Space Camera Fault Diagnosis Expert System Research and Implementation
|School||Graduate Schoo,Chinese Academy of Sciences|
|Course||Applied Computer Technology|
|Keywords||Expert System Fault Tree Analysis Neural Network Frame|
In order to improve the efficiency of space camera troubleshooting and shorten troubleshooting time, reduce troubleshooting human and material resources, the design can be more than a collection of the best experience of experts to achieve human space camera combined diagnosis of fault diagnosis expert system . Elaborate fault diagnosis technology research status at home and abroad to introduce expert systems , fault tree analysis and neural network fault diagnosis technology features and basic methods . For their advantages and disadvantages , will take reasonable combination of the three fault diagnosis method for fault diagnosis. In the comparative analysis , based on the fault tree has established itself as one of the means of knowledge acquisition , rule -based framework and knowledge base construction scheme . Using neural network self- learning ability , BP neural network model and network model set up systems Bayes learning machine . Since the main source of failure data FMEA table , according to the characteristics of knowledge acquisition , build space camera fault tree . Failure is mainly expressed in production rules . Then introduced the fault diagnosis inference mechanism . Reasoning using forward and reverse hybrid reasoning . Object-oriented language java and SQL Server as a development tool . Knowledge creation, inference engine and other experts explain procedures to implement parts of the system using java , knowledge of storage using SQL Server database to complete. For space camera for fault simulation , using the expert system for fault diagnosis , the results show that expert systems can quickly analyze the fault phenomena , proposed diagnostic results are consistent with reality.