Article Details

Abstract

Aiming at the problem that traditional hospital energy management modes find it difficult to balance energy saving and reliability, this paper constructs a new type of hospital energy management platform based on Internet of Things (IoT) edge-cloud collaboration. The platform adopts an improved Particle Swarm Optimization (PSO) algorithm and Reinforcement Learning (RL) algorithm, collects and processes data in partitions through edge-cloud collaboration, and realizes precise control by combining the two-layer algorithm of upper-layer weight allocation and lower-layer regulation. Tests show that after 1 000 iterations, the loss value of the platform is as low as 0.17, and the 7-day reliability index of key areas stabilizes between 0.98 and 1.0. Simulation results show that the total electricity consumption decreases by 14.8% and the heat consumption decreases by 13.8%, showing significant energy saving effects. In a 60-day long-term deployment in an actual hospital, the platform reduces the total electricity consumption and total heat consumption of the inpatient department by 11.8% and 11.5% respectively, and improves the temperature compliance rate of key areas to 99.6%, which verifies its energy saving and high reliability in the real environment. This study provides technical support and reference for the intelligent control of hospital energy and energy saving practice of public buildings.

Keywords

IoT; PSO; RL; energy management information platform; energy saving effect

References

1. [1] Zhang Tian, Gao Jianwei, Liu Haoyu, etc The integration and application of information and energy technology in the comprehensive energy system of computing power [J]. Electric Power Construction, 2025, 46 (2): 26-34
2. [2] Zhang Yan Analysis of several energy-saving design measures for building HVAC systems [J]. Shanxi Architecture, 2016, 42 (19): 183-184
3. [3] Zhong Yefeng, He Wei The Impact of Energy Storage Technology Integrating Grid Balance Regulation and Energy Management on Industrial Economic Development [J]. Modern Industrial Economy and Informatization, 2025, 15 (1): 145-147
4. [4] Wang Xinyi, Fei Zesong, Monday Qing, etc Integration of Sensing, Computing, and Intelligence for the Internet of Things: Key Technologies and Future Prospects [J]. Journal of Electronics and Information Technology, 2025, 47 (4): 888-908
5. [5] Kong Xiangqiang, Yue Zhenwei, Yin Pengyu, etc Optimization of microchannel structure for heat collection/evaporator based on multi-objective algorithm [J]. Journal of Agricultural Engineering, 2023, 39 (8): 86-94
6. [6] Wen Muqi, Wen Wushao Overview of security challenges and solutions of edge computing [J]. Computer System Applications, 2024, 33 (11): 38-47
7. [7] LONG W C，SU R Q，WU H Y，et al. Design of DTS Training Simulation System Networking Based on LoRaWAN Protocol [J].International Journal of Modelling, Identification and Control，2024，45（2-3）：110-119.
8. [8] Li Xiaonian, Tan Fang, Qi Fei, etc WKNN localization algorithm based on median and sliding window fusion filtering [J]. Sensors and Microsystems, 2025, 44 (5): 142-145
9. [9] ZHOU J，LU R. Microgrid Cooperative Scheduling Optimization Based on Quality of Service Constraint and Deep Q-network Algorithm [J].International Journal of Low Carbon Technologies，2025，20：47-57.
10. [10] Han Baojun, Gao Qiang, Dai Fei, etc Research on Intelligent Frequency Control Strategy Based on Collaborative Reward Function Multi Objective Reinforcement Learning [J]. Journal of Electric Power Science and Technology, 2023, 38 (2): 18-29

About the authors

Yang Heng (1985.03-), male, Han nationality, from Anqing, Anhui, engineer, bachelor's degree, research direction: hospital logistics informatization.