Study About Fault Diagnosis Of The Anti-skid Braking System Of Aircraft

The anti-skid braking system is an important part of equipment in aircraft, it does very important action in the course of flying-off and landing.If the anti-skid braking system is good or bad, it can affect the security of aircraft and pilot. In order to build up security in the anti-skid braking syetem,pledge natural flying educate,ferform the role of campaign,it is very important and necessary that we diagnose anti-skid braking system in the course of sliding and landing. But skeleton army often adopt experience and handbook to diagnose fault on the anti-skid braking system of aircraft.It's efficiency is low.,it's realtime is low and it's subjectivity is strong. It can't suffice the need of safe training.This dissertation has researched method of fault diagnosis. It has sumed up the advantage and disadvangtage of the method . This dissertation unites actual thing and technique in the skeleton army. This dissertation unites the elements and peculiarity of the anti-skid braking system of aircraft.Expert system on FTA is adopted to diagnose the whole anti-skid braking system of certain aircraft. It has adopted BP neural network and fault analysis on model to diagnose pivotal work.Fault tree on anti-skid braking system of certain aircraft is set up, depending on consulting data and suming up experience. Fault tree is qualitative analysised. Inorder to improve rational efficiency, fault tree is predigested. The knowlendges of expert system are adopted by introducing the least fault tree. Procreant rule is adopted in expressing knowledge. Knowledges are transformed the form of conclusion rule and condition. Knowledges are deposited repository. It can improve inferential efficiency and easy to look up. VB 6.0 is adopted as tool to open out expert system. ACCESS 2002 is adopted as tool to open out data-base. In the working of data-base SQL is adopted.BP neural network is adopted to diagnose the control box of defending slide. Three-floor neural network is set up, emulational study of fault diagosis is played by kit of neural network in MATLAB. Conclusion of diagnosis is came out. This dissertation adopts an unknown import compensate observer to diagnosis servovalve. Emulational study of fault diagnosis in servovalve is played by SIMULINK in MATLAB. The different resultt is contrasted between state observer and unkown import compensate observer.