Research On Fall Protection Strategies Based On Ambulatory Gait And Mobility Tracking With Wearable Sensors

Fall is a sudden dangerous behavior common in people’s daily life, and thepeople generally unable to respond in a timely manner. Ageing trend is exacerbated intoday’s society, for the large ageing population, the most salient problems faced is thefall which caused by lower extremity motor dysfunction, it has become one of themain causes of accidents of the elderly population in China. Therefore, the fallprediction and auxiliary protection during the human body movement becomeproblems to be solved. With the rapid development of sensor technology, theapplication of wearable fall detection device to solve the fall issues has great researchvalue.Based on the situation above, this article aims to study a human lower limbposture detection method with sensors. On the one hand, we can collect human lowerlimb movement gesture parameters, to distinguish normal walking and fall conditions,on the other hand, we can also use the alert or brake to prejudge or prevent theoccurrence of the fall, to help lower limb impaired elderly people and people duringthe lower extremity rehabilitation, then ensure the life and health.Research contents are divided into four parts, the second chapter describes a newalgorithm applied to the detection of Lower Extremity motion signal, that is thecoordinate transformation method for three-dimensional rotation matrix, andcalculated with the Matlab programming to get the posture parameters of the lowerlimbs. The third chapter describes hardware`s selection and structures, the mostimportant part “sensor part” use MAG3composite sensors, which mounted on thesimple retractable slide lower limb exoskeleton, the entire apparatus is worn on thedetection object`s lower limb, during walking, the sensors collect acceleration signaland magnetic signal, the data after A/D conversion, transmitted through the wirelessmodule to the computer, and the computer uses Matlab programming for dataprocessing. The fourth chapter which also the experimental part is mainly to verify theaccuracy of signal detection, we do the experiment compare with the three-dimensional image acquisition and analysis system, and the data accuracy rate is more than90%. In the last part of the article we put forward Fall prevention strategies basedon lower limb posture detection, introduce the theory of stability of humanoid robot,that is zero moment point theory to judge walking stability, and to predict whetherthere is a tendency to fall, and we initially demonstrate the feasibility.