长跑运动员耐力预测模型研究仿真
发布时间：2019-04-17 20:32
【摘要】：对长跑中运动员耐力进行准确预测,可以对运动员的身体变化进行合理分析,提出科学合理的训练指导意见。进行耐力预测时,需要考虑跑步中的加速度、天气等自变量,在这些自变量影响下提取运动员的运动负荷特征,以此为基础,建立耐力预测模型,但是传统方法很难有效表达这些因素产生的动态影响,无法提取准确的运动负荷特征,存在建模误差大的问题。提出基于改进人工鱼群神经网络算法的长跑中运动员耐力预测建模方法。先利用BP神经网络算法获取运动员在长跑中的踝关节加速度综合指数,利用计算的结果定义能耗预测中的自变量,提取长跑中运动员的运动负荷特征,利用人工鱼群算法对BP神经网络的初值权值和阈值进行优化,组建运动员耐力预测方程,利用支持向量机的默认RBF核函数搜索最佳的惩罚因子,利用搜索的结果对预测方差进行训练,获取最优的长跑中运动员耐力预测模型。仿真结果表明,所提算法可以为长跑中运动员耐力预测提供参考。
[Abstract]:The accurate prediction of athletes' endurance in long-distance running can reasonably analyze the changes of athletes' body and put forward scientific and reasonable training guidance. When carrying out endurance prediction, it is necessary to consider the acceleration, weather and other independent variables in running. Under the influence of these independent variables, the sports load characteristics of athletes should be extracted, and based on this, the endurance prediction model should be established. However, the traditional method is difficult to express the dynamic effect of these factors effectively, and can not extract the accurate motion load feature, which has the problem of large modeling error. In this paper, an improved artificial fish swarm neural network (Ann) algorithm is proposed to predict the endurance of athletes in long-distance running. Firstly, the BP neural network algorithm is used to obtain the ankle acceleration index of the athletes in the long-distance running, and the independent variables in the energy consumption prediction are defined by the calculated results, and the sports load characteristics of the athletes in the long-distance running are extracted. The initial value weight and threshold value of BP neural network are optimized by artificial fish swarm algorithm, and the endurance prediction equation of athletes is set up. The optimal penalty factor is searched by the default RBF kernel function of support vector machine. The prediction variance is trained by the search results, and the optimal endurance prediction model of long distance runners is obtained. The simulation results show that the proposed algorithm can be used as a reference for the endurance prediction of long-distance runners.
【作者单位】： 郑州大学西亚斯国际学院;
【分类号】：G822.3
本文编号：2459773
[Abstract]:The accurate prediction of athletes' endurance in long-distance running can reasonably analyze the changes of athletes' body and put forward scientific and reasonable training guidance. When carrying out endurance prediction, it is necessary to consider the acceleration, weather and other independent variables in running. Under the influence of these independent variables, the sports load characteristics of athletes should be extracted, and based on this, the endurance prediction model should be established. However, the traditional method is difficult to express the dynamic effect of these factors effectively, and can not extract the accurate motion load feature, which has the problem of large modeling error. In this paper, an improved artificial fish swarm neural network (Ann) algorithm is proposed to predict the endurance of athletes in long-distance running. Firstly, the BP neural network algorithm is used to obtain the ankle acceleration index of the athletes in the long-distance running, and the independent variables in the energy consumption prediction are defined by the calculated results, and the sports load characteristics of the athletes in the long-distance running are extracted. The initial value weight and threshold value of BP neural network are optimized by artificial fish swarm algorithm, and the endurance prediction equation of athletes is set up. The optimal penalty factor is searched by the default RBF kernel function of support vector machine. The prediction variance is trained by the search results, and the optimal endurance prediction model of long distance runners is obtained. The simulation results show that the proposed algorithm can be used as a reference for the endurance prediction of long-distance runners.
【作者单位】： 郑州大学西亚斯国际学院;
【分类号】：G822.3
【相似文献】
相关期刊论文 前7条
1 王炳彝;对运动成绩预测模型选定过程的研究[J];体育与科学;1993年02期
2 汪鸿,吴云飞;十项全能运动员成绩多因素非均衡性预测模型的研究[J];山西师大体育学院学报;2005年01期
3 师义民，，刘桂林，付刚，杨昭军;体能预测模型[J];预测;1994年05期
4 邓美兰;张珍;;奥运会男子十项全能前三名成绩的灰色预测模型的研究[J];数学的实践与认识;2009年18期
5 张杰,叶东惠;体育预测成绩的无差异性检验分析研究[J];北京体育大学学报;2004年07期
6 马祥海;;我国优秀十项全能运动员成绩的灰色关联分析及灰色GM(1,1)预测模型的建立[J];首都体育学院学报;2012年06期
7 ;[J];;年期
相关重要报纸文章 前1条
1 南方日报记者 彭琳 实习生 周鹏程;预测世界杯之外 大数据还能做什么[N];南方日报;2014年
相关硕士学位论文 前1条
1 马婷婷;基于序列与多因素分析的奥运会运动成绩预测模型[D];南京信息工程大学;2012年
本文编号：2459773
本文链接：https://www.wllwen.com/jiaoyulunwen/tylw/2459773.html
