Pareto multi-objective optimum design of vehicle-suspension system under random road excitations

This work presents a new framework by integrating some recent methods in the fields of road and vehicle modeling and optimization for multi-objective optimization of a passive vehicle model capable of estimating vehicle performance in facing random road excitation. To achieve this purpose, simulation of an actual random road power spectral density is employed, and a five-degree of freedom half-car model, capable of approximating vehicle performance, is developed. Furthermore, with the aid of stochastic theory, criteria of vehicle performance, including ride comfort (acceleration of seat) and road holding (working space and vertical tyre velocity), are calculated in terms of root mean square. These criteria are applied as objective functions in multi-objective uniform-diversity genetic algorithm optimization of the vehicle model. Based on different performance criteria, several design points are chosen from Pareto front, and frequency responses of those designs are depicted. Comparison between results of current work and those reported in a single-objective optimization study delineates a considerable improvement in the performance of the vehicle-vibration model. It is concluded from the obtained results that the proposed framework enables designers to select proper primary designs as a basis of later design stages with regard to the priority of desired performance criteria.