Ant colony optimization for solving the car side impact design optimization problem: a constraint-driven engineering study

Abstract

The Car Side Impact (CSI) design problem represents one of the most challenging benchmark cases in structural optimization due to its highly nonlinear objective function, multiple conflicting constraints, and strict safety requirements. While a wide range of metaheuristic algorithms have been applied to this problem, relatively limited attention has been devoted to the systematic adaptation of Ant Colony Optimization (ACO) for constrained continuous engineering design. This paper presents a comprehensive experimental assessment of an enhanced Ant Colony Optimization framework tailored for the CSI problem. The proposed approach incorporates continuous pheromone modeling, constraint-aware probabilistic sampling, adaptive evaporation mechanisms, and a dynamic penalty function to effectively balance exploration and exploitation in the constrained search space. Extensive numerical experiments demonstrate that the proposed ACO variant achieves competitive or superior performance compared with well-established algorithms in terms of weight minimization, constraint satisfaction robustness, and convergence stability. Statistical analysis across multiple independent runs confirms the reliability of the method. The findings validate ACO as a viable and competitive optimization strategy for complex automotive safety design problems and provide insights into its practical applicability in real-world engineering optimization.