Abstract: To address the challenges associated with the three-dimensional flow design requirements of hydrodynamic retarders, specifically the low braking performance and limited geometric adaptability of conventional straight blade configurations. This study proposes a novel three-dimensional geometric modeling methodology for cambered blade retarders based on the class-shape-transformation (CST) approach. In the proposed method, the toroidal circulation path, blade camber line, and thickness distribution are all parameterized using CST curves, enabling a highly flexible and precise design representation of the blade geometry. To verify the reliability and accuracy of the computational approach, a comparative analysis was conducted between experimental braking performance data and numerical simulations performed using computational fluid dynamics (CFD) for a baseline straight blade retarder model. The results revealed a maximum deviation of 4.92% and a minimum deviation of 4.43% between the simulated and experimental values, both of which fall below the generally accepted threshold of 8%. These findings confirm the validity of the CFD model and establish a robust foundation for subsequent design and optimization procedures. Building upon this validated simulation framework, the study proceeds to reconstruct the original straight blade retarder using CST-based parameterization to generate an improved cambered blade retarder design. Comparative CFD analyses between the straight blade and cambered blade configurations were conducted, and the results demonstrate that the cambered blade retarder not only meets feasibility requirements from a structural standpoint but also outperforms the conventional straight blade variant in terms of braking performance. This validates both the structural viability and performance advantages of cambered blade implementations in retarder applications. In pursuit of further improvements in braking torque and reductions in idling loss power, the study employs a design of experiments (DOE) methodology to systematically explore the influence of key blade cascade parameters on retarder performance. A response surface model (RSM) was developed to quantitatively describe the relationship between the cambered blade cascade parameters and two critical performance metrics: braking torque and idling loss power. A main effects analysis was then conducted on the RSM to identify the individual contributions of each design parameter. The analysis revealed that the peak height of the blade camber line has a significant positive impact on retarder performance, enhancing braking torque while mitigating idling loss power. Conversely, the blade deflection angle, incidence, and thickness factor were all found to exhibit negative correlations with overall retarder efficiency, indicating that these variables must be carefully managed during the design process to avoid performance degradation. To determine the optimal combination of blade parameters that can achieve a balanced improvement in both performance objectives, a multi-objective evolutionary optimization algorithm-non-dominated sorting genetic algorithm II (NSGA-II) was utilized. The NSGA-II optimization process generated a set of optimal solutions, from which a final optimized blade configuration was selected based on design priorities. A comprehensive evaluation of the optimized cambered blade retarder was conducted by comparing its external performance characteristics and internal flow field characteristics against those of the pre-optimization design. The results show that the braking performance of the optimized blade retarder is 23.5% higher than that of the cambered blade retarder, and the idling loss power is reduced by 30.9 %. In conclusion, this study presents a robust CST-based methodology for the geometric modeling and performance optimization of 3D cambered blade hydrodynamic retarders. By integrating advanced parameterization techniques, validated CFD analysis, and evolutionary multi-objective optimization algorithms, the proposed approach offers a comprehensive framework for the design of high-performance retarders. Beyond the immediate application to hydrodynamic braking systems, the modeling and optimization techniques developed herein provide a valuable reference for the broader field of turbomachinery design, offering insights and tools that can be extended to the efficiency improvement of turbines and other rotating fluid machinery.