A Methodology for Identification of Magic Formula Tire Model Parameters from In-Vehicle Measurements

Abstract

Accurate tire modeling is of key importance to the development of modern vehicles. Traditional Flat-Trac testing of tires is expensive and time consuming. In order to increase the efficiency of the vehicle development process, a new method for identifying Magic Formula tire model parameters has been investigated. By driving specific maneuvers with a vehicle instrumented with sensors to measure wheel forces and angles, tire model parameters can be estimated by use of a global optimization algorithm. Full vehicle simulations were made in order to identify the test procedures necessary to provide a representative range of data for a tire. A tire model parameter fitting tool was developed and validated with Flat-Trac datasets. Physical testing was carried out to reproduce the simulated driving maneuvers on a car equipped with wheel force transducers, an inertial measurement device, and a high-speed camera system to measure wheel angles and displacements. Results of estimating the pure slip lateral tire model parameters from the in-vehicle measurements shows good correlation to the original tire model. Furthermore, the tire model parameters identified by this method more accurately represent the behavior of the tire on the test vehicle, without the need to modify scaling parameters. This indicates that this newly proposed method can produce accurate tire model parameters that require less tuning of scaling parameters to accurately represent vehicle behavior.