Optimising vehicle handle characteristics using active chassis control subject to driver input limitations

Abstract

Consulta en la Biblioteca ETSI Industriales (Riunet)

[EN] It is well known that car manufacturers try to improve their vehicles performance each generation. That is why lap times are recorded in some circuits, such as Nürburgring. Most of the companies compete with the new models to achieve the best time for a lap on those circuits to highlight their vehicle performance when compared to the competitors. The focus of this Thesis is to improve vehicle performance by using front to rear torque vectoring. For that purpose, several car models, torque strategies and manoeuvres were analysed. Vehicles are complex systems (the main subsystem that can be found are: steering system, powertrain system and chassis as an example), for this reason, they need to be modelled for being analysed. The initial models used for the simulation were the bicycle model and a four wheel model. These models were created using Matlab. Moreover, the torque strategies selected went from absolute freedom of the torque to highly constrained torque values. With the three torque strategies selected, vehicle response was exanimated as well as the results variance between the torques configurations. Once all the parameters required for the manoeuvre were defined, the lap time optimisation was realised with GPOPS2. By using this optimisation software inside Matlab, the computer was able to establish the manoeuvre time for each case. After all the cases were simulated, the best torque strategy was able to be selected by looking at the results obtained.