Simulation on Life Distribution and Reliability of Long Life Mechanical and Electrical Hybrid System

The analysis on life and reliability needs data. For a complicated and hybrid system, life information and data are mainly obtained from the testing information of subsystems or components. It is a main approach to develop the model to transform all information on components and subsystems into the hybrid system and then determine its life distribution type. Because of differences on test conditions and life distributions of components and subsystems, it may carry a mass of reduced error if the test data of all components and subsystems is converted to the life data of the whole system. If the life distribution can be obtained according to structural relevance of the components, subsystems and the system, it will reduce the error produced during each stage of data transformation and improve the analyzing accuracy. Taking a complicated series system which consists of mechanical subsystems (Life obeys Weibull distribution) and electrical subsystems (Life obeys Exponential distribution) as an example, the distribution type of the hybrid system's life is proved to obey Exponential distribution, by establishing reliability model of the series system about Weibull and Exponential distribution and simulating the hybrid system's life distribution directly by life information of all subsystems. And according to the engineering conditions and requirements, on the basis of minimum integrated bias, maximum probability of estimated reliability no less than real reliability and minimum mean square error between estimated reliability and real reliability, the optimized fitting distribution curve of the hybrid system is obtained by genetic algorithm method. The result shows that the curve is very good for fitting the real curve. The model can reduce the error produced by data converting processes and hypothesis on the system's life type, simplify the analysis and calculating procedure, and can improve the life and reliability analyzing accuracy.