Study On Fault Safety Analysis Methods Of Mechanic-Electronic-Hydraulic System Based On Virtual Prototype

Mechanic-Electronic-Hydraulic system(MEHS)plays an important role in the national economy and military equipment.MEHSs usually have complicated constructions and and large scale.What's more,they always have tight relationship with human beings and environment.Faults in MEHSs might not only lead to the safety threat in the systems themselves,but also lead to the human injury and environment destruction.Thus,it is of great significant to perform comprehensive fault safety analysis for MEHS.However,because of the lack of historic fault data,traditional analysis methods of fault safety can hardly prove effective.In order to solve the above problems,this thesis applies virtual prototyping in the fault safety analysis of MEHS.The research content includes:(1)Modeling method of virtual prototyping of MEHS is studied.The concept and connotation of virtual prototyping are introduced.The modeling procedure of virtual prototying is proposed.Steering system is then studied to explain the key points in the modeling procedure.Verification,Validation and Accreditation(VV&A)of virtual prototyping is studied.The concept and connotation of VV&A are introduced.The procedure to relize VV&A for virtual prototyping of MEHS is proposed.As is the most important technique,result validation method in VV&A for virtual prototyping of MEHS is proposed.Steering system is finally studied to verify the proposed methods.(2)Fault injection and simulation method of virtual prototyping is studied,which provides data foundation of fault safety analysis based on virtual prototyping.On this basis,qualitative safety analysis method of MEHS based on virtual prototyping simulation is proposed.Qualititive conclutions of safety consequence leaded by typical fault are obtained.Quantitative probabilistic risk assessment method based on virtual prototping and Monte Carlo simulation is proposed.This method adequately takes the flexible and efficient advantages of simulation,and effectively relieve the problem of fauty data lacking in the traditional safety analysis methods.Steering system is studied to verify the proposed methods.(3)Fault diagnosis method based on virtual prototyping model is proposed.Virtual prototyping is taken as a special mathematical model.Model based diagnosis method of MEHS.Fault detection method based on neural network prediction is proposed.Input-output data of virual prototyping is taken as the training sample of neural network predition model.Accurate output predition of MEHS is performed and thus fault detection could be relized based on residual checkout.Fault diagnosis method based on simulation knowledge from virtual prototyping is proposed.Feature extraction is conducted based on the simulated data of virtual prototyping.Several pattern recognition methods are used to train the extracted features and thus the knowledge library for diagnosis could be founded.Fault diagnosis is achievable by the mean of ensemble classifier.Steering system is taken as case study to verify the proposed three diagnosis methods respectively.(4)Hybrid assessment-decision method based on fuzzy-event tree is proposed.Key parameters involved in safety status when fault happens are extracted.Safety level is obtained according to the fuzzy safety assessment theory.Fault event tree is illustrated and reverse deduction is conducted according to the safety level.Thus the maneuvers decision could be made.Underwater vehicle is studied to verify the proposed methods.Early warning information generation method based on generalized regression neural network(GRNN)is proposed.Key parameters in the maneuvers decision are extracted.Sample data of early warning information is obtained based on Simulink model.These samples are used to train the GRNN model and then the trained model is used to predict the early warning information.Underwater vehicle is taken as case study to verify the proposed method.(5)Design and realization method of fault safety analysis software platform for MEHS based on simulation is proposed.Steery system is taken as case study.System scheme is firstly design.Then the realization method of software platform based on the integtation of Matlab/Simulink,AMESim,Matlab/VR is proposed.Some vital techniques involved in the software platform are overcomed.Finally,some cases are studied to demonstrate the typical function of the software platform.To sum up,this thesis studies the the fault safety analysis methods of MEHS based on virtual prototyping.In this thesis,fault safety of MEHS is the main clue,virtual prototyping is the application basis and steering system is the case background.The thesis sythetically studies “how is going”,“how to diagnosis” and “how to save” when fault happens to MEHS.Most of the studied content are integrated in the software platform.The presented thesis is supposed to provide valuable reference for fault safety analysis of MEHS.