基于智能手机的健身跑数据采集及演化建模
发布时间:2018-03-06 16:34
本文选题:健身跑 切入点:智能手机 出处:《智能系统学报》2017年05期 论文类型:期刊论文
【摘要】:针对Brzostowski方法因采集的速度数据精度不高、采用的建模算法搜索能力受限,而导致难以获取高精度健身跑模型的问题,本文提出一种基于智能手机的健身跑速度数据获取及演化建模方法。首先,提出基于智能手机多传感器和中值滤波的健身跑速度采集方法,以滤除由于手机间歇性的姿态变化而产生的三轴加速度信号脉冲噪声,并结合方向传感器对手机三轴加速度中包含的重力分量进行过滤;然后,设计一种健身跑演化建模算法,以增大搜索空间为获取更优的健身跑模型提供支持。实验结果表明,本文提出的方法可以比Brzostowski方法获取更为精确的速度和健身跑运动模型。
[Abstract]:According to the Brzostowski method for the accuracy of the data acquisition speed is not high, the modeling algorithm uses search ability is limited, which is difficult to obtain high precision of fitness running model, this paper presents an intelligent mobile phone based on the fitness run speed data acquisition and evolution modeling method. First, put forward the speed acquisition method of intelligent mobile phone running multi sensor and median filtering based on fitness, to filter out due to attitude change caused by the intermittent mobile phone three axis acceleration signal of pulse noise, and combined with the direction of gravity sensor component containing a mobile phone three axis acceleration in filtration; then, design a fitness running evolutionary modeling algorithm to increase the search space to provide support for obtaining better fitness running model the experimental results show that this method can obtain more accurate speed and fitness running model than the Brzostowski method.
【作者单位】: 福建师范大学软件学院;武汉纺织大学数学与计算机学院;福建师范大学体育科学学院;
【基金】:国家自然科学基金项目(61305079,61370078) 福建省自然科学基金项目(2015J01235,2017J01498) 福建省教育厅JK类项目(JK2015006) 武汉大学软件工程国家重点实验室开放基金项目(SKLSE 2014-10-02)
【分类号】:TN929.53;TP274.2
,
本文编号:1575612
本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1575612.html