Research On The Fault Diagnosis Of Gun Automatic Load System

Gun automatic load system is one of the key subsystems. As an functional independent part, it has to coordination with the entire system, its dynamic response will directly affect the system's overall reliability. Automatic system for cannon is an mechanical-electrical integration system which includes mechanical, electrical and hydraulic components. Its structure is complex, and it has a lot of working procedures, therefore it's also one of the subsystems which have often broken down. The automatic load system is the key technology in the tower mounted guns'research; however, it's also the difficult technology. When the automatic load system have fault, artificial diagnosis is difficult to meet the intelligent requirement of modern weapons and equipment, so it is urgent to study corresponding fault diagnosis method to achieve the intelligent fault diagnosis of automatic load system.In recent years, computer technology and intelligent technology developed rapidly. At the same time, artificial intelligence is applied to fault diagnosis field more and more, and fault diagnosis is pushed into a new stage. Expert system and neural network are two active branches of artificial intelligence, so they are widely used in fault diagnosis. But the two technologies both have shortcomings. The study find that they are well complementary to each other, therefore this paper designs a kind of intelligent fault diagnosis system which is dominated by expert system and supplemented by the neural network. Then research the working principle from knowledge base, inference engine and explanation mechanisms aspects respectively, and test the system using Visual C + +, SQL Server 2000 and Matlab, combined with the fault mode of gun automatic load system. The application of this system can improve the fault diagnosis efficiency of gun automatic load system, so as to meet the intelligent requirement of modern weapons.