Bayesian Approach's Application In Accuracy Assessment Of Weapon System

The accuracy assessment of weapon system is a complex engineering , involving various information . How to make the most of these information and obtain reasonable estimate is always a problem. Bayesian approach can fuse various prior information effectively, and assess the performance of weapon system combining a little information. This paper discusses emphatically several methods about assessing the variation of impact points. First, this paper discusses the general theory about prior information, including the acquirement of prior information,prior information conversion,fusion of prior information of multiple sources, non-information priors and the expression of prior distribution. All of these studies provide Bayesian approach a sound basis. Considering that in the process of assessing, experimental data is general multistage, so the estimation of accuracy parameter is a dynamic problem. In the third chapter, based on the relativity of prior and posterior information, the method of dynamic revised Bayesian is introduced to estimate the variation. A method of dynamic recurrence estimation is developed the estimated accuracy, and solved the difficult brought by small sample. In the forth chapter, Bootstrap and Bayesian Bootstrap's application in accuracy assessment of weapon system is discussed. We study the multiple sensor fusion estimation of prior distribution. In the same time, a Bootstrap adjustment plan on the method of dynamic recurrence estimation mentioned in the third section is given. Finally, on the basis of power prior theory and information processing, a new scheme of dynamic parameter estimation is also put forward. This plan can make use of historical data effectively and reasonably.