| With the developments of product quality and operation requirements, reliability, maintainability, and supportability are playing increasingly crucial roles, and written in manufacture, production, management, and use documents as important controllable indexes. Maintainability not only reflects the difficult degree of maintenance, but also determines the designs and the levels of the product supportability, and also has an important effect on availability and life cycle costs.To small sample, maintainable and large complex system just like as a ship, how to choose maintainability index realistically, establish rational maintainability model scientifically, and carry out ship system maintainability prediction, is not only the objective requirement of the development of general reliability theory, but also the important subject of the shipbuilding industry for recent years. Based on the brief review of correlative reliability and maintainability theories, the paper discusses the involved maintainability prediction theories and analysis methods deeply from actual conditions, and advances a Monte Carlo simulation method to predict system maintainability. Then system mean time to repair is considered as the object, and a new concept, the ratio of efficiency to time for fault diagnosis is proposed. According to this ratio, we can prioritize these units and find the best procedure to remove the faults. Moreover, the system maintenance path is chosen by the discrete random variable sampling method, and the system mean time to repair is predicted by using MATLAB software. Finally, the method is further validated by a ship navigation system sample, and the conclusion is satisfied.The contributions of the paper lie in:a) Starting with the reliability and maintainability theories, according to the structure characteristics and missions, Failure mode and effect analysis, fault tree analysis and level of repair analysis are all chosen as the basic methods for ship maintainability prediction, and then the complexity of ship maintainability prediction model, lack of information and method can be well solved.b) Based on the analysis of the correlative factors of system mean time to repair, system maintenance procedure is researched. The ratio of efficiency to time for fault diagnosis is considered as the sorting basis to determine the system maintenance procedure, and then the best procedure to obtain the best diagnostic result with least cost in the shortest time is found out.c) According to actual situation that lack of information and calculational difficulties are difficult to solve by traditional mathematical methods, discrete random variable sampling method and Monte Carlo simulation method are advanced to predict system mean time to repair, and the corresponding MATLAB general purpose simulation program is also given. Finally, a ship navigation system maintainability prediction is made an example to validate this method. |
