Kinematics-based Fault-tolerant Techniques: Lane Prediction for an Autonomous Lane Keeping System

Abstract

In this paper, we propose the use of fault-tolerant techniques for an autonomous lane keeping system (LKS) under sensor failure of the camera vision sensor. When the output of the vision sensor is not available to the LKS due to malfunction and/or environmental conditions, it is necessary for the lateral control system to maintain its stability before the driver takes over control authority. We propose a method for fault-tolerant control using the lateral kinematic vehicle motion model. The kinematic motion model-based lane estimation scheme covers possible camera vision sensor failure that occurs in the presence of unreliable or unavailable data from the vision sensor due to complex shadowing, incomplete lane marks, and lighting changes. The proposed lane estimation method enables the LKS to maintain its performance in the presence of sensor failures. The developed algorithm was validated via computational simulation results with CarSim and MATLAB/Simulink. We also included experimental results with a test vehicle equipped with an AutoBox from dSPACE.