Optimization of Double Wishbone Suspension: Evaluating the Performance of Classes of Algorithms

Double wishbone suspension systems are employed in high-performance vehicles owing to their superior control over wheel motion. The design of a double-wishbone suspension is a challenging multi-objective problem due to the many inter- related quantities that are key in determining its performance and vary during its motion. The design activity is thus a time-consuming and complex iterative process that aims to tune their variation and minimize deviations from the desired performance characteristics while requiring the satisfaction of mechanical constraints. Computational optimization methods can be integrated into the design activity only after an assessment of their capability to support efficiently and effectively an iterative workflow. In this work, we start with the mathematical modeling of the camber angle of the double wishbone suspension. Then, we use the multi-model optimization approach that includes metaheuristic algorithms (Genetic Algorithms, Particle Swarm Optimization, Gradient Descent, and Ant Colony Optimization) and classical methods (Nelder-Mead, Interior Point, Simulated Annealing, and Random Search). These methods are assessed in terms of objective value, computational time, and memory usage in an effort to minimize the variation of camber angles from the specified constraints. The results prove that the Interior Point method is more accurate and efficient in optimizing the suspension's camber angle across its operational range, followed by other classical methods.