摘要
在组卷优化问题的研究中,组卷受到多约束条件的限制。为提高在线考试系统中试卷的质量,提出采用概率表示的二进制粒子群优化算法(BPSO)的智能组卷策略,采用粒子群优化算法有效克服遗传算法的局部搜索能力差,以及导致"早熟"和收敛速度不理想等缺陷。在标准粒子群算法基础上,利用贝叶斯公式对粒子群算法进行改进,克服人为因素对算法收敛速度的影响,同时算法的时间性能和空间性能得到进一步提升。通过仿真证明改进算法是一种切实可行的组卷策略。
In order to improve the quality of test sheets applied to Online Exam System, a Binary Particle Swarm Optimization (BPSO) algorithm based on probability was proposed in composing test sheets systems. The Particle Swarm Optimization (PSO) can overcome the shortcoming that Genetic Algorithm (GA) is easy to fall into a local op-timum and prematurity. Bayes formula was introduced to overcome the impact of human factors on algorithm conver-gence speed. Meanwhile, the algorithm performance was further improved in space and time domains. Simulation re-suits show that the algorithm is effective, feasible and practical to test-sheet composing strategy.
出处
《计算机仿真》
CSCD
北大核心
2012年第9期387-391,共5页
Computer Simulation
基金
国家自然科学基金项目(61073189)
上海市第四期本科教育高地建设项目(B-8515-10-0001)
关键词
组卷
贝叶斯公式
二进制粒子群优化
Composing test-sheets
Bayes formula
Binary particle swarm optimization(BPSO)