基于分散熵和支持向量机的运动状态识别

Motion state recognition based on dispersion entropy and support vector machine

针对传统的运动状态识别方法不能对非跌倒运动进行具体区分的问题,提出了一种结合了分散熵和支持向量机的运动状态识别方法。首先计算加速度传感器ADXL345的三轴加速度数据的均方根作为合成加速度,然后求解合成加速度的分散熵,最后将分散熵输入支持向量机进行识别和分类。选取SisFall数据集的跌倒、走路和上/下楼3类运动状态的加速度信号,对该方法的性能进行评估。实验结果表明,该方法比排列熵方法的分类精度提高了15.5%。该方法不仅具有较高的分类精度,而且明显提高了计算效率,可以更好地识别人体的运动状态。

Aiming at the problem that traditional motion state recognition methods can only distinguish fall and non-fall,a novel motion state recognition method based on dispersion entropy and support vector machine is proposed.First,the root mean square(RMS)of the three-axis acceleration data of ADXL345 is calculated as the combined acceleration.Then,the dispersion entropy of the combined acceleration is calculated.Finally,the feature vectors are sent into the support vector machine to classify and recognize different types of motion state recognition.In this paper,the acceleration signals of the three types of motion states(falling,walking and up/down stair)of the SisFall database are selected to evaluate the performance of the method.The experimental results show that the proposed method improves the classification accuracy by 15.5%compared with permutation entropy method.The method not only has a higher classification accuracy,but also significantly improves the calculation efficiency and can better recognize motion states of the human body.