Research on Fall Detection System of Wearable Devices Based on ELM Algorithm
Xiaolei Wang, Donghao Li, Xiaowan Zheng, Shuangjian Yan, Qingfang Zhang, Jitao Zhang, Lingzhi Cao · 2019 IEEE 3rd Information Technology, Networking, Electronic and Automation Control Conference (ITNEC) · 2019
The aging phenomenon in current society is getting more and more serious, and the health status of the elderly is facing challenges. A fracture or other disease caused by a fall, or a fall due to a sudden illness, may result in serious consequences if not treated in time. According to the difference between fall and normal activities, this paper proposes a method for detecting human fall by ELM based on inertial measurement unit. By placing the inertial measurement unit on the human body, the human acceleration and angular velocity data are sampled and the human body attitude is calculated. The ELM training is performed on the human characteristic data to obtain the ELM classification function, and the measurement data is classified by the ELM classification function to determine whether it is in a falling state. The human fall experiment is carried out, and the result shows that the ELM method can accurately detect the fall state of the human. The use and promotion of this method is of great significance for the treatment of falls and reduction of casualties.