山东电力技术2025，Vol.52Issue(1)：54-61,8.DOI:10.20097/j.cnki.issn1007-9904.2025.01.006
基于深度强化学习的空气源热泵供热系统温度控制策略
Temperature Control Strategy for Air Source Heat Pump Heating System Based on Deep Reinforcement Learning
摘要
Abstract
The air source heat pumps(ASHP)exhibits good adjustability,and the accuracy of its modeling and the design of control strategies are key to fully exploiting its regulation potential.This paper considers the thermal storage characteristics of air source heat pump heating systems and proposes a temperature control strategy based on deep reinforcement learning(RL)for ASHP heating systems.First,a mathematical model of the ASHP heating system is established based on parameter identification.Then,a Markov decision process(MDP)model for the ASHP heating system is developed,and a temperature control strategy based on deep reinforcement learning is designed using the Q-learning algorithm.Simulation results based on real operating data demonstrate that the proposed heating system mathematical model,which accounts for heating delays,can accurately predict the variations in supply and return water temperatures as well as indoor temperatures.Furthermore,the proposed deep reinforcement learning-based temperature control strategy effectively reduces electricity costs while maintaining the indoor temperature at the set value.关键词
空气源热泵/分时电价/强化学习/水温控制策略Key words
air source heat pump/time of use pricing/reinforcement learning/water temperature control strategy分类
信息技术与安全科学引用本文复制引用
刘伟,高嵩,宋宗勋,许晓康,刘萌..基于深度强化学习的空气源热泵供热系统温度控制策略[J].山东电力技术,2025,52(1):54-61,8.基金项目
国网山东省电力公司科技项目(520613220004).Science and Technology Project of State Grid Shandong Electric Power Company(520613220004). （520613220004）
