INTEGRATION OF MOTION PLANNING AND MODEL-PREDICTIVE-CONTROL-BASED CONTROL SYSTEM FOR AUTONOMOUS ELECTRIC VEHICLES

Guodong Yin,Xianjian Jin,Chentong Bian,Nan Chen,Jianghu Li

doi:10.3846/16484142.2015.1089322

Abstract

This paper introduces the development of an autonomous driving system in autonomous electric vehicles, which consists of a simplified motion-planning program and a Model-Predictive-Control-Based (MPC-based) control system. The motion-planning system is based on polynomial parameterization, which computes a path toward the expected longitudinal and lateral positions within required time interval in real scenarios. Then the MPC-based control system cooperates the front steering and individual wheel torques to track the planned trajectories, while fulfilling the physical constraints of actuators. The proposed system is evaluated through simulation, using a seven-degrees-offreedom vehicle model with a ‘magic formula’ tire model. The simulations and validation through CarSim show that the proposed planner algorithm and controller are feasible and can achieve requirements of autonomous driving in normal scenarios.

Highlights

Autonomous driving has become a fast-developing and promising area in recent years (Urmson et al 2008; Levinson et al 2011)
This paper introduces the development of an autonomous driving system in autonomous electric vehicles, which consists of a simplified motion-planning program and a Model-Predictive-Control-Based (MPC-based) control system
The motion-planning system is based on polynomial parameterization, which computes a path toward the expected longitudinal and lateral positions within required time interval in real scenarios

Summary

Introduction

Autonomous driving has become a fast-developing and promising area in recent years (Urmson et al 2008; Levinson et al 2011). An autonomous vehicle is required to perform two important tasks coherently: a motionplanning program computes the desired trajectories in various scenarios and a control system manipulates actuators to track the planned path (Park et al 2009). This paper will present a simplified polynomial parameterization method for motion planning in real scenarios and make the best use of initial and terminal motion states of the autonomous vehicle. The initial state can be detected via various senrate together with motion-planning and control system sors and the terminal state can be estimated based on for four-wheel-drive autonomous electric vehicles. The tween motion-planning system and control system can planned path is defined by fifth-order polynomials that enhance the ability to deal with complex driving require- have six parameters given as: ments quickly and effectively in autonomous electric vehicles. The paper is closed with concluding remarks and ideas for future work

Motion Planning

Vehicle Control System

Vehicle Model

Simulation and Analysis

Conclusions