| This thesis utilizes a new integrated diagnostic process that seamlessly employs a graph-based dependency model and mathematical models. This integrated process is also implemented in a Hardware-in-the-Loop (HIL) platform. The diagnostic techniques are developed using the MATLAB/Simulink off-line simulation environment. The integrated diagnostic process and the development platform enable test designers to design tests, evaluate their diagnostic performance, generate executable code, and validate the diagnostic strategies in a seamless way prior to implementation on vehicles. The focus of our application is on automotive diagnosis. However, the solution is, generic, and has the potential for application to a wide range of aerospace and industrial systems.; The proposed thesis consist of five major contributions to the theoretical extensions and applications of the integrated diagnostic process: investigating graphical models for diagnostic knowledge representation and inference; applying integrated model-based and data-driven diagnostic strategies to an anti-lock brake system (ABS); using model-based prognostic techniques for an automotive suspension system; implementing the integrated diagnostic process by way of application to a hardware-in-the-loop (HIL) automotive engine control system; and developing distributed diagnosis algorithm for networked, embedded automotive systems. |
