Design Of Remote Monitoring Network And Intelligent Fault Diagnosis System For Hydraulic System

The erecting launcher test-bed is mainly used to simulate the leveling,erecting and launching conditions of the special vehicle.It is a large complex system composed of electrical,mechanical and hydraulic systems.It plays an important role in the research and manufacture of the special vehicle.When the test-bed fails,it will affect the development and manufacturing progress of the special vehicle and cause great economic losses.Therefore,in the case of relative lack of field maintenance personnel,the establishment of remote monitoring and diagnosis system,rapid diagnosis of faults and providing solutions will effectively avoid the economic losses caused by long-term maintenance In.In the establishment of remote fault monitoring and intelligent fault diagnosis system for hydraulic system of erecting launcher test-bed,the spiral development method is adopted,and only the most important function design is completed without considering each situation.With this method,the design of the system can be completed quickly and preliminarily.When the system is further improved and developed again,only some modifications and perfections are needed on the basis of the system,which paves the way for the further development of the system.In the design of remote monitoring network and intelligent fault diagnosis system for hydraulic system,firstly,on the basis of understanding the structure and principle of the vertical launcher test-bed system,the fault tree analysis method is used to analyze some faults of the pump station of the hydraulic system,in order to prepare for the knowledge acquisition of the expert system.Then,the mechanism of common hydraulic faults is analyzed,and the data information needed for fault simulation of AMESim is prepared in advance.In the simulation of the pumping station system,the schematic diagram of the pumping station is analyzed and simplified,and then the simulation model of the pumping station is established according to the simplified schematic diagram of the pumping station.When the simulation model is established,the existing test bench is used to carry out experiments under certain conditions to verify the accuracy of the simulation model.When the accuracy of the model is verified,the fault signals obtained from the fault analysis of the hydraulic pump before are used to simulate the fault of the pump station system by changing the parameters,so as to collect the fault samples and overcome the difficulty of collecting the fault samples in the new test-bed.In the design of expert system,by comparing the advantages and disadvantages of rulebased reasoning and neural network reasoning,combined with their fault situations in the field of electromechanical and hydraulic systems,a neural network expert system is proposed,which is mainly based on neural network and supplemented by reasoning based on rule and framework fusion.By analyzing the acquisition and representation of comparative knowledge,the acquisition method based on fault tree knowledge and the knowledge representation based on rule and framework fusion are selected.In the design of inference engine,traditional methods are used to design rule-based inference engine and neural network-based inference engine,CLIPS is used to develop rule-based inference engine,and MATLAB mathematical tools and fault samples from fault simulation are used to train neural network.Gateway and server system are built,and CANET is redeveloped to meet the requirements of the system.When developing fault diagnosis expert system,use case diagram and sequence diagram of the system are analyzed,and then the accuracy of the diagnosis results of the fault diagnosis expert system is checked.In the process of inspection,the test samples obtained from simulation and the fault cases encountered from experiments are used as detection respectively.Then the advantages of reasoning mechanism based on neural network reasoning and rule and framework fusion are analyzed.