Open Access

ARTICLE

Research on Performance Improvement of Gas Boiler Coupled with Solar Heating System Based on Artificial Neural Network

Dong Liu¹, Xinyu Li¹, Yuanqiang Zhao¹, Jinhuan Liu¹, Xiangfei Kong^2,*
1 Beijing District Heating Group Co., Ltd., Beijing, China
2 School of Energy and Environmental Engineering, Hebei University of Technology, Tianjin, China
* Corresponding Author: Xiangfei Kong. Email:
(This article belongs to the Special Issue: Integrated Renewable Energy Systems for Heating, Cooling, Power Generation and Energy Management)

Energy Engineering https://doi.org/10.32604/ee.2026.077639

Received 14 December 2025; Accepted 27 January 2026; Published online 26 February 2026

Abstract

Despite the significant carbon dioxide emissions associated with the combustion of gas-fired boilers, they remain widely used in modern heating systems. The integration of gas-fired boilers with solar thermal utilization systems enables the synergistic use of traditional and sustainable energy sources, offering an effective pathway for energy conservation and carbon reduction in the heating sector. This study innovatively proposes an advanced predictive control strategy that combines mass flow regulation with artificial neural network modeling. This approach allows for real-time hourly control of the system’s thermal output, effectively addressing the limitations of traditional control strategies that struggle to adapt to the variability of solar energy and the dynamic changes in building thermal loads. The heating system is modeled using TRNSYS and predictive control is developed using MATLAB. The results demonstrated a remarkable 36.06% reduction in energy consumption when comparing the gas-fired boiler coupled with clean energy heating system (GCCH) against conventional gas-fired boiler heating (GH) systems configurations. Comparative analysis revealed that the predictive control strategy achieved significant performance enhancements of GCCH: a 24.03% reduction in energy consumption, a 23.58% improvement in the coefficient of performance, and an increased solar contribution ratio reaching 43.61%. The system maintained optimal thermal comfort conditions with an average indoor temperature of approximately 21.1°C. By accurately predicting hourly outdoor temperatures and building thermal loads, the proposed strategy achieves dynamic matching between energy supply and demand. This not only minimizes energy waste to the greatest extent but also provides a novel technical solution for the intelligent optimization of heating systems that integrate gas-fired boilers with renewable energy sources.

---

Keywords

Building heating system; solar energy; building heat; predictive regulation; artificial neural network

---