一种基于改进遗传算法的组卷算法

A exam-paper generating algorithm based on improved genetic algorithm

为了能更好地解决组卷质量和组卷速度之间的矛盾,提出了一种基于整数分段编码的遗传算法.该算法在保证组卷预期效果的前提下,不仅搜索速度快,而且能够避免遗传算法中经常出现的“早熟现象”,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约束优化问题.

A high quality test paper combination is generally at the cost of considerable time. In order to handle the contradiction between quality and speed of combination, a genetic algorithm based on integer segment is proposed in the thesis. The algorithm guarantees expected result of a test paper combination and has satisfying astringency and practicability, with not only high speed but also avoidance of prematureness phenomenon which is often arisen in genetic algorithm. Practice proves that this algorithm is able to solve the problem of restriction optimizing in the process of intellect test paper combination.