摘要： Various metaheuristic optimization methods have been applied in high-performance glazing and shading system design due to the complexity and nonlinear impact of the fenestration characteristics on the daylighting performance. However, the optimal solutions found by different optimization methods may vary because of the stochastic nature and the configurations of the optimization methods. This paper studied the improvements in the reliability, consistency and robustness of the genetic algorithm (GA) using hybridization with simulated annealing (SA) considering different cooling strategies in the SA, including the initial temperature and state transitions for each temperature, which controls the comprehensiveness and the convergence acceleration of the search by SA. Analysis of the reliability, consistency and robustness of the optimization methods based on the mean values and the variances of the objective function values of the best cases found in different methods revealed that there is a significant difference between the hybrid GA/SA with higher temperature and GA, where hybrid algorithm performed better than the GA.