Reliability Information Fusion Of Electromechanical Equipment Under Different Conditions

Information fusion technology is widely used in various fields and plays a decisive role in the development of intelligent information society.In the reliability system of mechanical and electrical equipment,information fusion technology also plays a key role.There is a lot of information data to be used in the process of design,research and development of mechanical and electrical equipment.If the relevant information can be fused,the reliability of products would be improved.With more and more complex electromechanical system and more information sources,the reliability requirements of electromechanical system are also getting higher and higher.Information fusion technology is promoting the development of reliability field.This dissertation mainly focus on the high conflict,strong correlation and multi-layer situations of reliability information fusion of electromechanical equipment as follows:(1)A high conflict reliability information fusion model based on D-S evidence theory was established based on the different importance of each evidence in reliability information fusion.A high conflict reliability information fusion model based on interval epistemic uncertainty and fuzzy number epistemic uncertainty was established aiming at the possible uncertainty in high conflict information fusion.The random sampling method,sequential quadratic programming algorithm,particle swarm optimization algorithm and genetic algorithm were used to solve the problem respectively,and the results were compared with the calculation time.The feasibility and accuracy of the model were verified,and the optimal algorithm and adjustment rules that were most suitable for the model were obtained.(2)The optimization and digital simulation methods of Copula function under uncertainty were proposed to solve epistemic uncertainty and the coexistence of random and epistemic uncertainty in the distribution parameters of Copula function.The uncertainty strong correlation reliable information fusion model based on Copula function method was established.Taking gear as the research object,the random sampling method,the data fusion method and the data fusion method were adopted Sequential quadratic programming algorithm,particle swarm optimization algorithm and genetic algorithm were solved and compared to verify the feasibility and accuracy of the model.(3)A reliability information fusion model of multi-layer all voting system was proposed in order to improve the reliability life of multi-layer system in a short time.A multi-layer complete voting system model was proposed in view of the fact that the failure rate of components may be uncertain.The information fusion of multi-layer series,multilayer parallel and hybrid systems based on information entropy was discussed at the same time.The accuracy and feasibility of the model were verified by an example.