Fault Detection for Lateral and Vertical Vehicle Dynamics

Abstract

Model based methods have been developed for sensor and process fault detection and diagnosis of an active hydraulic suspension system. Dynamic mathematical models are used for describing the lateral vehicle behavior and vertical behavior of an active suspension in real-time. Based on parameter estimation and parity equations symptoms for fault detection are generated. This is followed by fault diagnosis with fuzzy logic. Results are presented for measurements with a vehicle and a suspension test rig. The fault detection method is intended to be used for mechatronic suspensions, driver assistance, and vehicle control systems.