Abundant test data are required in assessment of weapon performance. When weapon test data are insufficient, Bayesian analyses in small sample circumstance should be considered and the test data should be provided by ...Abundant test data are required in assessment of weapon performance. When weapon test data are insufficient, Bayesian analyses in small sample circumstance should be considered and the test data should be provided by simulations. The several Bayesian approaches are discussed and some limitations are founded. An improvement is put forward after limitations of Bayesian approaches available are analyzed and the improved approach is applied to assessment of some new weapon performance.展开更多
A Bayesian method for estimating human error probability（HEP） is presented.The main idea of the method is incorporating human performance data into the HEP estimation process.By integrating human performance data an...A Bayesian method for estimating human error probability（HEP） is presented.The main idea of the method is incorporating human performance data into the HEP estimation process.By integrating human performance data and prior information about human performance together,a more accurate and specific HEP estimation can be achieved.For the time-unrelated task without rigorous time restriction,the HEP estimated by the common-used human reliability analysis（HRA） methods or expert judgments is collected as the source of prior information.And for the time-related task with rigorous time restriction,the human error is expressed as non-response making.Therefore,HEP is the time curve of non-response probability（NRP）.The prior information is collected from system safety and reliability specifications or by expert judgments.The（joint） posterior distribution of HEP or NRP-related parameter（s） is constructed after prior information has been collected.Based on the posterior distribution,the point or interval estimation of HEP/NRP is obtained.Two illustrative examples are introduced to demonstrate the practicality of the aforementioned approach.展开更多
Failure prediction plays an important role for many tasks such as optimal resource management in large-scale system. However, accurately failure number prediction of repairable large-scale long-running computing （RLL...Failure prediction plays an important role for many tasks such as optimal resource management in large-scale system. However, accurately failure number prediction of repairable large-scale long-running computing （RLLC） is a challenge because of the reparability and large-scale. To address the challenge, a general Bayesian serial revision prediction method based on Bootstrap approach and moving average approach is put forward, which can make an accurately prediction for the failure number. To demonstrate the performance gains of our method, extensive experiments on the data of Los Alamos National Laboratory （LANL） cluster is implemented, which is a typical RLLC system. And experimental results show that the prediction accuracy of our method is 80.2 %, and it is a greatly improvement with 4 % compared with some typical methods. Finally, the managerial implications of the models are discussed.展开更多
文摘Abundant test data are required in assessment of weapon performance. When weapon test data are insufficient, Bayesian analyses in small sample circumstance should be considered and the test data should be provided by simulations. The several Bayesian approaches are discussed and some limitations are founded. An improvement is put forward after limitations of Bayesian approaches available are analyzed and the improved approach is applied to assessment of some new weapon performance.
基金supported by the Specialized Research Fund for the Doctoral Program of Higher Education(20114307120032)the National Natural Science Foundation of China(71201167)
文摘A Bayesian method for estimating human error probability（HEP） is presented.The main idea of the method is incorporating human performance data into the HEP estimation process.By integrating human performance data and prior information about human performance together,a more accurate and specific HEP estimation can be achieved.For the time-unrelated task without rigorous time restriction,the HEP estimated by the common-used human reliability analysis（HRA） methods or expert judgments is collected as the source of prior information.And for the time-related task with rigorous time restriction,the human error is expressed as non-response making.Therefore,HEP is the time curve of non-response probability（NRP）.The prior information is collected from system safety and reliability specifications or by expert judgments.The（joint） posterior distribution of HEP or NRP-related parameter（s） is constructed after prior information has been collected.Based on the posterior distribution,the point or interval estimation of HEP/NRP is obtained.Two illustrative examples are introduced to demonstrate the practicality of the aforementioned approach.
基金supported by the National Natural Science Foundationof China (60701006 60804054 71071158)
文摘Failure prediction plays an important role for many tasks such as optimal resource management in large-scale system. However, accurately failure number prediction of repairable large-scale long-running computing （RLLC） is a challenge because of the reparability and large-scale. To address the challenge, a general Bayesian serial revision prediction method based on Bootstrap approach and moving average approach is put forward, which can make an accurately prediction for the failure number. To demonstrate the performance gains of our method, extensive experiments on the data of Los Alamos National Laboratory （LANL） cluster is implemented, which is a typical RLLC system. And experimental results show that the prediction accuracy of our method is 80.2 %, and it is a greatly improvement with 4 % compared with some typical methods. Finally, the managerial implications of the models are discussed.
