Surrogate Model based NSGA-II for Suspension Optimization for Two-Wheelers

Abstract Optimizing two-wheeler suspension systems for rider comfort traditionally requires computationally expensive dynamic simulations. This paper presents an efficient framework combining Kriging surrogate modeling with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to optimize suspension parameters on a two-wheeler. To bypass costly continuous MATLAB evaluations, a surrogate model is trained on a 4-DOF half-car model, mapping the non-linear relationships between front/rear stiffness, damping coefficients, and dynamic responses. NSGA-II is deployed across this surrogate space to search for parameters that minimize sprung mass vertical and rotational acceleration while satisfying constraints of vehicle stance, and suspension travel limits. The proposed surrogate-assisted methodology reduced computational time by 419% compared to traditional numerical integration while successfully identifying a Pareto-optimal front. The selected optimal configuration yielded a 16% improvement in rider comfort, providing a robust, time-efficient approach to fine-tuning two-wheeler suspensions.