Multisensor Fault-Tolerant Automotive Control

You are here: Homepage > Search

Multisensor Fault-Tolerant Automotive Control

Martinez, John J. / Varrier, Sebastien

This chapter deals with the problem of obtaining fault-tolerant guarantees of a multi-sensor switching strategy for automotive control. It is assumed that each sensor (or a family of sensors) has an associated observer that performs a good estimation under normal operation conditions. In presence of sensor failures the related observer provides an estimation that is biased by signals (that often depend of the references). Since the automotive vehicle is modeled as a linear parameter varying (LPV) system by taken the vehicle speed as a scheduling parameter, the main problem concerns the computation of robustly positively invariant-sets for the state trajectories of the controlled system during fault-free operation. These invariant-sets could be used as bounds and/or thresholds for the residuals (here the tracking error estimations for instance) allowing to detect a sensor failure even in presence of nominal disturbances. The invariant-sets provide a support for fast faultdetection avoiding selecting faulty-sensors. Then, these invariant-sets together with a sensor switching mechanism allows to obtain fault-tolerant guarantees of the controlled system. Here, the proposed approach is applied for a vehicle lateral dynamics control.