Research And Implementation Of Elderly Fall Detection Monitoring System Based On Multi-sensors

China's population aging situation is serious. It presents the three features of the accelerated aging population, old aging and empty-nested elderly people. It makes social medical and social insurance system face huge pressure, and elders nursing care issues are increasingly become a significant social problem. Falling is the leading cause of older accidental injury. And it is likely to lead to fall when accidental slip or some disease attack suddenly and deterioration. If not promptly timely rescue, delay salvage timing when the fall happened will be life-threatening. Therefore, the development designed for elderly service products—elderly fall monitoring system which can accurately differentiate between fall events and active daily life events, and can timely alarm and contact ambulance in the cases of not affect older people's normal life, it has important role in improving the life quality of elderly and promoting the stability of our society.Real-time ambulatory monitoring system in fall detection for elderly people is described.Embedded processor MSP430F169 is the processing core in this system. We used three-axis accelerometer to get the acceleration information, magnetic sensors to obtain angle information on the human body, pressure sensor to collect pressure data underfoot and then judged fall by the threshold algorithm. Completing the waist data acquisition module and the foot data acquisition module's hardware circuit design based on understanding of the sensor's the buffer circuit. The system is small and low power consumption. It meets people wear habits because it is dressed in the waist belt and is embedded in the insole. We finished lots of fall events and daily life events experiment. We also used MATLAB tools to analyze the collected data, determined the acceleration sensor SMV threshold, inclination angle threshold and pressure threshold, completed threshold algorithm based on multi-sensor information, and finished lower level computer programming in EW4.21 IAR development environments. We completed the buttons interactive function which can cancelle alarm. Finished debugging interface based on Teechart, GPRS send text message for contact rescue, voice alarm according to the different events and so on in VC++ 6.0 development environment. Software realized real-time communication function and defined each other's communication protocols and then debugged communications to normal.Finally, we completed the prototype of Elderly Fall Detection Monitoring System.We designed two kinds of experimental in order to validate the system and algorithm. Firstly, examine the recognition of fall judgment through all sorts of real fall. Secondly, detect the misstatement rate through the normal daily activities events of this system. These experiments show that the system have high reliability and accuracy. Finished application of prototype in the old man service robot system in the last of this project, and the results show that the system has good adaptability and portability.