CN-121981696-A - Household-based care calendar management method, device, medium and equipment

CN121981696ACN 121981696 ACN121981696 ACN 121981696ACN-121981696-A

Abstract

The application discloses a schedule management method, a device, a medium and equipment for home care, which relate to the technical field of health management and can be applied to medical health business scenes, and the method comprises the steps of collecting multi-mode data of a user in the home care scene in real time based on a multi-equipment linkage mode, wherein the multi-mode data at least comprises health index data and behavior data; the method comprises the steps of carrying out data fusion processing on multi-mode data to generate a standardized data set, calling an AI large model to analyze association rules between behavior data and health index data in the data set, generating an initial personalized schedule based on the association rules, monitoring the health index data in real time through an anomaly identification algorithm, and dynamically adjusting the initial personalized schedule based on the anomaly type of the health index and a preset schedule adjustment strategy database to generate a target personalized schedule when the health index is determined to be anomalous. The application can solve the problems that the related technology has low intelligent degree and can not dynamically adjust the schedule based on the health state.

Inventors

WU XIAO
WANG YANQI

Assignees

平安健康互联网股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. The schedule management method for home care is characterized by comprising the following steps: Acquiring multi-mode data of a user in a home care scene in real time based on a multi-device linkage mode, wherein the multi-mode data at least comprises health index data and behavior data, and the behavior data comprises schedule operation data, voice interaction data, behavior gesture data and sleep state data of terminal equipment; carrying out data fusion processing on the multi-mode data to generate a standardized data set; invoking an AI large model to analyze the association rule between the behavior data and the health index data in the data set, and generating an initial personalized schedule based on the association rule; and monitoring the health index data in real time through an abnormality identification algorithm, and dynamically adjusting the initial personalized schedule based on the abnormal type of the health index and a preset schedule adjustment strategy database to generate a target personalized schedule when the health index is abnormal.
2. The method of claim 1, wherein performing data fusion processing on the multi-modal data to generate a normalized data set comprises: Respectively executing analysis and cleaning operations on the health index data and the behavior data, and eliminating invalid redundant data; The health index units and the timestamp formats of different devices are subjected to standardized conversion so as to unify data format standards; And storing the health index data and the behavior data after the analysis and cleaning operation and the standardized conversion according to the time dimension association to form a standardized data set.
3. The method of claim 1, wherein the invoking the AI large model to analyze a rule of association between behavior data and health indicator data in the data set, generating an initial personalized schedule based on the rule of association, comprises: Inputting the data set to an AI large model based on a transducer architecture and with a self-attention mechanism; Calculating characteristic association weights of the daily operation data, the voice interaction data, the behavior gesture data, the sleep state data and the health index data in the data set respectively through the self-attention mechanism, and mining association rules of the behavior data and the health index data; and generating an initial personalized schedule adapting to the behavior habit and the health state of the user based on the association rule, the user demand reflected by the voice interaction data and the work-rest habit reflected by the sleep state data.
4. The method according to claim 1, wherein the monitoring the health index data in real time by the anomaly identification algorithm dynamically adjusts the initial personalized schedule based on the health index anomaly type and a preset schedule adjustment policy database when the health index anomaly is determined, and generating a target personalized schedule includes: comparing the health index data with a preset health threshold value through an abnormality identification algorithm to determine the abnormality type of the health index; A preset schedule adjustment strategy database is called, and a target schedule adjustment strategy corresponding to the abnormal type of the health index is matched, wherein schedule adjustment strategies corresponding to different abnormal types of the health index are stored in the preset schedule adjustment strategy database, and the schedule adjustment strategies comprise schedule pause, schedule addition, time adjustment and content replacement rules; and performing content adjustment on the initial personalized schedule by executing the target schedule adjustment strategy to generate a target personalized schedule.
5. The method according to claim 1, wherein the method further comprises: Collecting historical behavior data of the user and schedule data corresponding to similar user groups; Mining a potential demand schedule of the user through a collaborative filtering algorithm based on the historical behavior data and schedule data of the similar user group; and outputting schedule recommendation information corresponding to the potential demand schedule, so that the user triggers generation of an adjustment instruction for the target personalized schedule based on the schedule recommendation information.
6. The method according to claim 1, wherein the method further comprises: Based on the user position data, the intelligent equipment state data, the behavior gesture data and the voice interaction data, performing scene recognition, and judging the current scene of the user; a preset equipment combination strategy library is called, and reminding equipment combinations corresponding to the current scene of the user are matched, wherein preset reminding equipment combinations corresponding to different scenes are configured in the preset equipment combination strategy library; Triggering the reminding equipment to perform initial reminding operation on the user in a combined way when judging that the execution time node of the target personalized schedule is reached, and monitoring a user response state through an equipment sensor; if the user response is not monitored, the reminding mode is updated according to a preset gradient until the user response feedback is received.
7. The method according to claim 1, wherein the method further comprises: constructing a daily schedule base line library based on the historical schedule operation data and the historical sleep state data of the user, wherein the daily schedule base line library is used for recording a fixed schedule rule formed by the user for a long time; Comparing the execution condition of the target personalized schedule with the daily schedule baseline library in real time, and identifying abnormal conditions which are not executed according to the baseline and have no manual adjustment record; If the target personalized schedule is identified to be abnormal, judging the security risk level of the user by combining the health index data and the behavior gesture data; If the security risk level is judged to be a high-risk, automatically generating an emergency schedule containing rescue guidelines; and synchronously pushing the emergency schedule to a user emergency contact terminal and a community rescue terminal.
8. A home-oriented calendar management device, comprising: The system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring multi-mode data of a user in a home care scene in real time based on a multi-equipment linkage mode, the multi-mode data at least comprises health index data and behavior data, and the behavior data comprises schedule operation data, voice interaction data, behavior gesture data and sleep state data of terminal equipment; The processing module is used for carrying out data fusion processing on the multi-mode data to generate a standardized data set; the analysis module is used for calling an AI large model to analyze the association rule between the behavior data and the health index data in the data set and generating an initial personalized schedule based on the association rule; The generation module is used for monitoring the health index data in real time through an abnormality recognition algorithm, and dynamically adjusting the initial personalized schedule based on the abnormal type of the health index and a preset schedule adjustment strategy database to generate a target personalized schedule when the health index is judged to be abnormal.
9. A storage medium having stored thereon a computer program, which when executed by a processor, implements the method of any of claims 1 to 7.
10. An electronic device comprising a storage medium, a processor and a computer program stored on the storage medium and executable on the processor, characterized in that the processor implements the method of any one of claims 1 to 7 when executing the computer program.

Description

Household-based care calendar management method, device, medium and equipment Technical Field The application relates to the technical field of health management, in particular to a schedule management method, device, medium and equipment for home-based care. Background With the deepening of the aging degree of population, home care has become one of the main current care modes in China. The scale of solitary old people and chronic disease old people is continuously enlarged, and the demands of the population on life schedule planning, health state monitoring and safety risk early warning are increasingly urgent. Under the background, the pension auxiliary system integrating intelligent equipment and algorithm technology gradually becomes an industry research and development hot spot, and the core objective of the related technology is to provide convenient and safe life and health management service for the elderly users through data acquisition and intelligent analysis. Related technologies and products currently facing home care scenes in the industry are mainly divided into two categories. The first category is home care content pushing technology or product, and the core function of the technology or product is to collect user behavior data and health data, generate user portraits, predict the demands of users on the content and further adjust the content pushing priority and display form. The product comprises the functions of the part of intelligent televisions and the intelligent sound boxes for pushing the content in the pension mode, and focuses on 'information distribution optimization', and does not have the core capability of schedule management. The second category is a general schedule management technology or product, which covers a mobile phone with a schedule APP, a professional care APP and the like, and has the core functions of providing a schedule manual input and timing reminding service, and part of advanced functions can be associated with single health equipment data, such as medication reminding of a smart watch. However, existing content push-type technologies can only passively respond to content demands, have no capability to actively generate and adjust task schedules, and data processing is limited to structured data, with incomplete perception of user status. The universal schedule management technology has low intelligent degree, relies on manual schedule entry, and does not design a special function aiming at the core pain points such as independent safety, chronic disease management and the like of home care scenes. These drawbacks have resulted in the prior art being difficult to meet the comprehensive schedule management needs of elderly users, particularly solitary, chronically ill users. Disclosure of Invention In view of the above, the application provides a method, a device, a medium and equipment for managing schedules for home-based care, which can solve the problems that the content pushing technology is low in intelligent degree, the equipment linkage is insufficient, and the schedules cannot be dynamically adjusted based on the health state. According to a first aspect of the present application, there is provided a schedule management method for home-based care, including: Acquiring multi-mode data of a user in a home care scene in real time based on a multi-device linkage mode, wherein the multi-mode data at least comprises health index data and behavior data, and the behavior data comprises schedule operation data, voice interaction data, behavior gesture data and sleep state data of terminal equipment; carrying out data fusion processing on the multi-mode data to generate a standardized data set; invoking an AI large model to analyze the association rule between the behavior data and the health index data in the data set, and generating an initial personalized schedule based on the association rule; and monitoring the health index data in real time through an abnormality identification algorithm, and dynamically adjusting the initial personalized schedule based on the abnormal type of the health index and a preset schedule adjustment strategy database to generate a target personalized schedule when the health index is abnormal. According to a second aspect of the present application, there is provided a schedule management apparatus for home-based care, comprising: The system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring multi-mode data of a user in a home care scene in real time based on a multi-equipment linkage mode, the multi-mode data at least comprises health index data and behavior data, and the behavior data comprises schedule operation data, voice interaction data, behavior gesture data and sleep state data of terminal equipment; The processing module is used for carrying out data fusion processing on the multi-mode data to generate a standardized data set; the analysis