The desiccant wheel-assisted atmospheric water harvesting system (DW-AWHS) effectively mitigates the performance degradation of a heat pump-based atmospheric water harvesting system (HP-AWHS) in arid desert climates by elevating the air dew point temperature. Comparative multi-environmental performance experiments between the HP-AWHS and DW-AWHS revealed that the DW-AWHS exhibits reduced susceptibility to environmental fluctuations. The water harvesting rate (WHR) and dew point temperature rise (Delta T-DP) are the core metrics for evaluating the DW-AWHS performance. However, an inherent trade-off exists between these two objectives during system optimization. This study proposes a hybrid optimization framework integrating response surface methodology (RSM) and the non-dominated sorting genetic algorithm-II (NSGA-II) to resolve this multi-objective optimization challenge. First, four critical design variables were identified: treated air fan speed (n(tr)), regeneration air fan speed (n(re)), condensation temperature (T-co), and regeneration power (W-he). A Box-Behnken experimental design was implemented to construct regression models. Analysis of variance (ANOVA) validated the adequacy and reliability of the regression models. Response surface analysis elucidated the interactive effects between paired design parameters. Subsequently, the Pareto optimal frontier was derived using NSGA-II. Sensitivity analysis demonstrated that WHR is predominantly influenced in the order W-he > T-co > n(re) > n(tr), while Delta T-DP is governed by T-co > W-he > n(re) > n(tr). Under optimized conditions (20 degrees C, 40 % RH), the system achieved a WHR of 0.2164 kg/h and a Delta T-DP of 23.72 degrees C, corresponding to operational parameters of n(tr) = 3.43 krpm, n(re) = 2.99 krpm, T-co = 17.89 degrees C, and W-he = 0.993 kW. Experimental validation confirmed the prediction accuracy, with deviations between simulated and measured results below 3 %, thereby substantiating the robustness of the proposed methodology.