本文结合遗传算法的原理和思想,分析了传统遗传算法在求解题库组卷问题中存在的缺点和不足。并采用了新的编码方式、交叉算子和变异算子对其进行优化。验证得到若题库所含的题目数适中,分布合理,所组卷的试卷质量较好的满足各个控制指标的分布要求。 Based on the principle and idea of ​​genetic algorithm, this paper analyzes the shortcomings and shortcomings of the traditional genetic algorithm in solving the test paper problem. And adopted a new coding method, crossover operator and mutation operator to optimize it. If the problem is validated, if the number of questions contained in the question bank is moderate and the distribution is reasonable, the quality of the test paper in the test paper meets the distribution requirements of each control index.