FAULT DIAGNOSIS TRAINING SYSTEM FOR HYDRAULIC SYSTEM OF LARGE-SCALE ARMAMENT BASED ON PHYSICAL SIMULATION MODEL AND BP NEURAL NETWORK

摘要

In order to take the training of fault diagnosis in hydraulic system instead of practical armament, the fault diagnosis training system was built as a function module of the training simulation system of large-scale armament based on HLA in this paper. Firstly the physical simulation model was built to create and analyze the model of the archetypal system's dynamic character. Then its fault mathematical model was designed and simplified, which analyzed and calculated the injected fault information to simulate the state parameters of the fault system. Finally the faults were located and classified by the BP neural network of the fault diagnosis training system, thereby their locations, causes and disposals are presented and displayed on the visual interface. And the simulation results validate the feasibility and validity of this method.