Comparison Of Two Model-Based Methods Of Remote Diagnosis Of Technological Systems

This paper presents a scheme for remote diagnosis of technological systems, which decomposes the diagnostic task into two subtasks. Due to the utilization of remote resources (off-board), the requirements on the on-board component are reduced with respect to memory and computing efforts. Two model-based diagnosis methods for remote diagnosis are described, which differ in the measurement data required for diagnosis and in using a hybrid or a discrete event model, respectively. Both methods are compared by means of a bench-scale manufacturing cell.