The application of a distributed system-level diagnosis algorithm in dynamic positioning system

This paper introduces the application of a distributed system-level fault diagnosis algorithm for detecting and diagnosing faulty processors in dynamic positioning system (DPS) of an offshore vessel. The system architecture of DPS is a loose coupling distributed multiprocessor system, which adopts the technique of Intel´s MULTIBUS II and develops a software application on the platform of iRMX OS. In this paper a new approach to the diagnosis problem is presented, including an adaptive PMC model, distributed diagnosis including self-diagnosis and interactive-diagnosis, and system graph-theoretic model. The self-diagnosis fully utilises the individual results of built-in self-tests as a part of diagnosis work. Interactive-diagnosis means that in the system fault-free units perform simple periodic tests on one another under the direction of the graph-theoretic model by interactively communicating, and every unit can only send the diagnosis information to its considered fault-free units. Finally, we illustrate the procedure of diagnosis verification. The results obtained show that the adaptive PMC model is applicable, the distributed system-level diagnosis algorithm is proper, and the applications of diagnosis and verification are reliable and practicable.