Research On Key Technology Of Abnormal Behavior Detection Of Elderly Living Alone Based On Location Information

With the change of the family structure and the deepening of the aging population structure,the issue of old-age care for the elderly living alone in the vulnerable groups has made China's future development facing severe challenges.In order to actively cope with the aging population and solve the pension problem of the elderly living alone,the family-based home-based pension model promoted by the government is an ideal pension service model at present and in a certain period in the future.Due to the decline of the physical function of the elderly and the impact of a variety of chronic diseases,it is of great social significance to improve the safety monitoring of the daily activities of the elderly living alone indoors,reduce the accidental risk of indoor activities,and build a safe and intelligent home care service environment.This thesis takes the abnormal behavior detection of the elderly living alone in the home care service environment as the research background.Based on the trajectory of the daily behavior of the elderly living alone,this thesis conducts the following research work:(1)This thesis first summarizes the research status of the existing elderly living system for the elderly,and analyzes and compares the indoor positioning methods of the elderly living alone.This thesis adopts the Bluetooth indoor positioning system based on low power consumption,low cost and high precision.Real-time location data is collected and stored for indoor activities of elderly people living alone.(2)Aiming at the unpredictability and uncertainty of the movement state of elderly people living alone in the indoor space environment,this thesis adopts the kalman filter tracking algorithm based on interactive multi-model to conduct the filtering and tracking of various movement states in the dailybehaviors and activities of elderly people living alone.The simulation results show that the filter tracking effect of this algorithm is better than that of single mode filter.(3)Considering the difference of individual living alone,the time points of activities in various indoor areas and the length of staying time in each spatial area are different.This thesis proposes a method for extracting the length of stay time based on the 3? criterion.Through the analysis of the length of the historical residence time of different spatial positions of the elderly living alone,the length of stay is subject to the Gaussian normal distribution model;the characteristics of Gaussian normal distribution are combined,and the residence time threshold of the normal behavior of indoor space in the elderly living alone is established based on the 3? criterion.method of extraction.Then,the abnormal state detection is performed by judging whether the elderly living alone in a certain spatial position in the room and the length of the staying time in the spatial location area exceed the length of the staying time,thereby realizing remote monitoring and emergency alarm for the elderly living alone.