Based On The Genetic Algorithm Of Automatic Group Volume System Design And Implementation

Test-sheet determines the effiency and quality of test-sheet generating. The traditional algorithm for generating test-sheet is time-consuming with poor success ratio and generates test-sheet not qualifying for a real test. On the other hand, genetic algorithm meets the requirements of generating test-sheet better due to its predominant features such as adaptable flexibility, intelligent search and good convergence. In the environment of Visual Basic6.0and SQL Server2005, use C/S (client/server) for building framework, gen Test is deemed as an important approach to measure the education quality and level of students. However, more and more experts and scholars start close studies and researches on the intelligent test-sheet generating theory so that teachers might be released from numerous test-related works while teaching could be separated from test to ensure a justy and fair result of the test and reflect the actual ability level of the testees.Test-sheet generating is always an issue of multi-objective constrained optimization. As the core of test-sheet generating system, the design of the algorithm for generating etic algorithm will develop a user-friendly, convenient in operation and practically useful system interface.To resolve the incidental the prematurity and slow convergence in global optimization by normal genetic algorithm, this system adopts a coding strategy of natural number partition, which reduces procedures in coding and decoding and simplifies the solving issues. It chooses operators by using the method of best individuals protection combining with roulette algorithm, adjusts dynamically the probability of crossover and probability of mutation by using adaptive function so that it can improve the convergence spped and improve the possiblity of find out global optimum. The test-sheet and answer-sheet will be saved in Microsoft word for convenient layout settings and printing. Practices also prove that the improved genetic algorithm shortens the time for system generating test-sheet of high quality and meets the practical demands of users.