文章摘要

文章信息/Info

Title:: Design of body condition monitoring device for the elderly based on inertial sensors

Author(s):: WU Cong; ZHANG Kun; YANG Licai; School of Control Science & Engineering, Shandong University, Jinan 250061, China

Keywords:: Fall detection; Posture calculation; Posture recognition algorithm; Inertia sensor; Wearable monitoring; Remote monitor

摘要:: 为解决人口老龄化社会所带来的老人跌倒事件频发，减轻跌倒等异常状态对老年人的伤害，研发了一种基于惯性传感器的穿戴式姿态监测装置。该装置基于姿态解算特性，通过采集三轴加速度计、三轴陀螺仪和三轴磁力计的数据,以多种运动特征为依据进行姿态判别，不仅可以检测跌倒状态，还可对老年人的日常姿态及位置信息进行连续监测并进行步数统计，利用GPRS网络向服务器上报数据。装置中加入了语音推送功能，可对服务器端推送的天气信息、问候信息等进行语音播报。试验结果表明，该装置可对佩戴者的姿态及位置信息进行连续监测，当出现跌倒等异常状态时，可实现实时报警及对事故发生地的准确定位，姿态识别的准确率不低于98%。

Abstract:: In order to solve the problem of frequent falls in elderly caused by aging society and reduce the harm led by abnormal state such as falls, a wearable posture monitoring device based on inertial sensors has been designed. The device, based on the characteristics of posture calculation, could discriminate current posture by analyzing multiple motion features acquired through three-axis accelerometer, three-axis gyroscope and three-axis magnetometer. It not only could detect fallen state,but also could monitor daily posture and position information of the wearer, count the number of footsteps and report data to the server through the GPRS network. The device was also equipped with vocal broadcasting function, which could broadcast weather information and regards came from server. The test results show that the device can detect abnormal situations such as falls and send out alarms to concerned party instantly. The posture recognition accuracy is higher than 98%.