A variable importance guided optimization framework based on whale optimization algorithm for energy-efficient HVAC operation

To reduce energy consumption of HVAC systems, this paper proposes a Variable Importance Guided Optimization Framework for optimization algorithm of HVAC operation. By dynamically quantifying variable importance during the optimization process, an adaptive strategy adjusts operator selection probabilities and variable update behaviors, thereby allocating search resources more effectively. By incorporating the proposed mechanisms into the Whale Optimization Algorithm (WOA), Variable Importance Guided Whale Optimization Algorithm (VIG-WOA) is developed and evaluated using operational data collected from an actual HVAC system. Results show that the overall energy-saving rate reaches 22.43%. Compared with the conventional WOA, the optimized system coefficient of performance (COP) under two typical load conditions is improved by 9.48% and 5.45%, respectively, demonstrating superior convergence accuracy and robustness. The proposed framework provides a potential approach for incorporating variable-importance information into HVAC energy optimization and demonstrates its effectiveness in the investigated HVAC case.