求解多目标随机动态经济调度问题的场景解耦方法

发布时间：2018-10-20 07:26

【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。
[Abstract]:For the power system with wind power access, a multi-objective stochastic dynamic economic scheduling model with minimum total fuel consumption and power purchase cost is established. The model is transformed into a large-scale multi-objective deterministic dynamic economic scheduling model with the help of scene method. The multi-objective optimization problem is transformed into a series of single-objective optimization problems by normal boundary crossing method, and the nonlinear primal dual interior point method is used to solve the problem. In the iterative process, the coefficient matrix of the simplified modified equation is arranged into diagonal and edge-added form according to the scene sequence, which is easily decoupled, and solved by the asynchronous block iterative method. Thus, the solution of a set of high-dimensional modified equations is transformed into the solution of several low-maintenance positive equations corresponding to the prediction scenario and the error scene, respectively. The calculation results on a provincial 50 computer system show that the scenario decoupling method can greatly reduce the memory requirement and is suitable for solving dynamic economic scheduling problems of large system, multi-scene and multi-objective.
【作者单位】：华南理工大学电力学院;
【基金】：国家重点基础研究发展计划(973计划)资助项目(2013CB228205)~~
【分类号】：TM614

【参考文献】