved by means of the nonlinear primaldual interior point method.During the iteration process,the coefficient matrix of the reduced correction equation set can be rearranged as a block bordered diagonal structure according to the sequence of scenarios.Thus this reduced correction equation set is conveniently decomposed into several low-dimensional correction equation sets corresponding to forecast scenario and error scenarios,respectively to be solved by the asynchronous block iteration method.Furthermore,the computational results concerning aprovincial power system with 50generators show that the proposed scenario decomposition method is able to greatly reduce the memory demand while suitable for solving the dynamic economical dispatch problem of large-scale power systems involving both multiple scenarios and multiple objectives.
出处
《电力系统自动化》
EI
CSCD
北大核心
2014年第9期34-40,共7页
Automation of Electric Power Systems
基金
国家重点基础研究发展计划(973计划)资助项目(2013CB228205)~~
关键词
风电场
随机动态经济调度
多目标
场景法
非线性原对偶内点法
对角加边结构
场景解耦
异步块迭代法
wind farm
stochastic dynamic economical dispatch
multiple objectives
scenario method
nonlinear primal-dual interior-point method
block bordered diagonal structure
scenario decomposition
asynchronous block iteration method
