Study On The Assessment Of Missile's Terminal Guidance Accuracy Under Small Samples

The terminal guidance accuracy assessment is an important aspect of weapon system appraisement in tactical and technical capability, which has always been drawn attention from the design and test department, and the users. Analyzing and evaluating the weapon system's performance scientifically by the test information can discrease the develop time and save the expense. Thus it is of great importance for weapon system appraisement and investigation.The assessment scheme of weapon system's final guidance has been developed. The problem of three dimensional precision index is discussed, including spherical error probable (SEP) caculation and its mathematical relations to other precision indices. Further more, the approach of bootstrap evaluation for SEP by test information is proposed.The approach of consistency test for small-sample is discussed. The methods of static test data's consistency checkout are summarized systemically. The technique of consistency test for very small sample number is developed. In addition, the approach for one sample is also proposed. The maximum entropy spectrum estimation method is discussed for dynamic test data's consistency checkout.The problem of utilizing test data to assess weapon system terminal guidance accuracy is investigated. Some methods for multi-sources information fusion are discussed, including modified-cross-entropy and adaptive-cross-entropy information fusion approaches, the prior information fusion method based on Bayesian networks. Two integral inference approaches which are reapectively based on across and vertical information have been proposed to fuse test data and prior information. These methods can help to deal with the test imformation more effectively.The factors which affect terminal guidance accuracy are analyzed. The main terminal guidance error models are established and simulated.Then, the assessment of final guidance accuracy is simulated by different methods discussed above based on different kinds of test imformation. The simulation results show that these approaches are appropriate to accuracy evaluation and make the assessment more reliable.