The Research Of Failure Mechanism Based On Audio-visual Integration And Bond Graph Language For Complex Electromechanical Systems

With increasing levels of domestic equipment technology and increasing demands forengineering practice, more and more complex electromechanical system was put intoproduction and life. Complex electromechanical system is not just the electromechanicalcoupling, but also has the characteristics: functional coupling, energy coupling andstructural coupling. In order to make the system a long time, high precision, stable tocomplete various tasks, it is necessary to monitor the system failure. Therefore, the researchof failure mechanism based on audio-visual integration and bond graph language forcomplex electromechanical systems has become the focus of fault monitoring system.With deep understanding of the concept of complex electromechanical system andcareful study of the components of complex electromechanical system, I put forward themodeling approach on energy for complex electromechanical system. Order to compensatethe sidedness of using single sensor to monitor the status; I establish the monitoring systemof complex electromechanical system with the fusion of visual and auditory information.Bond graph model can not only establishment the physical models and mathematicalmodels, can also research energy changes relying on bond graph model. In the dissertation,doing simulation experiment with MATLAB and20-SIM and realizing the dynamicsimulation for the system. Analyze establishment method of mathematical model for bondgraph model, and make internal state information link perfect. The visual and auditoryinformation as input information, impact power bond graph model, and establish thediagnosis of bond graph model.In order to overcome the problem that different sources of information have differentrepresentations. Unicode different information is easy for data analyzing and data storingand use parameter estimation to evaluate fault.Combined with qualitative reasoning method and quantitative reasoning method,deduce fault mechanism and evaluate fault degree. Defines failure models of bond graphelements, and make the failures of the actual physical components map into the changes ofthe specific parameters and the generalized variables in the model. Establish inference rulesof failure mechanism for complex electromechanical system, and make the faults can be transmitted into any one parameter in the model via the reverse inference and forwardinference.In the dissertation, in order to verify the correctness of the research of failuremechanism based on audio-visual integration and bond graph language for complexelectromechanical systems has become the focus of fault monitoring system, I have donelots of experiments with the software, have done a number of visual processing experimentsand auditory processing experiments, have done the experiment of failure reasoning withaudio-visual integration. Experimental results show that the research can achieve faultlocation and fault evaluation, can inference failure mechanism quickly.