Research On Detection Method Of Fall Based On Multi Sensor Joint

Today,China has entered the ranks of aging countries,the elderly pension problem has become a hot issue in today’s society,has been widespread concern in the community.The fall is a frequent occurrence in the elderly group of events,the incidence is extremely high,and the fall for the elderly is a great harm,and even life-threatening,according to statistics,leading to death among the many reasons the elderly,falls rank very before.However,more and more "empty-nesters" and some social problems make the old people often can not get timely assistance after the fall.Thus,there is a great need for a remote alarm system capable of targeting fall behavior.With the development of embedded technology,communication technology,network technology and computer technology,remote communication technology is very mature,but the difficulty lies in the detection of falling behavior.Based on this background,this paper studies the wearable fall detection method.According to the characteristics of fall behavior,falls were divided into rapid falls and slow falls or severe abnormal falls,and the detection method based on multi-sensor combination was designed for the fall behaviors.And the fall test platform is designed to validate the fall detection method designed in this paper.The main contents of this paper are as follows:(1)To analyze the daily behavior and fall behavior of the human body,analyze the characteristics of fall behavior,according to fall in different ways fall behavior is divided into rapid fall and slow fall two categories,and fall for the two types of detection methods.(2)To study the changes of physical information parameters during human activities,mainly to study the changes of acceleration and plantar pressure.According to their changing characteristics,a series of characteristic variables were selected to describe the human behavior.A Fast Fall Detection Method Based on Triaxial Acceleration Sensor and Foot Pressure Sensor.(3)The characteristics of slow falls were studied and analyzed,and a new detection method for slow fall behavior or serious abnormal fall behavior was designed according to its characteristics.(4)By using the embedded technology,the control technology and the sensor technology,a set of design and design experiments for the fall test platform,including the hardware and software,is designed according to the detection method proposed in this paper.And the feasibility of the fast fall detection method proposed in this paper is validated by the system.The experimental results show that the method of fall detection has high accuracy and can control the occurrence of false alarm effectively.It can be used to in daily life,especially for the elderly to detect the falls event.