This paper presents the fault detection and diagnosis(FDD) algorithm to enhance reliability of a longitudinal controller for an autonomous All-Terrain Vehicle(ATV). The FDD is designed to monitor and identify faults which may occur in distributed hardware used for longitudinal control, e.g., DSPs, CAN, sensors, and actuators. The proposed FDD is an integrated approach of decentralized and centralized FDD. While the former is processed in a DSP and suitable to detect faults in a single hardware, it is sensitive to noise and disturbance. On the other hand, the latter is performed via communication and it detects and diagnoses faults through analyzing concurrent performances of multiple hardware modules, but it is limited to isolate faults specifically in terms of components in the single hardware. To compensate for disadvantages of each FDD approach, two layered structure including both decentralized and centralized FDD is proposed and it allows us to make more robust fault detection and more specific fault isolation. The effectiveness of the proposed method will be validated experimentally.
[1]
Rolf Isermann,et al.
Application of model-based fault detection to a brushless DC motor
,
2000,
IEEE Trans. Ind. Electron..
[2]
Masayoshi Tomizuka,et al.
A complete fault diagnostic system for automated vehicles operating in a platoon
,
2001,
IEEE Trans. Control. Syst. Technol..
[3]
Steven X. Ding,et al.
Model Based Diagnosis of Sensor Faults for ESP-Systems
,
2000
.
[4]
J. Karl Hedrick,et al.
Robust stabilization and ultimate boundedness of dynamic surface control systems via convex optimization
,
2002
.
[5]
J. K. Hedrick,et al.
Fault tolerant nonlinear control with applications to an automated transit bus
,
2005
.