Font Size: a A A

Design And Development Of Intelligent Test Paper Generation System Based On Improved Genetic Algorithm

Posted on:2016-09-24Degree:MasterType:Thesis
Country:ChinaCandidate:J K HuangFull Text:PDF
GTID:2428330473464913Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the development of computer technology,and great changes have taken place in teaching technologies.Examination system has been play an important role in teaching activity,it can effectively evaluate the students' knowledge level.On-line automatic generation system can significantly shorten the test cycle,reduce the repetition labor teachers,improve the work efficiency.But the traditional group volume system error is big,the success rate is low,cannot satisfy the needs of users.In this paper,south of south China university of technology school of sports discipline for the object in this paper,a smart group volume system based on improved genetic algorithm.This thesis points the problem of automatic group volume constraints,established the mathematical model of the smart group volume,and on the basis of the traditional genetic algorithm a new improved genetic algorithm is presented.Mainly from the following several aspects are studied.1)Study computer smart group volume system mechanism and system requirements,design meet the smart group algorithm model of volume,to effectively solve the traditional genetic algorithm fast convergence and the lack of guidance,improve the selection and mutation mechanism,realize the request of the smart group volume,in the selection mechanism of traditional adaptive genetic algorithm was improved.In traditional algorithms cannot meet the requirements of smart group volume,so it is necessary to improve the traditional genetic algorithm,makes the algorithm can well meet the demand of the smart group volume,volume to complete the online examination system group.The algorithm firstly random initialization n species,and then according to the problem domain of individual fitness in every generation value size,use the roulette wheel selection operator ACTS on the populations of individual choice.And with the aid of natural genetic recombination crossover and mutation genetic operators,produced on behalf of the new solution set of the population.This process will lead to the later generation populations adapt to value higher than the previous generation,the last the best individual in the population after decoding,the approximate optimal solution can be used as a problem.Algorithm simulation experimental results show that the improved algorithm is moreefficient than the traditional algorithm to solve the problem of automatic group volume.2)With the improved adaptive genetic algorithm(ga),research and design a more efficient,scientific,strong smart group algorithm.After improvement of genetic algorithm,can be achieved for different constraint conditions for weight assignment,conform to the requirements of the constraint conditions to random variation in the fitness of chromosomes,in the process of the group,each item has a certain orientation,and the performance of the algorithm is applied to the group.3)The design and implementation of automatic group volume system.On the basis of the improved genetic algorithm,this paper designed and developed the smart group volume system to adopt three layer architecture model,the system includes four basic function module exam question bank management module,respectively,the examinee examination process management module,test score management module,and system information management module.In this thesis,the development of smart group volume system through the functional and performance testing,the results show that this system is easy to operate,human-computer interaction friendly interface,and realizes the on-line automatic group volume of all functional modules;And on the performance,this system is superior to the traditional system,have certain advancement..
Keywords/Search Tags:Genetic algorithm, intelligent test paper generation, roulette wheel method, module
PDF Full Text Request
Related items