Fault detection and handling for sensor and communication faults of an automatic valet parking system

In this paper, a fault detection and handling method for automatic valet parking in the presence of faults in sensors and packet loss and time delay of communication is presented. Two possible fault scenarios based on the analysis of field test data are considered, i.e., faults of position and heading angle measurements and communication. The proposed FDD method is based on multiple nonlinear observers to estimate the position and heading angle with consideration of packet loss via V2I communication, and a double-threshold-statistical-testing technique to reduce a false alarm. Finally, the performance of the proposed algorithm will be validated through the field test data including the faults.