基于大数据分析的运动损伤估计模型设计

发布时间：2018-06-05 11:40

本文选题：大数据分析 + 运动损伤　；参考：《现代电子技术》2017年17期

【摘要】：为了预防运动损伤,保证运动员的身体安全,提出一种基于大数据分析的运动损伤估计模型。介绍了大数据分析技术,将引发运动损伤的原因划分成A内部致伤因子、B外部致伤因子、C刺激诱发因子。在大数据分析技术的基础上,通过RBF神经网络构建运动损伤估计模型。分析了基本RBF神经网络,将高斯函数看作隐含层单元的激活函数,通过一种简单的方式设计隐含层,令所有风险等级和一个高斯函数相对应。对径向基函数中心、权值和宽度进行更新,通过梯度下降法对径向基函数中心和其余参数进行学习。依据运动损伤风险样本库对RBF神经网络进行训练,将运动损伤数据输入到RBF神经网络中,当传输数据和某运动损伤风险等级相对应时,RBF神经网络将输出相应值,从而实现运动损伤估计。实验结果表明所设计模型精度和效率都高。
[Abstract]:In order to prevent sports injury and ensure athletes' physical safety, a model of sports injury estimation based on big data analysis is proposed. In this paper, big data analysis technique is introduced. The causes of sports injury can be divided into A internal injury factor and B external injury factor C stimulation factor. On the basis of big data analysis technology, the motion damage estimation model is constructed by RBF neural network. The basic RBF neural network is analyzed and the Gao Si function is regarded as the activation function of the hidden layer unit. The hidden layer is designed in a simple way so that all risk levels correspond to a Gao Si function. The center weight and width of radial basis function are updated and the center of radial basis function and other parameters are studied by gradient descent method. The RBF neural network is trained according to the sample base of motion injury risk, and the data of motion injury is input into the RBF neural network. When the transmission data correspond to a certain risk level of sports injury, the corresponding value will be outputted by the neural network. Thus, motion injury estimation is realized. The experimental results show that the designed model has high accuracy and efficiency.
【作者单位】：郑州工程技术学院体育学院;
【分类号】：R873;TP311.13

【相似文献】