Sensor anomaly detection and recovery in a nonlinear autonomous ground vehicle model

Publication Type:
Conference Proceeding
Citation:
2017 Asian Control Conference, ASCC 2017, 2018, 2018-January pp. 430 - 435
Issue Date:
2018-02-07
Metrics:
Full metadata record
Files in This Item:
Filename Description Size
08287208.pdfPublished version517.85 kB
Adobe PDF
© 2017 IEEE. This work considers the problem of sensor anomaly detection, estimation and recovery for an autonomous vehicle model (e.g. a self-driving car). The proposed algorithm employs particle filtering and maximum likelihood methods to detect and estimate the anomaly. The estimated anomaly is used to correct the sensor readings. The system is nonlinear and it is assumed that both the system model and sensor outputs are corrupted by noise (not necessarily Gaussian). Simulation results are presented highlighting the performance of the proposed method.
Please use this identifier to cite or link to this item: