制冷技术2025，Vol.45Issue(1)：61-67,7.DOI:10.3969/j.issn.2095-4468.2025.01.203
基于强化学习的直膨空调系统温湿度同时控制方法
Reinforcement Learning Based Method for Simultaneous Control of Temperature and Humidity in Direct Expansion Air Conditioning System
闫玲玲 1李钊 1沈晨彬1
作者信息
- 1. 上海理工大学环境与建筑学院,上海 200093
- 折叠
摘要
Abstract
In order to solve the problems of difficulty in constructing predictive models and weak adaptability in the simultaneous control of temperature and humidity in variable frequency direct expansion air conditioning systems,a reinforcement learning based method for simultaneous temperature and humidity control of direct expansion air conditioning systems is proposed.The direct expansion air conditioning system is taken as the controlled object,and the Q-learning algorithm is used to train the intelligent agent.By building a variable speed direct expansion system air conditioning room model,intelligent agent training and simulation experiments are conducted to test the control stability of the system.The results show that the method can make the room temperature reach the set value quickly and keep it within a certain range of the set value in the subsequent long-term operation.In the experiments of step change of dry bulb temperature 20-25℃and wet bulb temperature 10-15℃,the time for the system to adapt to the set dry bulb temperature change is about 456 s,and the time for the system to adapt to the set wet bulb temperature change is about 673 s;in the test cycle of 8 000 s,the maximum dynamic deviation of the dry bulb temperature is 1.2%,and the average deviation is 0.24%;the maximum dynamic deviation of the wet bulb temperature is 7.6%,and the average deviation is 0.4%.关键词
变速直膨式空调系统/温湿度同时控制/强化学习/智能体训练/仿真Key words
Variable speed direct expansion air conditioning system/Simultaneous control of temperature and humidity/Reinforcement learning/Intelligences training/Simulation分类
通用工业技术引用本文复制引用
闫玲玲,李钊,沈晨彬..基于强化学习的直膨空调系统温湿度同时控制方法[J].制冷技术,2025,45(1):61-67,7.
