为了解决学生耐力典型评价标准中台阶试验和耐力跑评分标准不一致的问题,引入单边和对称Kullback—Leibler距离,作为不同得分评价体系所决定的耐力跑学生得分概率分布与台阶实验得分概率分布的测度,使用基于元启发算法评价体系的5个算法对两种问题分别进行了求解。比较结果显示,DE在求解单边和双边KL距离问题时,均展现了较好的特性。实验同时证实,最佳算法得到的新得分体系显著缩短了与台阶实验概率分布的KL距离,说明通过优化得分体系提高耐力跑测试科学性的尝试具有可行性。
To solve the inconsistent standards of the step test and endurance running in the endurance evaluation criteria, this paper introduces the concept of unilateral and symmetrical Kullback-Leibler distance as the measure of the probability distribution of step test and endurance running. And five meta-heuristics algorithms are employed to obtain the results. The comparison indicates that DE shows the best characteristics. The experiment also confirms that the new KL distance of the step test and endurance running is significantly shorten, which verified the feasibility of the new strategy.