Silica gel desiccant wheel is the main dehumidification technology to realize humidity control of ambient air. Optimizing the wheel size and reducing the operating energy consumption are necessary trends in technology development. In this study, based on computational fluid dynamics (CFD), we first established and verified a three-dimensional dynamic simulation model of the dehumidification and regeneration process of the desiccant wheel. We conducted a dynamic simulation for the silica gel desiccant wheel used for the protection of bridge cables. We then analyzed and discussed the influence of wheel thickness, wind speed, and regeneration wind volumetric flow rate and temperature on the moisture removal capacity (MRC) and specific energy consumption (SEC) to optimize the design of the desiccant wheel. The results showed that the optimal thickness increased with the increase in wind speed. Compared with the original design, the optimized wheel improved the average MRC by approximately 10% while reducing the SEC by approximately 15%, demonstrating the effectiveness of geometry optimization. Notably, the SEC decreased with increasing regeneration temperature. Therefore, in practical design, the regeneration temperature should be selected in accordance with the dehumidification requirements and maintained as low as feasible to enhance the overall economic performance.