CN-122004843-A - Identification and early warning method and system for abnormal behaviors of old people

CN122004843ACN 122004843 ACN122004843 ACN 122004843ACN-122004843-A

Abstract

The invention is suitable for the technical field of monitoring and early warning, and provides a method and a system for identifying and early warning abnormal behaviors of old people, comprising the following steps of collecting multi-modal behavior data, and analyzing the multi-modal behavior data to obtain multi-dimensional behavior characteristics, wherein the multi-dimensional behavior characteristics comprise night path characteristics, social interaction characteristics and feeding movement characteristics; the method comprises the steps of analyzing night path characteristics, social interaction characteristics and feeding movement characteristics, determining night risk scores, social risk scores and feeding risk scores, generating early warning levels of each dimension according to the night risk scores, the social risk scores and the feeding risk scores, wherein the early warning levels comprise attention levels, intervention levels and alarm levels, and generating an intervention scheme for the early warning levels of the intervention levels. The invention can comprehensively monitor the night path, social interaction and eating movement behavior information of the old, discover subtle changes and sudden abnormal conditions in the behavior of the old in time, and improve the real-time performance and accuracy of monitoring.

Inventors

LI YUNNA
ZHANG YAN
WU DONGYING

Assignees

中国人民解放军海军青岛特勤疗养中心

Dates

Publication Date: 20260512
Application Date: 20260306

Claims (10)

1. The method for identifying and early warning the abnormal behaviors of the old is characterized by comprising the following steps of: Collecting multi-modal behavior data, and analyzing the multi-modal behavior data to obtain multi-dimensional behavior characteristics, wherein the multi-dimensional behavior characteristics comprise night path characteristics, social interaction characteristics and eating movement characteristics; Analyzing night path characteristics, social interaction characteristics and feeding movement characteristics, and determining night risk scores, social risk scores and feeding risk scores; Generating an early warning level of each dimension according to the night risk score, the social risk score and the feeding risk score, wherein the early warning level comprises a concern level, an intervention level and an alarm level; generating an intervention scheme for the early warning level of the intervention level, reminding a guardian to carry out remote accompany when the night risk or the feeding risk reaches the intervention level, and reminding the guardian to search for an active contact person and a social activity when the social risk reaches the intervention level.
2. The method for identifying and warning abnormal behaviors of the elderly according to claim 1, wherein the step of collecting multi-modal behavior data and analyzing the multi-modal behavior data to obtain multi-dimensional behavior features specifically comprises: Determining night activities through the cooperative work of the mattress pressure sensor and the infrared motion sensor to obtain a bed leaving time period, and generating a moving track according to the triggering sequence of the infrared motion sensor to obtain night path characteristics; Collecting call records and household monitoring data, and determining initiative ratio, interaction time length, contact breadth, outgoing time length and visitor time length to obtain social interaction characteristics; The inertial measurement sensor on the tableware is used for collecting hand movement information, determining feeding speed, movement smoothness and movement intermittence, and obtaining feeding movement characteristics.
3. The method for identifying and pre-warning abnormal behaviors of elderly people according to claim 2, wherein the step of determining the initiative ratio, the interaction time length, the contact breadth, the outgoing time length and the visitor time length specifically comprises the following steps: Obtaining an initiative ratio according to the ratio of dialing and answering calls in the call records, obtaining interaction time according to the call time, and obtaining the contact breadth according to the number of the non-repeated contacts; determining an old person exit time stamp and an old person entrance time stamp according to the entrance monitoring data, and summarizing to obtain an outgoing time length; and determining a visitor entry time stamp and a visitor exit time stamp according to the entry monitoring data, and summarizing to obtain visitor duration.
4. The method for identifying and warning abnormal behaviors of the elderly according to claim 2, wherein the step of determining the feeding rate, the movement smoothness and the movement intermittence specifically comprises the steps of: Determining the cycle times of the tableware from a low position to a high position in unit time according to the hand motion information to obtain the feeding rate; calculating the smoothness of each acceleration curve from a low position to a high position to obtain the movement smoothness; The pause time between two adjacent feeding actions from the low position to the high position is calculated, and the movement pause is determined.
5. The method for identifying and warning abnormal behaviors of elderly people according to claim 2, wherein the determining the night risk score, the social risk score and the feeding risk score specifically comprises: judging whether the time stamp of the moving track is matched with the bed leaving time period, and determining that the activity mode is normal night or abnormal loitering according to the moving track when the time stamp of the moving track is matched with the bed leaving time period; Comparing each social factor in the social interaction characteristics with a corresponding reference range, determining the degree of decline of each social factor, and carrying out weighted summation on all the degrees of decline to obtain social risk score; comparing each eating factor in the eating exercise profile with a corresponding reference range, determining the degree of deviation of each eating factor, and weighting and summing all the degrees of deviation to obtain an eating risk score.
6. The utility model provides an old person's unusual action discernment early warning system which characterized in that, the system includes: The multi-dimensional behavior feature module is used for collecting multi-modal behavior data and analyzing the multi-modal behavior data to obtain multi-dimensional behavior features, wherein the multi-dimensional behavior features comprise night path features, social interaction features and eating movement features; the multidimensional risk assessment module is used for analyzing night path characteristics, social interaction characteristics and feeding movement characteristics and determining night risk scores, social risk scores and feeding risk scores; The early warning level determining module is used for generating early warning levels of each dimension according to the night risk score, the social risk score and the feeding risk score, wherein the early warning levels comprise a concern level, an intervention level and an alarm level; The intervention scheme generation module is used for generating an intervention scheme for the early warning level of the intervention level, reminding the guardian of carrying out remote accompany when the old starts at night or eats when the night risk or the eating risk reaches the intervention level, and reminding the guardian of searching for the active contact person and the social activity when the social risk reaches the intervention level.
7. The system for identifying and pre-warning abnormal behavior of elderly people according to claim 6, wherein the multi-dimensional behavior feature module comprises: The night path characteristic unit is used for determining night activities through the cooperative work of the mattress pressure sensor and the infrared motion sensor to obtain a bed leaving time period, and generating a moving track according to the triggering sequence of the infrared motion sensor to obtain night path characteristics; The social interaction feature unit is used for collecting call records and household monitoring data, determining initiative ratio, interaction time length, contact breadth, outgoing time length and visitor time length and obtaining social interaction features; And the feeding motion characteristic unit is used for acquiring hand motion information through an inertial measurement sensor on the tableware, determining feeding speed, motion smoothness and motion intermittence and obtaining feeding motion characteristics.
8. The system for identifying and pre-warning abnormal behavior of elderly people according to claim 7, wherein the social interaction feature unit comprises: The call condition subunit is used for obtaining an initiative ratio according to the ratio of dialing and answering calls in the call record, obtaining an interaction time length according to the call time length and obtaining a contact breadth according to the number of non-repeated contacts; The outgoing time length subunit is used for determining an old person outgoing time stamp and an old person incoming time stamp according to the incoming monitoring data, and summarizing to obtain outgoing time length; And the visitor duration subunit is used for determining a visitor entry time stamp and a visitor exit time stamp according to the entry monitoring data and summarizing to obtain the visitor duration.
9. The system for identifying and warning abnormal behavior of elderly people according to claim 7, wherein the eating exercise characteristic unit comprises: The eating rate subunit is used for determining the cycle times of the tableware from the low position to the high position in unit time according to the hand motion information to obtain the eating rate; The motion fluency subunit is used for calculating the smoothness of each acceleration curve from a low position to a high position to obtain the motion fluency; And the motion intermittence subunit is used for calculating the pause time between two adjacent feeding actions from the low position to the high position and determining motion intermittence.
10. The system for identifying and pre-warning abnormal behavior of elderly people according to claim 7, wherein the multi-dimensional risk assessment module comprises: The night risk scoring unit is used for judging whether the time stamp of the moving track is matched with the bed leaving time period, and determining that the activity mode is normal night or abnormal loiter according to the moving track when the time stamp of the moving track is matched with the bed leaving time period; the social risk scoring unit is used for comparing each social factor in the social interaction characteristics with a corresponding reference range, determining the descending degree of each social factor, and carrying out weighted summation on all the descending degrees to obtain social risk scores; and the feeding risk scoring unit is used for comparing each feeding factor in the feeding movement characteristics with the corresponding reference range, determining the deviation degree of each feeding factor, and carrying out weighted summation on all the deviation degrees to obtain the feeding risk score.

Description

Identification and early warning method and system for abnormal behaviors of old people Technical Field The invention relates to the technical field of monitoring and early warning, in particular to a method and a system for identifying and early warning abnormal behaviors of old people. Background With the acceleration of the aging process of society, the population of the aged is continuously increased, and the health and safety problems of the aged are increasingly focused by various communities of society. The elderly are prone to various abnormal behaviors during the progressive decline of physiological functions, which are often closely related to potential health risks, psychological problems, or life dilemmas. At present, monitoring and early warning of abnormal behaviors of the elderly mainly rely on manual observation and regular health examination. The manual observation mode not only consumes a large amount of manpower and material resources, but also is difficult to realize real-time and comprehensive monitoring, and some fine but key abnormal behavior signals are easy to miss. Although some potential health problems can be found in regular health examination, the examination interval time is long, and sudden abnormal conditions in the daily behaviors of the old cannot be captured in time, so that the intervention opportunity is delayed. Therefore, it is necessary to provide a method and a system for identifying and early warning abnormal behaviors of the elderly, which aim to solve the above problems. Disclosure of Invention Aiming at the defects of the prior art, the invention aims to provide a method and a system for identifying and early warning the abnormal behaviors of the aged, so as to solve the problems in the prior art. The invention discloses a method for identifying and early warning abnormal behaviors of old people, which comprises the following steps: Collecting multi-modal behavior data, and analyzing the multi-modal behavior data to obtain multi-dimensional behavior characteristics, wherein the multi-dimensional behavior characteristics comprise night path characteristics, social interaction characteristics and eating movement characteristics; Analyzing night path characteristics, social interaction characteristics and feeding movement characteristics, and determining night risk scores, social risk scores and feeding risk scores; Generating an early warning level of each dimension according to the night risk score, the social risk score and the feeding risk score, wherein the early warning level comprises a concern level, an intervention level and an alarm level; generating an intervention scheme for the early warning level of the intervention level, reminding a guardian to carry out remote accompany when the night risk or the feeding risk reaches the intervention level, and reminding the guardian to search for an active contact person and a social activity when the social risk reaches the intervention level. As a further scheme of the invention, the steps for acquiring the multi-mode behavior data and analyzing the multi-mode behavior data to obtain the multi-dimensional behavior characteristics specifically comprise the following steps: Determining night activities through the cooperative work of the mattress pressure sensor and the infrared motion sensor to obtain a bed leaving time period, and generating a moving track according to the triggering sequence of the infrared motion sensor to obtain night path characteristics; Collecting call records and household monitoring data, and determining initiative ratio, interaction time length, contact breadth, outgoing time length and visitor time length to obtain social interaction characteristics; The inertial measurement sensor on the tableware is used for collecting hand movement information, determining feeding speed, movement smoothness and movement intermittence, and obtaining feeding movement characteristics. As a further aspect of the present invention, the step of determining the initiative ratio, the interaction time length, the contact breadth, the outgoing time length and the visitor time length specifically includes: Obtaining an initiative ratio according to the ratio of dialing and answering calls in the call records, obtaining interaction time according to the call time, and obtaining the contact breadth according to the number of the non-repeated contacts; determining an old person exit time stamp and an old person entrance time stamp according to the entrance monitoring data, and summarizing to obtain an outgoing time length; and determining a visitor entry time stamp and a visitor exit time stamp according to the entry monitoring data, and summarizing to obtain visitor duration. As a further aspect of the present invention, the step of determining the feeding rate, the movement smoothness and the movement intermittence specifically includes: Determining the cycle times of the tableware from a low position to a high