基于贝叶斯分类算法的电子呼啦圈设计

Design of electronic Hula hoop based on Bayes classification algorithm

针对实体的呼啦圈体积大,不易携带,使用不当可能对身体造成伤害的问题,设计了一种基于贝叶斯分类算法的电子呼啦圈。系统使用三轴加速度计采集呼啦圈旋转时腰部的加速度信号,经卡尔曼滤波后,提取特定周期内的波形特征并以此训练贝叶斯分类器,利用训练后的分类器对实时滑动窗口内加速度计数据进行判断是否符合转实体呼啦圈腰部的运动特征,并将分类结果通过蓝牙发送至Android终端。实验证明:该设计功耗低,实时性好,准确率高。

Aiming at problem that Hula hoop is bulky,hard to carry,and improper usage may do harm to the body,an electronic Hula hoop based on Bayes classification algorithm is designed. A three-axis accelerometer is used to capture acceleration signal at waist when tester doing hula hoop exercise,train the Naive Bayesian classifier by the wave characteristics within a given period after Kalman filtering ,the trained classifier is carried out to analyze the real-time data in sliding window,then send the result to android terminal by Bluetooth. The results prove that the system is precision with low-power consumption,good real-time.