A Physical-Based Observer for Vehicle State Estimation and Road Condition Monitoring

M Terzo,B Lenzo,G Ottomano,S Strano,C Tordela

doi:10.1088/1757-899x/922/1/012005

Abstract

The performance of the vehicle’s active safety systems depends on accurate knowledge of the vehicle state, and the frictional forces resulting from tyre contact and the road surface. This paper aims to estimate the vehicle states and tyre-road coefficient of friction through and Extended Kalman Filter (EKF), integrated with the Double-Track model and the Pacejka Magic Formula that allows knowledge of the lateral force of the tyre. Besides, this approach can estimate the overall coefficient of lateral friction on each side of the vehicle, left and right respectively. Simulations based on a reference vehicle model are performed on different road surfaces and driving manoeuvres to verify the effectiveness of the proposed estimation method, in order to obtain good performance from different vehicle control systems.

Highlights

Over the past decades driver assistance system have become a standard in automotive industry [1]
The performance of the vehicle's active safety systems depends on accurate knowledge of the vehicle state, and the frictional forces resulting from tyre contact and the road surface
This paper aims to estimate the vehicle states and tyre-road coefficient of friction through and Extended Kalman Filter (EKF), integrated with the Double-Track model and the Pacejka Magic Formula that allows knowledge of the lateral force of the tyre

Summary

Introduction

Over the past decades driver assistance system have become a standard in automotive industry [1]. In order to try to reduce this very high value, it is possible to try to improve the performance of these driver assistance systems The performance of these systems could be further improved if more accurate knowledge on the vehicle state, inputs and parameters would be available. Many of these variables, as the sideslip angle, cannot be measured directly in commercial vehicles because the sensors are very expensive. The first approach uses a kinematic vehicle model, independent from tyre parameters and road condition, in combination with measurement from standard vehicle sensor. This estimation technique is sensitive to sensor errors (noise and bias).

Objectives

Results

Conclusion