Maintainability Test Data Processing And Demonstration Methods Based On Small Sample For Amored Equipment

Maintainability of armored equipment is an important design feature,which has direct impact on battle effectiveness.Maintainability demonstration can directly verify the maintainability of armored equipment,prove whether the maintainability meets the requirements of military and identify the existence of defective design, so that the maintainability of the equipment will continue to grow.At present, the maintainability demonstration of armored equipment is merely limited to the methods of GJB.Usually the sample size is large and this brings on not only serious wear and tear of equipment,but also high cost and long cycle of development.It's very hard to equip troops as soon as possible.At the same time,the maintainability test data processing is not fine enough, such as direct use of abnormal data or giving up missing data, all of which result in large deviation of assessment and demonstration. Research on armored equipment maintainability data processing and demonstration based on small sample is very important in this case.Based on an armored equipment maintainability research task, with the engineering background, armored equipment maintainability test data processing and demonstration based on small sample are researched in-depth. First of all, armored equipment maintainablity test procedures and the general steps are presented; For the abnormal data and missing data in the traditional maintainability test based on large sample, the appropriate statistical methods are given such as statistical test and multiple imputation; various consistency test and compatibility test have been done for the maintainability data information, and the maintainability characteristics of armored equipment have been analysed intensively to ensure the rational use of the maintainability information; Finally,how to use Bayes sequential posterior odd test to demonstration when the test sample size is small is studied,focused on the determination of prior distribution, particularly as the historical information and the scene test information are incompatible. The mean time to repair of armored equipment has been demonstrated and an example illustrates the effectiveness of these methods.