基于加速度传感器的自行车运动强度评估研究  

Exercise intensity evaluation of cycling based on accelerometer

作　　者：赵月民 陈培友[1] 刘晓翠 ZHAO Yue-min;CHEN Pei-you;LIU Xiao-cui(Nanjing Normal University,Nanjing 210023,Jiangsu,China)

机构地区：[1]南京师范大学

出　　处：《山东体育学院学报》2019年第5期70-76,共7页Journal of Shandong Sport University

基　　金：国家哲学社会科学基金资助课题(13BTY014)

摘　　要：研究目的:基于加速度传感器,建立自行车运动强度预测方程,确定4.8 MET和7.2 MET对应的VM轴加速度计数最佳临界值点。研究方法:101名普通大学生(实验组81人,验证组20人)实验过程中同时佩戴K4b2和GT3X(腰部、大腿、脚踝),在功率自行车上依次进行不同强度(①较低强度:37%~45%VO2max;②中等强度:46%~63%VO2max;③较大强度:64%~91%VO2max)的骑行。选取逐步回归方法建立运动强度预测模型,采用ROC曲线建立VM轴加速度计数最佳临界点。结果:1)脚踝处GT3X三分轴对MET的解释力最高(R2=0.80),且VM轴加速度计数与MET存在较高的相关关系(r=0.89,P<0.01)。2)方程MET=0.00019007×VM+3.121,R2为0.80,SEE为0.61,SEE/Y(%)为10.33%。3)经验证组数据回代检验,各强度水平下方程预测值与K4b2实测值相关系数在0.80~0.85之间(P<0.01);相对误差为8.69%~9.61%;95%的残差基本均落在Bland-Altman散点图Mean±1.96 SD的区间内。4)4.8 MET和7.2MET对应的VM轴加速度计数最佳临界值点分别为9764 counts/min和21138 counts/min。结论:脚踝处是自行车运动GT3X的最佳适配位置,新建方程能够有效地预测运动强度MET,预测准确度较高,所建切点能有效地区分运动强度,可为自行车运动科学监测提供参考。Objective: Based on the acceleration sensor, it established the bicycle exercise intensity prediction equation, in order to determine the best threshold points of VM axis acceleration counts for 4.8 MET and 7.2 MET.Methods: During the experiment,101 ordinary college students(experimental group 81, verification group 20) wear K4 b2 and GT3 X(waist, thigh, ankle) at the same time,then followed by different intensity of riding: 1) Lower intensity: 37%~45% VO2max;2) Moderate intensity: 46%~63% VO2max;3)Greater intensity: 64%~91% VO2max) on the power bike. At last, it established the stepwise regression method for the exercise intensity prediction model,and the optimal critical point of VM axis acceleration counting by using ROC curve. Result: 1) The GT3X third-axis at the ankle showed the highest explanatory power( R2= 0.80),and the VM axis acceleration was higher correlated with MET( r = 0. 89,P < 0. 01). 2) The equation MET =0.00019007×VM+3.121,R2 is 0.80,SEE is 0.61,and SEE/Y( %) is 10.33%.3) The data from the verified group back-tested can conclude that the correlation coefficient between equation project value and K4b2 measured values at each level is between 0.80 and 0.85( P<0.01);the relative error was 8.69% ~ 9.61%;the 95% all fall within the range of Mean ±1.96 SD for the Bland-Altman scattergram.4) 4.8 MET and 7.2 MET VM axis acceleration corresponding to the best cut-off points were 9764 counts/min and 21138 counts/min. Conclusion: Ankle is the best place for bike sports GT3X.The new equation can effectively predict the intensity of exercise MET,prediction accuracy is high,and the built-point can effectively distinguish exercise intensity,that can provide reference for the scientific monitoring of cycling.

关 键 词：三轴加速度传感器 自行车运动 运动强度 MET 

分 类 号：G804.49[文化科学—运动人体科学]