An Experimental Design Method for Extended Translation Latin Hypercube Based on Sudoku Grouping

Abstract

The invention proposes an extended translation Latin hypercube test design method based on Sudoku grouping, which belongs to the field of computer simulation test design. The present invention groups the factors on average, and generates an extended translation Latin hypercube design according to the number of factors in each group; the number of factors is rounded up and the square number is taken to generate the Sudoku of the square number grid; the row / column of the grid number before the Sudoku is selected , delete redundant factor numbers, and group them evenly according to the number of factor groups; output the sampling results of each group of factors in sequence, serially combine the results of each group in each row / column and rearrange them in ascending order of factor numbers; The sampling results corresponding to the columns are arranged in parallel to generate the final Sudoku group-based extended translation Latin hypercube sampling space. The invention solves the problems of long time for the extended translation Latin hypercube design, too many generated sample points and poor effect, thereby effectively reducing the simulation test time and making the test results more representative.

Description

technical field [0001] The invention belongs to the field of simulation test design, and relates to a test sample space design method, in particular to an extended translation Latin hypercube test design method based on Sudoku grouping. Background technique [0002] Under the condition of weapon system confrontation, different weapon system equipment data and tactical methods will affect the final confrontation effect. It is necessary to select different parameter combinations for these influencing factors and conduct multiple experiments to obtain the optimal weapon system technical indicators and tactical approach. However, it takes a lot of manpower and material resources to conduct multiple real-scenario experiments, so a global understanding of the problem is often obtained through the running results of a large number of simulation samples. Although computer hardware capabilities have been increasing rapidly, it is unrealistic to execute all possible situations in a s...

AI Technical Summary

Problems solved by technology

[0005] The weapon system countermeasure model needs to comprehensively consider the influence of multiple factors (the number of factors>>6), and the extended translation Latin hypercube design cannot obtain certain levels of factors, so the test results are not representative, and a method that can quickly generate approximate optimal The design method of Uradin hypercube provides a representative sample space for weapon system combat simulation test

Images