Design And Implementation Of Fault Diagnosis System For A Certain Autopilot

Ships usually navigates on sea for a long time,and far away from land.The diagnosis of the electronic equipments is usually very difficult for mariners.Especially when the auto pilot was out of order,the navigation would be stopped if the mariner couldn’t diagnose it timely.This thesis develop a diagnosis system which can help mariner repairing the fault auto pilot.This thesis distinguishes wires diagnosis and circuit board diagnosis,according to the character of auto pilot.Electro circuit board is only a small part of the whole auto pilot circuit,and there have spare board for each circuit.So circuit board usually didn’t need diagnosis on navigation.Auto pilot is usually composed by many different parts which was arranged in different rooms.More wires and multi-switches were needed to contact the isolated components together,which makes it difficult to diagnosis wires by the diagnosis methods mentioned in current scientific articles.This article improved the structure of the Fault Tree and enriched the node information of the Fault Tree.Then this method not only can help to diagnose but also exert users’ positivism.So it can meet the need of maintenance requirement.In the aspect of circuit board fault diagnosis,a diagnosis method based on improved wavelet neural network is proposed,and a fault diagnosis model is designed according to the circuit board fault of the autopilot.The board output is analyzed to determine the location of the faulty component or faulty subcircuit on the board.The crew can use this fault diagnosis system to locate the fault with simple operation.