Quantify and mimic the feedback through the steering wheel during specific driving conditions: Reproducing the steering feel in a conventional steering system in a Steer-by-Wire system

Abstract

The mobility industry is going through a transformation where mechanical-connected steering system is replaced with wired-connected steering system, called Steer-by-Wire. Steer-by-Wire steering feedback is set by a Electronic Control Unit. Based on this transformation Polestar’s dynamic department wants to create steering feedback in the coming Steer-by-Wire cars that mimic the feedback felt in a mechanical steering system during some specific conditions. This thesis aim is therefore to create software that creates feedback that mimics the mechanical system during these conditions: • Driving over crest while turning • Driving on a dashed road marking • Driving across rutted roads • Driving with front or rear axle saturated The method starts with logging the sensors data from the car which should be con verted to Steer-by-Wire when it is driving through the specific conditions to see how they could be detected. Then a car with a mechanically connected system is driven through the same conditions to see what the desired feedback should be. The result is a Simulink model consisting of several functions, a main feedback function that sets the basic feedback and additional functions that are connected to the specific conditions. The main feedback function has promising results. The functions that detect driving over crest while turning, dashed road marking and axle saturation show potential. These functions detect the situations and output the aimed feedback. For rutted road conditions, no solution for a function has been found. In addition to the functions detecting the situation there are two functions for friction coefficient estimation to enable change of feedback in the steering wheel with available friction. These functions only show reasonable values when close to limit handling or steady state.