Keywords: motorsport, vehicle design, vehicle modelling, dynamics, evolutionary algorithms, genetic algorithms, optimisation.

In order to achieve optimum performance from a racing car, a racing team must balance a large number of variables. In order to achieve this, the team will draw on a combination of experience, testing results and simulation to evaluate the effectiveness of different vehicle set-up configurations. Parameters such as suspension set-up, weight distribution, engine tune, traction-launch control settings, tyre pressures, brake bias, gear ratios and shift-point can all be adjusted in order to improve vehicle performance. However, many of these variables are inter-linked since adjusting one variable, such as tyre pressure, will affect the optimal setting of another variable, such as brake bias. Achieving this balance between variables usually relies on the experience of team members and the extensive use of vehicle dynamics simulations.

This paper presents the results of a model developed to assist with the design and set-up of Cardiff University's Formula Student car. This car, designed and built by undergraduate students, competes in a series of time-trial events at the annual inter-university Formula Student competition. In order to assist inexperienced students to optimise the car design and set-up within a short period of time, a simulation of one of the time-trials (a straight line acceleration event over 75 metres) was developed. The simulation allows various car set-up parameters to be adjusted and the effect of these changes on event times to be calculated. It was tested against actual times for the acceleration event achieved at previous competitions and found to be in good agreement. This model was then coupled to a genetic algorithm to find high performance vehicle configurations. The genetic algorithm mimics Darwinian evolution to rapidly evolve successively more optimal solutions by sampling only a small percentage of the total possible solutions. Such algorithms allow near-optimal vehicle set-ups to be obtained without having to exhaustively test every possible combination of variables.

Whilst the simulation is, in itself, relatively simple, this work demonstrates the applicability and potential power of evolutionary computing when coupled with vehicle dynamics models. Such computational tools, whilst unlikely to completely replace practical experience and testing, are set to see increasing use as motorsport of all levels further embraces digital design and simulation methods.

Further work to extend the current simulation, together with the potential for linking algorithms to more comprehensive lap-time simulations, is also discussed in the paper.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £125 +P&P)