Lower hybrid current drive (LHCD) is an efficient method for noninductive current drive in fusion devices. The LHCD system has been constructed on the Experimental Advanced Superconduct Tokamak (EAST). It is a complex system due to lots of devices involved. Each device has possibility of faults, which causes great difficulties in fault diagnosis. Consequently, a fault diagnosis expert system is essential for a safe and steady operation of the LHCD system. This paper proposes an expert system called LFDES (lower hybrid current drive fault diagnosis expert system) to aid operators in diagnosing and analyzing abnormal situations of the LHCD system. After a brief description of the structure of LHCD system, the LFDES architecture, the knowledge base, the inference engine and the database are presented in detail. Based on an empirical knowledge, the diagnostic tree of LHCD system is built. A fuzzy group multiple attribute decision making method is used to determine the priorities of nodes in the diagnostic tree. KDevelop tool, QT Designer tool and Linux operation system have been used in developing the proposed system. In the study, satisfactory results were obtained. The analyses of the results indicated that LFDES can provide reliable, efficient and economical service.
