基于加速度传感器的无线跌倒检测系统  

Wireless Fall Detection System Based on Acceleration Sensor

作　　者：秦昉[1] 孙子文[1] 白勇 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122 [2]无锡宏创电子有限公司,江苏无锡214072

出　　处：《控制工程》2016年第1期124-128,共5页Control Engineering of China

基　　金：江苏省自然科学基金(Bk20131107);国家自然科学基金(61373126)

摘　　要：针对老年人跌倒检测问题,设计了一种基于加速度传感器的跌倒检测装置。装置安置于腰上采集加速度传感器数据,提取对人体运动具有较好区分度的加速度特征和躯干角度特征并对特征设定合理的阈值,建立基于多重阈值的人体跌倒检测算法,区分跌倒和人体日常生活活动。在跌倒时利用GPRS自动发出报警信息,该装置具有便携性、实时性等特点。仿真及实验表明:系统能有效识别出跌倒和日常行为,算法具有较高实时性、灵敏度和特异度。A fall detection device based on acceleration sensor is designed for monitoring the elderly fall. The fall detection algorithm is proposed based on the acceleration feature and torso angle feature which have good differentiation of human movement by analyzing the acceleration data captured from the acceleration sensor placed on the waist, then the fall detection algorithm based on multiple thresholds is established to distinguish the body fall and human activities of daily living. The device will automatically send alarm information by using GPRS when a fall is detected. The device has the characteristics of portability and real-time. Simulative and experimental results show that the system can achieve high accuracy and reliability.

关 键 词：加速度传感器 跌倒检测 阈值算法 老年人 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]