Research On Design Of Integrated Test

With the increasing of complex and accelerating of renewal speed of weapon and equipment, traditional test and evaluation method(T&E) is more and more difficult to meet the new requirements of short cycle and low cost. Integrated Test and Evaluation(IT&E) is in the mainstream in recent years. The concept of IT&E emphasizes to make full use of the information from different test methods, different test stages and different subsystem tests. Comparing to experiment design problem in T&E, the number of factor variables is increased and the distribution of response variables extends from the normal distribution to the exponential family distribution in experiment design problem in IT&E. Prior information and requirements of the experiment design problem change as well. This paper focus on the research on design of experiment in IT&E and the main contents are as follows:Firstly, according to the selection problem of factor variables in IT&E, the method of analysis of factor variables in design of experiment based on system engineering theory is presented. Interpretative structural model of the system is established to analyze the arrangement and relevance of the system and Analytic Hierarchy Process(AHP) is used for further analysis in the variable selection problem.Secondly, according to the changes in the distribution of response variables, prior information and the purpose of one experiment, the improved method in design of experiment based on generalized linear regression model is presented, including D-optimal design of generalized linear regression model, D-optimal design of logistic regression model, precision design, probability density weighted D-optimal design, Bayesian D-optimal design and multiple responses D-optimal design problem.Finally, using the simulation data, design of experiment in the performance test of a certain type of air to ground missile is presented and the correctness of the method presented in this paper is proved.