Model-Based Fault Diagnosis of an Automated Manual Transmission Shifting Actuator

摘要

This paper presents a model-based methodology of residuals design for fault diagnosis of an Automated Manual Transmission (AMT) shifting actuator by employing Structural Analysis (SA). A group of sensors are suggested to obtain the maximal capability of Fault Detection and Isolation (FDI) after performing SA. Then, Minimal Structurally Over-determined (MSO) sets are identified to generate four residuals. To ensure stable and robust residuals, concepts from Analytical Redundant Relation (ARR) and observer-based parameter evaluation techniques are utilized. The proposed FDI scheme for AMT actuator has been successfully tested and verified using numerical simulations in MATLAB Simulink. The presented scheme offers a cost effective solution by using only two sensors to monitor five critical faults in AMT actuator.