多种群变异遗传算法在自动化组卷中的运用

A Genetic Algorithm Based on Multi-population Nutation Applied in Automatically Formulating Test Paper

基于自动化组卷模型中对各个组卷参数要求精确程度的不同,通过对传统模型的改进,本文建立了新的自动化组卷的数学模型;并在此数学模型基础上,针对传统遗传算法组卷速度较慢,存在"封闭竞争"问题等等,本文借鉴生物界中存在的"杂交"优势原理,提出了基于多种群变异技术的单亲遗传算法,提高了组成试卷的质量,缩短了组卷时间。

Because of the different accuration requirement of the parameters used in formulating test paper automatically,By improving traditional model,this paper establishes a new mathematic model of formulating test paper.Referring to this mathematic model and hybridize superiority existing in organism world,this paper puts forward a new partheno-genetic algorithm basing on multi-population nutation.It aims at the slow speed of formulating test paper and closing competitiom lying in traditional genetic algorithm.This new genetic algorithm can improve test paper quality and shorten the time of formulating test paper considerably.