Exergy-based optimization of heating system performance under variable operating conditions

To address the low efficiency caused by unstable operation of heating systems under variable operating conditions, this paper proposes an innovative dynamic optimization framework based on exergy analysis, filling a research gap in quantifying energy quality loss and enabling collaborative optimization across multiple operating conditions. A novel scientific partitioning method for variable operating conditions is developed using SCADA data combined with k-means clustering and silhouette coefficients, overcoming the limitations of traditional steady-state analyses. A refined flow model is constructed that integrates dynamic constraints of variable operating conditions and accurately quantifies loss distribution across different scenarios. Furthermore, a multi-parameter collaborative optimization strategy driven by a genetic algorithm is designed to target and reduce key loss mechanisms. Experimental verification shows that this method achieves an average increase of 6.6% in the exergy index. Under light-load and high-load conditions, the system’s exergy efficiency improves by 6.1% and 5.2%, respectively, providing both a theoretical breakthrough and a new engineering paradigm for efficient energy utilization under dynamic operating conditions.