基于改进遗传算法的智能组卷研究

On Intelligent Test Paper Auto-Generation Based on Improved Genetic Algorithm

通过在遗传算法中引入个体浓度的选择机制和记忆机制,确保了进化过程中种群内个体的多样性,避免局部收敛,保证了算法朝优化方向进化.实验结果表明改进算法能跳出局部收敛,有效避免了早熟产生和遗传退化现象出现.

Intelligent test paper auto-generation deals with a multi-objective combinatorial optimization. As a common tool to generate test paper automatically, Genetic algorithm could not achieve the best solution frequently due to the local convergence. In this paper the concentration mechanism and memory mechanism are combined into the genetic algorithm to ensure the population diversity and avoid local convergence. Experiment results show the proposed algorithm can be applied to generate test paper automatically and effectively.