基于居民用户调控潜力精细评估的台区侧灵活资源多目标日前优化调度

Day-ahead optimal scheduling of flexibility resources for distribution area based on the fine evaluation of residential users'regulatory control potential

以空调和电热水器为代表的居民温控负荷和电动汽车是台区用户侧重要的灵活可调资源,具有较大的调控潜力。由于居民负荷的参数不尽相同,配电调度中心难以直接准确评估其调控响应潜力,导致调度计划与实际调度要求存在较大差异。针对上述问题,构建了台区侧灵活资源协同调度框架,并建立各温控负荷和电动汽车聚合模型;提出了基于物理特征量分类聚合的灵活资源调控潜力精细评估方法;在此基础上,以电网削峰需求偏差最小和电网收益最大为目标,兼顾调峰目标和光伏出力的不确定性,建立了台区侧温控负荷和电动汽车灵活资源模糊机会约束多目标日前优化调度模型,并通过模糊约束条件清晰等价类转换实现模型求解。仿真结果表明该方法通过选择合适的权重系数可以实现电网调峰需求和经济利益的平衡,模糊机会约束规划置信水平取0.90时结果最优。

Residents'temperature control load represented by air conditioners and electric water heaters and electric vehicles are important flexible adjustable resources on the user side of the station area,and they have great potential for regulation and control.Due to the different parameters of the residents'load,the distribution dispatch center has the difficulty to assess the potential for regulatory and control directly and accurately,resulting in a large difference between the dispatch plan and the actual dispatch requirements.In response to the above problems,the paper builds a platform-side flexible resource coordination scheduling framework;aggregate models of various temperature control loads and electric vehicles are established.A fine evaluation method of flexible resource regulation potential based on classification and aggregation of physical characteristics is proposed.Based on this,the minimum deviation of grid peak shaving demand and the maximum grid revenue are considered as targets,taking into account the uncertainties of peak shaving goals and photovoltaic output,a multi-objective day-ahead optimal scheduling model was established for the temperature control load on the platform side and the fuzzy opportunity constraint of flexible resources of electric vehicles.And the model is solved through the conversion of fuzzy constraint conditions to clear equivalence classes.The simulation results show that the method in this paper can achieve the balance of power grid peak shaving demand and economic benefits by selecting appropriate weight coefficients,and the result is optimal when the confidence level of fuzzy chance constrained programming is 0.90.