FAULT DETECTION BASED ON PROBABILISTIC ROBUSTNESS TECHNIQUES FOR BELT CONVEYOR SYSTEMS

摘要

In this paper, problems and application of the computation of false alarm rate (FAR) and threshold for belt conveyor systems are studied. Based on an information system which is developed to meet the requirements on monitoring and fault detection for large scale belt conveyor systems, the probability distribution of model uncertainty is assumed to be known and will be taken into account for the design of the fault detection system. The use of the probabilistic information will get a less conservative result, compared with the worst case handling of the system uncertainties. The solution and the application on the belt conveyor system will be illustrated.