基于自定义权重遗传算法的组卷算法仿真研究

Study onTest Paper Generating Algorithm Simulation Based on Self-Definition Weight Genetic Algorithm

针对目前众多组卷算法对组卷质量和组卷速度不能同时提高的缺陷,导致了组卷系统收敛效果不佳,迭代次数较多,组卷效率不高,难符合用户要求。为解决上述问题,提出了基于自定义的约束权重比的自适应遗传算法,强调在最有可能成为最终个体的基因里进行搜索,并增加优良个体的保留机制,使其在寻优准确率和搜索速度上均明显优于其他组卷算法,且具有很好的收敛性和实用性。实践结果表明,该方法可以有效地解决智能组卷中的多约束优化问题,并同时提高了组卷速度。

Genetic algorithm is used to study test paper generation speed and multi-restraint optimization question.In view of the defect that present numerous test paper generating algorithms cannot enhance the test paper generating quality and speed simultaneously,which causes the low efficiency of test paper generating system,the auto-adapted genetic algorithm based on the restrained weights is proposed.The algorithm searches the genes which are most likely to be the ultimate individuals and adds the retention mechanism of fine individuals.The optimization accuracy and search speed of the algorithm were much better than the other test paper method,with good convergence and practicability.The practice shows that the method can effectively generate intelligent test paper in the multi-constrained optimization and simultaneously improve the speed of test paper.