Safety Monitoring System For Elderly People Living Alone Based On Abnormal Behavior Detection

With the increasing degree of aging in our country,the home safety of the livingalone elderly has gradually become a social problem that needs urgent attention and solution.Due to their poor physical fitness and lack of care,the living-alone elderly are prone to accidents such as falls or sudden illness at home.It is difficult to detect such incidents timely,which poses a huge threat to their safety.In order to solve this issue,the real-time monitoring technology of the home safety of the living-alone elderly has become one of the hot research topics at present.Some of the current researches have some shortcomings in practical applications,such as wearable devices requiring frequent charging and easily forgotten to wear,video surveillance devices having privacy issues as well as blind spots,and some fixed devices unable to monitor sleep conditions at night.Therefore,it is a research topic with important application value to study the monitoring technology that can avoid or solve the above problems.Based on this background,this study designs and implements a home safety monitoring system for the living-alone elderly based on sensor technologies and anomaly detection technologies to achieve imperceptible and round-the-clock monitoring.The main tasks of this study are:1.A framework of a monitoring system was proposed,which achieves the detection of abnormal behaviors of the living-alone elderly at home.The home activities of the elderly are mainly divided into three aspects: indoor walking behavior,indoor affairs behavior and sleep-related behaviors.Therefore,a smart electricity meter,infrared sensors,and a sleep monitoring belt based on a piezoelectric film sensor were selected to design and develop sensor nodes,and together with the communication equipment to form the home safety monitoring system for the living-alone elderly,which is used to collect data related to the above behaviors of the elderly in real time.2.An algorithm for detecting abnormal activities of the living-alone elderly was developed.Using the electric meter data and infrared sensing data,a method for abnormal activities detection was proposed,which can automatically detect the abnormal activities of the elderly indoors.Using the sleep monitoring data based on the piezoelectric sensor,a multi-aspect sleep monitoring method was proposed,which can obtain multiple information such as time of going to bed,time of getting up,time of leaving bed at night,physical activities during sleep and heart rate during sleep of the elderly,so as to detect related anomalies.In terms of calculating heart rate during sleep,an improved method for calculating heart rate was proposed.It can obtain more accurate results in the case of poor signal quality.3.A data visualization page was developed to display recent monitoring data and processing results.In addition,the functionality and performance of the system were tested.The experimental results show that the system can effectively improve the home security of the living-alone elderly,and has certain practicability and promotional value.In the future,efforts will be made to improve and perfect the current system based on its shortcomings.Further exploration will be conducted on how to expand its functionality to better meet the needs of the living-alone elderly at home.