Research On A High Effectiveness Fall Detection System Based On Low-power Bluetooth

Owing to the deepening of the aging of population and the increasing of the number of the disability and the empty nest elderly,the need for health and safe care for the elderly is increasing.Fall is one of the most important factors that threaten the health and the safety of the elderly over 60 years old,and will be directly or indirectly leads to the disabled,and even death.Fall activities can seriously affects the elderly’s mental health,declining the quality of life.It is the research hotspot that effectively detect the falls so as to reduce the impacts of falls on the elderly.Most existing researches on wearable fall detection system adopts single sensor fall detect mode.This mode gets low accuracy of fall detection and is very easy to make mistakes with activities of daily living(ADL).Furthermore,it produces truly high fall false rate and gets poor user experience.Presently,difficulties of the wearable fall detection system research attribute to reducing device size,cutting down power consumption,raising accuracy of fall detection and decreasing fall false rate.Therefore,the aim of this paper is to design a wearable fall detection system with the merit of small size,low system cost,low power consumption,high accuracy of fall detection and low fall false rate.Using low-power Bluetooth technology combined with the pressure sensor and the.tri-axial accelerometer to realize the capability of classifying falls and ADL in real-time.Firstly,to realize the design above,this research completes the hardware design of the beneath-foot pressure detection module based on pressure sensor and the waist-used posture detection module based on the tri-axial accelerometer.The hardware of the system is tested as well.Secondly,finishing the fall detection algorithm and the design of system software.Lastly,the feasibility and effectiveness of the algorithm and software are verified by correlative experiments,and the results show that the designed system is stable and reliable,which can realize the design intention.The system measurement of power consumption show that the average current consumption of the modules is very low in standby status.In this condition,a coin cell can afford two modules several months.The result meets the requirement of low power consumption.Finally,the performance of the fall detection system is tested by simulating falls and ADL experiments in laboratory environment.Results show that the system has 100%ADLdetection rate,0%fall false rate and 93%fall detection rate,which satisfies the design requirement of high fall detection rate and low fall false rate.