摘要
试卷的质量直接决定了考试的效果,优秀的考试组卷算法能够更好的满足用户设定的标准。结合现有算法和项目实际需求,基于粒子群优化算法的流程,提出了五项改进措施,包括编码方案和试卷适应度值计算方法的改进,提出了新的粒子学习过程,使粒子能够逐代变好,更快收敛到全局最优解。在改进的粒子群优化算法中,又加入了混沌初始化和混沌扰动,通过算法实验与改进,设计并实现一种新的组卷方法以期得到更好的结果,并采用QTI标准进行试题存储,实现试题共享,证明在实际应用中的可行性。
The quality of test paper directly determines the test result,so the excellent test?paper combination algorithm can better satify the standard set by the user. In combination with the practical demand of the project and existing algorithms,5 improvement measures are proposed based on the particle swarm optimization algorithm. The computing method of test?paper fit?ness and coding scheme were improved. The new particle learning process was improved to make the particles better for each generation and have the performance of fast convergence to the global optimal solution. The chaos initialization and chaos distur?bance are added into the improved particle swarm optimization algorithm. A new test?paper combination method was designed and implemented by means of algorithm experiment and improvement to obtain the better results. The QTI Standard is used to store the test questions and share them. The method feasibility was verified in practical application.
出处
《现代电子技术》
北大核心
2016年第13期112-115,120,共5页
Modern Electronics Technique
基金
国家自然科学基金项目(81271284)
关键词
组卷
粒子群优化算法
改进与优化
QTI标准
test-paper combination
particle swarm optimization algorithm
improvement and optimization
QTI Standard
