自适应遗传算法在智能组卷中的应用

Self-adaptive Genetic Algorithm is Proposed in the Intelligent Test Paper Application

遗传算法以其具有自适应全局寻优和智能搜索技术,并且收敛性好的特性,能很好地满足自动组卷的要求,但简单遗传算法比较容易引起"未成熟收敛"和"搜索缓慢"等问题。本文提出自适应性的交叉和变异算子,能够降低多重约束目标,有效维持种群的多样性,避免产生局部最优解,改善未成熟收敛。

With the genetic algorithm with adaptive global optimization and intelligent search technology,and convergence good characteristic,can meet the requirements of the automatic group roll,but simple genetic algorithm is easy to cause ＂premature convergence＂ and ＂slow＂ search.This paper puts forward the adaptive crossover and mutation operators,can reduce multiple constraints goal,maintain effective the diversity of population,avoid to produce local optimal solution,improving premature convergence.