Bayesian Estimation based K-1 Gas-Mask Shelf Life Assessment using CSRP Test Data |
Jong-Hwan Kim, Chi-jung Jung, Hyunjung Kim |
1Department of Mechanical & Systems Engineering, Korea Military Academy 2Defense Agency for Technology and Quality |
CSRP 시험데이터를 사용한 베이시안 추정모델 기반 K-1 방독면 저장수명 분석 |
김종환, 정치정, 김현정 |
1육군사관학교 기계.시스템공학과 2국방기술품질원 신뢰성시험센터 |
|
Abstract |
This paper presents a shelf life assessment for K-1 military gas masks in the Republic of Korea using test data of Chemical Materiels Stockpile Reliability Program(CSRP). For the shelf life assessment, over 2,500 samples between 2006 and 2015 were collected from field tests and analyzed to estimate a probability of proper and improper functionality using Bayesian estimation. For this, three stages were considered; a pre-processing, a processing and an assessment. In the pre-processing, major components which directly influence the shelf life of the mask were statistically analyzed and selected by applying principal component analysis from all test components. In the processing, with the major components chosen in the previous stage, both proper and improper probability of gas masks were computed by applying Bayesian estimation. In the assessment, the probability model of the mask shelf life was analyzed with respect to storage periods between 0 and 29 years resulting in between 66.1 % and 100 % performances in accuracy, sensitivity, positive predictive value, and negative predictive value. |
Key Words:
Chemical Materiels Stockpile Reliability Program, Shelf Life Analysis, Bayesian Estimation, Classification |
|