Research And Application On Reliability Calculation Method For Equipment Integrated System

With the continuous development of modern science and technology, Equipment Integrated System (EIS) is formed by integrating equipment unit in a specific connection with highly coupled, highly integrated. EIS has been widely become the mainstream model of modern production systems, its reliability plays an important role on the system of production, on the industry’s normal, efficient operation. However, the reliability analysis for EIS is also limited to the use of traditional single method, which cannot meet the needs of its reliability analysis, because of its complex structure, diverse equipment connections and so on. It is especially important that analysis and forecasting system reliability based on the equipment unit reliability.In view of the characteristics of the EIS and sophisticated network theory, then a reliability hierarchical network model that can describe the system structure is built based on equipment units and subsystem as nodes respectively, the failure rate as node attributes, the edge is the relations between the node, right of edge is the strength. This method has structured, modular, network modeling characteristics that can effectively analyze system reliability by adopting form-the-bottom-up method. Then propose a reliability calculation method for node, the method takes into account the reliability of the correlation between the nodes and the node itself. The maximum probability of system state space is obtained based on the reliability of nodes and the generate algorithm. To improved Markov process transition matrix in modal space of the system. And describe the system reliability to use the availability, mean time between failure frequency and failure parameters. Then get the key nodes of the system through structural reliability-related indicators and importance of nodes in the system structure. By building and analyzing a reliability model for the High-speed train system verify the practicability of model. This study indicates that this method for the modern complex large system analysis provides a new way.