This grant provides funding for the development of a test bed, Advanced Diagnostic Vehicle Agent Network System(ADVANS) for automotive engineering fault diagnosis. The ADVANS will be used to investigate the feasibility of an innovative engineering diagnostic model, Distributed Diagnostic AgentS(DDAS). The proposed DDAS model consists of a number of signal diagnostic agents(SDA) and a device diagnostic agent(DDA). A SDA is responsible for the fault diagnosis of one particular signal using either a single signal or multiple signals depending on the nature of the signal. Each SDA will be developed using a common framework that incorporates technologies in signal segmentation, automatic signal feature selection, and machine learning. Some of the agents contain information concerning the cause of faults for other agents, while other agents merely report symptoms. Together these signal agents present to the device agent a full picture of the behavior of the machine currently under diagnosis. The DDA applies case based reasoning techniques to the output of the SDA's to obtain the fault diagnosis of a machine. The initial test bed, ADVANS, will consist of three SDA's implemented within the application area of automotive engineering diagnosis, EngineCoolantTemperature (ECT), Throttle Position (TP) and Engine RPM. ADVANS will be used to study the following research issues related to the DDAS model: signal features meaningful for fault diagnosis, signal diagnosis at three levels of details, fuzzy learning from normal samples only, and feasibility of a distributed agent architecture for engineering fault diagnosis.
The results of this research will lead to a full investigation of the DDAS model. The success of the fully developed DDAS model will provide a powerful and reliable engineering fault diagnostic technology to a broad range of US industries. Speedy, reliable and cost effective product service will significantly increase the global competitiveness of the U.S. manufacturing and service industry.