A method based on constraint multi-objective evolutionary algorithms approach is proposed to optimise the handling performance of a race car dynamic. This involves a software-in-the-loop optimisation (SiLO) between a high-fidelity model of a formula vehicle modelled in VI-Grade and MATLAB for hosting the optimisation algorithms. The optimisation process involves the tuning of the suspension design variables, i.e. spring stiffness, anti-roll bar stiffness, damper coefficient, toe angle, and camber angle for both front and rear suspensions against eight objectives functions, i.e. sideslip overshoot, yaw rate overshoot, linear understeer gradient, limit understeer gradient, maximum lateral acceleration, yaw rate phase delay and lateral acceleration phase delay. A custom constraint function was introduced to improve the convergence speed of the optimisation process. The optimised result showed a promising increase in handling performances, i.e. the maximum achievable lateral acceleration has increased by 8.9% whilst maintaining the stability of the vehicle close to neutral steer behaviour. The sideslip angle was well controlled under 1 deg at 95% of maximum lateral acceleration, and slip angle overshoot has improved by 1.65% in step steer manoeuvre. Additionally, the skid-pad events and autocross event laps time were reduced by 5.19% and 1.71%, respectively.
Handling performance optimisation for formula vehicle using multi-objectives evolutionary algorithms
Vehicle System Dynamics ; 58 , 12 ; 1823-1838
01.12.2020
16 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Unbekannt
Helper-Objectives: Using Multi-Objective Evolutionary Algorithms for Single-Objective Optimisation
British Library Online Contents | 2004
|Robust Multi-Objective AeroStructural Optimisation Using Advanced Evolutionary Algorithms
British Library Conference Proceedings | 2009
|