计及权系数影响的区域电网时段多目标调度优化策略

为了解决区域电网受风光机组出力的不确定因素影响的调度优化问题，提出了含风光机组在内的多电源多目标调度优化模型，目标函数中考虑了经济运行成本及环境治理成本，同时增加了备用容量和爬坡能力等约束。考虑权系数对多目标优化结果的影响，基于不同的负荷情况，对多目标优化的权系数进行动态选取，得到各阶段负荷下对应的权系数。该文将负荷分为基荷、腰荷及峰荷三种情况，根据模糊综合评价法对各种负荷情况动态确定权系数，以此提出四种优化方案。为了降低综合成本，在考虑了各种约束的前提下，利用改进的粒子群算法求解模型，求得四种方案下的分时段综合成本，并确定调度优化方案，同时得到运行周期内区域电网中机组出力的最佳策略，最后通过实例验证了所提基于动态权系数的多目标调度优化策略的有效性。

Multi-Objective Scheduling Optimization Strategy of Regional Power Grid Considering the Influence of Weighting Factors

The output of wind and solar power units is uncertain, which will affect the dispatching optimization of regional power grid. To solve this problem, a multi-source multi-objective scheduling optimization model including wind and solar power units is proposed. The objective function takes into account the economic operation cost and environmental treatment cost, and at the same time, the constraints such as reserve capacity and climbing ability increase. Considering the influence of weighting factors on the multi-objective optimization results, based on different load conditions, the multi-objective optimization weighting factors are dynamically selected, and the corresponding weighting factors under different load stages are obtained. The load is divided into three situations: base load, waist load and peak load. According to the fuzzy comprehensive evaluation method, the weighting factors of each middle load situation are dynamically determined, and four optimization schemes are proposed. In order to reduce the comprehensive cost, under the premise of considering various constraints, an improved particle swarm optimization algorithm was used to solve the model, obtains the comprehensive cost of four schemes in different periods, determines the scheduling optimization scheme. At the same time, the optimal strategy for unit output in the regional power grid during the operation period was obtained. Finally, the effectiveness of the proposed multi-objective scheduling optimization strategy based on dynamic weighting factors was verified by an example.