Robust Observer Based Intermittent Forces Estimation for Driver Intervention Identification

Abstract

On the way to fully automated driving, Level 3 automated driving systems (ADS) seem to be a more realistic choice in the short term. Within operational design domain (ODD), although the driver is not required to monitor the vehicle all the time, it is expected that the driver is ready to intervene in the ADS whenever certain risks are encountered. Hence, timely identification of driver intervention in the vehicle lays the foundation for safe switching from the ADS to the driver. Although many researches have been conducted on driver torque estimation in the last decade, most of them are based on electric power steering (EPS) systems, therefore are not quite suitable for application in steer-by-wire (SBW) systems, which shall be the next generation’s steering systems. In view of this, the SBW system parameters with uncertainties are first identified through experiments. Then, a PID controller and a robust observer considering Coulomb friction are designed and analyzed for the SBW systems, so that the system is asymptotically stable and the driver torque can be identified irrespective of parameter uncertainties and system noises. Detailed experimental results on a hardware-in-the-loop test bench verify the effectiveness of the proposed method, which shall also be implementable in a real vehicle.

Citation

Recommended citation: C. Huang, L. Li, Y. Liu and L. Xiao, “Robust Observer Based Intermittent Forces Estimation for Driver Intervention Identification,” in IEEE Transactions on Vehicular Technology, vol. 69, no. 4, pp. 3628-3640, April 2020. (Paperurl)