基于集成学习的含电气热商业楼宇群的分时电价求解

Optimal Solution of Time-of-Use Price Based on Ensemble Learning for Electricity-Gas-Heat Commercial Building

当前,越来越多的含电气热商业楼宇配备有各类分布式发电设备,商业楼宇群的负荷构成越来越复杂。针对商业楼宇群分时电价的制定问题,该文在考虑了售电公司和商业楼宇群双方利益后,提出一种分时电价制定的双层优化模型。在该模型中,售电公司通过发布分时电价以及收集商业楼宇的用电信息,实现自身收益最大化;商业楼宇群根据售电公司发布的分时电价制定其考虑舒适度的日前用电计划,实现自身运营成本最低的目标。同时,结合强化学习,提出一种集成学习算法和单纯形法相结合的混合优化方法,实现模型的求解,通过算例结果验证其对降低商业楼宇运行成本和提高售电公司收益的有效性。

At present, the load of commercial buildings is becoming more and more complex. This paper proposed a bi-level optimization model for time-of-use prices, taking into account the interests of both electric retailer and commercial buildings. In this model, electric retailer maximizes his own revenue by publishing time-of-use prices and collecting electricity usage information for commercial buildings;commercial buildings make their own daily electricity usage plans based on the time-of-use prices and reduce their own operating costs. At the same time, a hybrid optimization algorithm based on ensemble learning(EL) was proposed in this paper to achieve the solution of the model. It was verified by example results that it can reduce the operating costs of commercial buildings and improve the interest of the electric retailer.