A Deep Q-Learning Based Optimization Method of Energy Management for Peak Load Control of Residential Areas with Cchp Systems

Wu Jiekang,Jin Feng,Tang Huiling,Chen Lingmin,Wang Yanan
DOI: https://doi.org/10.2139/ssrn.4034068
2022-01-01
SSRN Electronic Journal
Abstract:Aiming at the coastal residential area, considering the natural characteristics of wind and photovoltaic energy, the supply characteristics of gas and electric energy, the energy consumption law of cold, heat and electricity and the complementary and coupling relationship of various energy sources, this paper studies the formation conditions and various scenarios of peak load in the residential area with CCHP systems. Using the regulation characteristics of controllable loads such as electric vehicle, air conditioner, water heater and refrigerator, considering the coordination of cooling, heating and electric energy storage, the interchangeability and controllability of air conditioner based on electric/gas energy, and the transferability, interchangeability and controllability of water heater based on electric/gas/soalr/heat pump energy, A collaborative energy management model for coastal resident area with electricity, gas, wind and solar energy is presented for peak load adaptive control of cooling, heating and electric energy. The deep Q-learning algorithm is used to solve the energy management optimization problem for peak load control in coastal residential area, and good results are obtained, such as effectively reducing load fluctuation and improving the balance relationship of multiple energy sources.
What problem does this paper attempt to address?