Soft‐computing techniques for cruise controller tuning for an off‐road electric vehicle

Abstract

Speed controllers may be employed to provide safer along with more secure vehicles. They may also be used to minimise environmental pollution, e.g. speed controllers can be employed to track speed optimal velocity profile based on energy consumption minimisation. Consequently, accurate speed tracking is important. However, despite soft-computing techniques have been proved successful in controller tuning, there is a limited amount of research on these techniques applied to speed controller optimisation. Therefore, this study performs a comparison study on PI cruise controller tuning for an off-road electric vehicle. A cost function is designed to reach an accurate EV speed tracking while considering safety aspects, such as no reverse speed. The ACO-NM algorithm has been demonstrated to be the most efficient compared to GA, ALO, DE, and PSO. Indeed, ACO-NM reached high-quality solutions for lower computational cost for three driving cycles. Moreover, contrary to the majority of published work on the subject, experimental validations have been carried out with the optimised PI cruise controllers. The experimental results have validated the ACO-NM efficiency with a maximum overshoot average <10% for the hardest acceleration of the real driving cycle.