Parameter Identification Method for Air Conditioning Load Models Based on Improved Harris Hawk Optimization Algorithm

The air conditioning load holds a significant proportion in demand-side resources, and the air conditioning load model serves as the foundation for air conditioning load regulation. However, it overly relies on the accuracy of model parameters. This paper addresses this issue by proposing a parameter identification method for the air conditioning load model based on the Improved Harris's Hawks Optimization (IHHO) algorithm to enhance parameter identification accuracy. The Harris's Hawks Optimization (HHO) algorithm simulates the cooperative predation behavior of Harris's hawks in nature and is effective in solving problems in parameter space but is prone to local optima. To tackle this, an Improved Harris's Hawks Optimization (IHHO) algorithm is introduced for parameter identification. Initially, a Latin Hypercube Sampling method is employed for population initialization to achieve a more uniform initial population distribution. Subsequently, algorithm performance is enhanced, and the issue of optimizing local optimal solutions is addressed through a nonlinear energy updating strategy and a Cauchy-Gaussian mutation perturbation strategy. Finally, the IHHO algorithm is combined with a second-order equivalent thermal parameter (ETP) model to minimize the error between simulated and measured temperatures of the second-order equivalent thermal parameter model as the optimization objective. Experimental results indicate that the fitness values of parameters obtained through IHHO optimization for the second-order ETP model are optimal in various experimental scenarios compared to those obtained through HHO and PSO optimizations. The simulated temperatures based on IHHO optimization are also closer to the actual temperatures. In conclusion, the proposed air conditioning load model parameter identification based on the IHHO algorithm exhibits improved accuracy, satisfying the requirements of load regulation.