A method and system for intelligent scheduling of elderly care service resources based on multi-dimensional portraits

Abstract

This invention discloses a method and system for intelligent scheduling of elderly care service resources based on multi-dimensional profiling, belonging to the field of intelligent scheduling technology for elderly care service resources. The method includes the following steps: integrating elderly health records, wearable data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles; synchronizing the digitalization of caregiver skills, locations, schedules, and equipment resources; developing care plans for the elderly, selecting 3-5 caregivers with matching skills and geographical proximity from a resource pool according to the complexity of care to form a dedicated service team, and designating a lead caregiver; simultaneously, determining the number of elderly people the lead caregiver can also assist with based on the level of care complexity. This invention, through the combination of multi-dimensional profiling, lead / assistant matching, risk prediction, path optimization, and dynamic scheduling, achieves precise allocation and efficient scheduling of elderly care service resources, balancing care efficiency and human warmth, improving service continuity, safety, and satisfaction, and promoting the intelligent, refined, and humanized development of elderly care.

Description

Technical Field

[0001] This invention relates to the field of intelligent scheduling technology for elderly care service resources, and more specifically, to an intelligent scheduling method and system for elderly care service resources based on multi-dimensional profiling. Background Technology

[0002] Currently, the allocation of elderly care service resources relies heavily on traditional dispatching models or algorithm-based scheduling that is solely driven by efficiency. This often results in problems such as inefficient matching of supply and demand, insufficient humanistic care, imbalanced allocation of manpower, and delayed emergency response. Traditional dispatching methods struggle to integrate multi-dimensional data on the elderly's health, behavior, and care needs, failing to create accurate and dynamic individual profiles. Caregiver resources suffer from low digitization and limited matching criteria, often resulting in skill mismatches, frequent staff turnover, and neglect of the elderly's psychological dependence on familiar caregivers, leading to a lack of human touch and low satisfaction. Furthermore, existing systems do not implement differentiated staffing based on the complexity of care, resulting in insufficient care for high-needs elderly and wasted resources for low-needs elderly, leading to uneven workloads for caregivers and overall low efficiency in human resource utilization. In addition, most dispatching platforms can only execute pre-defined care plans, lacking the ability to anticipate and mitigate sudden, non-standard fluctuations in elderly care, making them unable to cope with unexpected increases in care demands and prone to care gaps and safety hazards. Regarding path planning, traditional algorithms solely pursue the shortest path and highest frequency, failing to consider the intensity of caregivers' workloads and emotional burdens, easily leading to service fatigue in the long run and failing to balance dispatching efficiency with humanistic care. The aforementioned problems hinder the development of elderly care services towards refinement, intelligence, and humanization. There is an urgent need for an intelligent scheduling method based on multi-dimensional profiling, balancing efficiency and warmth, predicting risks, and balancing human resources, in order to break through existing technological bottlenecks and improve the overall quality of care and service stability. No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0003] To address the problems in related technologies, this invention proposes an intelligent scheduling method and system for elderly care service resources based on multi-dimensional profiling, in order to overcome the aforementioned technical problems in existing related technologies, such as extensive matching, lack of humanistic care, imbalance of manpower, delayed risk assessment, and imbalance between efficiency and care. Therefore, the specific technical solution adopted by the present invention is as follows: A method for intelligent scheduling of elderly care service resources based on multi-dimensional profiling, comprising the following steps: S1. Integrate elderly health records, wearable data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles; synchronize digital caregiver skills, locations, schedules, and equipment resources. S2. To develop a care plan for the elderly, select 3-5 caregivers with matching skills and from nearby locations from the resource pool according to the complexity of the care to form an exclusive service team, and designate a lead caregiver; at the same time, determine the number of elderly people that the lead caregiver can also assist in based on the level of care complexity. S3. Continuously analyze the risk factors in the elderly profile and predict the possible non-standard fluctuations; identify high-risk periods and automatically push the corresponding nursing skills and emergency plans to their primary caregiver and assistant caregiver. S4. Divide the city into community grids and cluster the needs of the elderly in the same time period and the same grid. Within the appointment time window, the algorithm plans the optimal route connecting multiple points for the dedicated service team caregivers and dynamically inserts work intervals without travel distance as buffers based on the physical and emotional load of the previous task. S5. Real-time monitoring of caregiver location, task progress and elderly person's condition; dynamic rescheduling in case of emergencies: priority is given to having an auxiliary caregiver from the same group take over, or a mobile caregiver from a nearby grid is dispatched to provide support, and service records are synchronized to minimize service interruption. S6. After the service is completed, collect feedback, satisfaction and core indicators, analyze the scheduling effect, and continuously optimize the profile weight, risk model and path algorithm to achieve intelligent and humanized iterative upgrades of scheduling. As a preferred implementation, the process of integrating elderly health records, wearable data, behavioral habits, and care plans to generate a dynamically updated multi-dimensional personal profile, and synchronizing digital caregiver skills, locations, schedules, and equipment resources, includes the following steps: S11. Collect heterogeneous data such as the elderly’s health indicators, daily activities, medication records and care preferences, and clean, deduplicate and standardize the data to form a unified and structured raw data pool. S12. Based on the cleaned data, a machine learning model is used to construct a quantifiable personal profile label system from five dimensions: physical health, self-care ability, behavioral patterns, psychological and social aspects, and care needs. This profile will be dynamically updated as new data flows in, forming a digital twin that reflects the elderly person's real-time condition. S13. Digitally model resources such as caregivers, equipment, and facilities: Define the professional skills, service radius, work schedule, and real-time location of caregivers; Mark the type, status, and geographical location of resources such as rehabilitation equipment, transportation vehicles, and beds; S14. Based on the profile and resource digital model, establish a multi-dimensional matching rule library to associate the disability level with the caregiver's qualifications and the type of chronic disease with the function of rehabilitation equipment. S15. Based on caregivers' historical service data, user reviews, skill certificates, and real-time status, generate a multi-dimensional resource profile for each caregiver, covering their professional competence, service style, responsiveness, and workload. As a preferred implementation method, the process of developing a care plan for the elderly, selecting 3-5 caregivers with matching skills and geographical proximity from the resource pool according to the complexity of the care to form a dedicated service team, and designating a lead caregiver; and determining the number of elderly people that the lead caregiver can also assist with based on the level of care complexity, includes the following steps: S21. Based on the multidimensional health profile of the elderly, identify the core nursing needs and risk points; then, in combination with professional nursing standards and available resource types, transform the needs into specific nursing goals, service items, execution frequency and quality standards to form a nursing plan. S22. Analyze the complexity level of the nursing care plan and classify the nursing care plan into low, medium and high levels; search for caregivers who meet the core conditions of qualifications, skills and experience according to the complexity level, and further combine factors such as real-time location, available time and historical service evaluation to screen out a list of candidate caregivers with high skill matching and geographical proximity. S23. Select 3-5 people from the candidate caregivers to form an exclusive service team for the elderly. Based on indicators such as service continuity, past interaction evaluation with the elderly, and matching of professional expertise, designate one caregiver as the main caregiver and the rest as assistant caregivers to form a main-assistant collaborative structure. S24. When the nursing plan is high-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 1; when the nursing plan is medium-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 2; when the nursing plan is low-level, the number of service recipients for which the primary caregiver also serves as an assistant caregiver is automatically limited to no more than 3. Caregivers can serve multiple elderly people at the same time, serving as the primary caregiver for at least one elderly person and also assisting in the care of other elderly people. Through this clear division of labor, caregivers can concentrate on meeting the in-depth care needs of the corresponding elderly people, ensuring service continuity, and improving the overall efficiency of human resource utilization. S25. A shared workbench is set up for the dedicated caregiver team. The lead caregiver is responsible for maintaining and synchronizing the latest status of the elderly, care points and precautions. The assistant caregiver can obtain updated information in real time when participating in the service, ensuring information continuity and consistent operation during the service process, and reducing the risk of care gaps caused by staff rotation. In a preferred embodiment, the analysis of the complexity level of the nursing care plan, classifying the nursing care plan into low, medium, and high levels, includes the following steps: S221. Construct an assessment framework that includes daily living activities, cognitive function, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and frequency of nursing care, and clarify the quantitative grading standards for each dimension. S222. Convert the elderly person's profile data into quantifiable indicator data, and input the quantified data into a preset scoring model to calculate the comprehensive score of care complexity; the model formula is: ; in, The calculated overall score; These include scores for activities of daily living, degree of cognitive impairment, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and nursing frequency coefficient. The weighting coefficients are respectively for the scores of activities of daily living, the degree of cognitive impairment, the number of chronic diseases, the need for special nursing skills, emotional and behavioral problems, and the nursing frequency coefficient; S223. Based on the comprehensive scoring results, classify the grades according to the preset threshold range into low, medium and high. As a preferred implementation, the continuous analysis of risk factors in the elderly person's profile to predict potential non-standard fluctuations, and the automatic push of corresponding nursing skills and emergency plans to the primary and secondary caregivers during high-risk periods, includes the following steps: S31. Continuously collected multi-dimensional profile data of the elderly is correlated and analyzed in real time with the preset medical risk model and behavior pattern library to construct a dynamic risk factor knowledge graph that is exclusive to each individual. S32. By setting threshold rules and machine learning models, identify data patterns that deviate from the baseline or show a worsening trend, including abnormal blood pressure fluctuations for three consecutive days, a sudden drop in daytime activity levels of more than 30%, etc., and automatically determine them as potential risk signals. S33. Based on the identified risk signals, combined with environmental factors such as the elderly’s medical history, age, season and weather, a predictive model is used to infer the non-standard events that may occur during the next care. S34. When a high-risk event is predicted, the system will automatically match the corresponding standardized nursing intervention measures, skill operation points and precautions from the contingency plan library to form a structured and executable contingency plan package. S35. The generated contingency plan package will be pushed to the primary and secondary caregivers of the elderly person via the caregiver's mobile app, and the caregivers must confirm receipt online. The pushed content will highlight risk warnings, key time windows, and specific operation instructions. In a preferred embodiment, the step of using a predictive model to infer potential non-standard events during the next care session based on identified risk signals, combined with environmental factors such as the elderly person's medical history, age, season, and weather, includes the following steps: S331. Search historical data to find the group that is closest to the current elderly profile, and calculate the probability of various non-standard events occurring in this group during the next care. S332. Transform each elderly person's individual nursing record into a set of standardized feature data, which includes static features, dynamic features, and environmental features. Static features include age, gender, basic medical history, and ADL baseline. Dynamic features include recent blood pressure fluctuations, sleep quality, mood score, and medication adherence. Environmental features include season, sudden drop in temperature, and changes in air pressure. S333. Clearly define the non-standard events to be predicted, namely the probability of severe pain, the probability of emotional breakdown, and the probability of needing emergency medical intervention. S334. Based on the collected data, calculate the actual incidence of various non-standard events in this group during the next care session. The formula for the prediction model is: ; in, ; This represents the probability of a non-standard event occurring, with a value ranging from 0 to 1. For the input feature values, These are weighting coefficients, trained using historical big data, representing the degree of influence of each factor on risk. It is a natural constant. The base risk score of the model, i.e., the intercept term, is the baseline probability of an event occurring when all features are 0. S335. The numerical range of the model output is 0-1. The preset risk threshold is P≥0.6. If the value exceeds 0.6, it will be automatically marked as high risk. As a preferred implementation, the process of dividing the city into community grids and clustering the needs of elderly people within the same grid and at the same time period includes the following steps: Within the appointment time window, the algorithm plans the optimal multi-point connection route for dedicated service team caregivers, and dynamically inserts work intervals without travel distances as buffers based on the physical and emotional load of the previous task. S41. Based on the geographic information system and community administrative boundaries, the urban area is divided into several standardized service grids. The system automatically aggregates all elderly service requests within the same service period and located in the same grid to form a task pool to be scheduled for that grid. S42. From the established dedicated caregiver groups for the elderly, select caregiver members who are currently available and whose current location is in or near the grid, and combine the matching degree of the main and auxiliary caregivers' responsibilities and skills to initially allocate service requests in the task pool to suitable caregivers. S43. Under the premise of meeting the appointment time window of each elderly person, the path optimization algorithm is used to calculate the optimal geographical order for each caregiver to visit all their assigned service points, ensuring the shortest total travel time and generating a preliminary connected service route. S44. Read the type and duration of the previous task completed by the caregiver, and combine the preset physical exertion coefficient and emotional load assessment model to dynamically calculate the pure work buffer interval (excluding travel time) required by the caregiver before starting the next task. S44. The calculated personalized work buffer intervals are inserted between each service node in the preliminary path planning to form a final scheduling scheme that includes the complete sequence of movement-buffering-service, and to clarify the arrival time of the caregiver, the buffer start time and the service start time at each point. A smart scheduling system for elderly care service resources based on multi-dimensional profiles. The system adopts a smart scheduling method for elderly care service resources based on multi-dimensional profiles as described above, including a dynamic multi-dimensional profile construction module, a primary and secondary caregiver matching module, a risk prediction-based nursing ability push module, a service route planning module, a dynamic scheduling module, and a feedback evaluation module. The dynamic multidimensional profile construction module is responsible for integrating elderly health records, wearable device data, behavioral habits and care plans, completing data cleaning, standardization and structuring, and constructing a dynamically updated multidimensional personal profile of the elderly from five dimensions: physical health, self-care ability, behavioral patterns, psychosocial aspects and care needs; at the same time, it digitally models caregiver skills, location, scheduling, service capabilities and equipment resources. The primary and secondary caregiver matching module assesses the nursing needs and complexity level based on the elderly person's multi-dimensional profile, selects caregivers with matching skills and nearby locations from the resource pool, forms a dedicated service team of 3-5 people for each elderly person and assigns a primary caregiver; and determines the number of elderly people that the primary caregiver can also serve as a secondary caregiver based on the nursing complexity. The risk prediction nursing ability push module continuously analyzes the risk factors in the elderly profile, identifies abnormal signals through risk models and machine learning, and predicts non-standard fluctuations and high-risk events; during high-risk periods, it automatically matches emergency plans and nursing skill points to form a standardized plan package and pushes it to primary and secondary caregivers to improve care foresight and emergency response capabilities. The service route planning module clusters service demands for the same time period by community grid, and uses a path optimization algorithm to plan the optimal route connecting multiple points for caregivers within the appointment time window; combined with the physical and emotional load of the caregiver's previous task, it dynamically inserts pure work buffer intervals without routes, and generates a complete time sequence plan including movement, buffering and service. The dynamic scheduling module monitors the location of caregivers, task progress and the elderly person's condition in real time. In case of emergencies, it immediately performs dynamic rescheduling, prioritizing the replacement by auxiliary caregivers in the same group, or dispatching mobile caregivers from nearby grids to provide support, and synchronizing service records and service information. The feedback evaluation module collects user feedback, satisfaction, and core operational metrics after the service is completed, and comprehensively analyzes the scheduling effect and service quality. Based on the analysis results, it continuously optimizes the profile weights, risk models, path algorithms, and scheduling rules, driving the system to continuously iterate and upgrade. The beneficial effects of this invention are as follows: 1. This invention constructs a multi-dimensional dynamic profile of the elderly and a digital model of caregiver resources to achieve precise matching between the needs of the elderly and care resources. It sets up a dedicated service team for each elderly person with a fixed primary caregiver and multiple suitable auxiliary caregivers. This effectively solves the problem that traditional algorithm scheduling only focuses on efficiency and lacks humanistic care. It avoids the psychological discomfort and lack of security caused by frequent visits from unfamiliar caregivers, stabilizes the emotional dependence of the elderly, and preserves the human touch and warmth of elderly care services in intelligent scheduling. It significantly improves the psychological comfort and overall satisfaction of the elderly when receiving care. 2. This invention configures the rules for primary and secondary caregivers to take on additional care based on the level of care complexity. For elderly people with high care complexity, the primary caregiver focuses on their care and strictly controls the number of secondary caregivers. For elderly people with low care complexity, the primary caregiver is reasonably arranged to take on additional care services. This can ensure that key elderly people such as the disabled, the elderly, and those with multiple diseases receive sufficient and stable dedicated care, while also balancing the workload and service load of all caregivers, avoiding idle or overly fatigued manpower. This can significantly improve the utilization rate of human resources while ensuring the quality of care, and achieve a dual improvement in service accuracy and operational efficiency. 3. This invention establishes a proactive prediction and advance push fault tolerance mechanism through real-time analysis of risk factors and prediction of non-standard fluctuations. It can identify potential abnormal conditions and sudden care needs of the elderly in advance, and accurately push the corresponding nursing skills and treatment plans to the primary and secondary caregivers. This effectively makes up for the shortcomings of static profiles in covering temporary and sudden care needs, enhances the system's adaptability to special situations, prevents service gaps caused by insufficient care capabilities, and comprehensively improves care safety, continuity and emergency response levels. 4. This invention uses community grids as units to cluster needs, combines path optimization algorithms to form an efficient chain of services, and dynamically inserts no-route buffer intervals according to the work intensity of caregivers. This not only shortens commuting time and improves the efficiency of home visits, but also fully considers the physical and emotional burden of caregivers, making scheduling more scientific and humane, achieving a balance between scheduling efficiency and humanistic care, and promoting the overall upgrading of elderly care towards refinement, intelligence and stability. Attached Figure Description To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1This is a flowchart of an intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to an embodiment of the present invention; Figure 2 This is a block diagram of an intelligent scheduling system for elderly care service resources based on multi-dimensional profiling, according to an embodiment of the present invention. Detailed Implementation To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components. According to an embodiment of the present invention, a method and system for intelligent scheduling of elderly care service resources based on multi-dimensional profiling is provided. The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figures 1-2 As shown, according to an embodiment of the present invention, an intelligent scheduling method for elderly care service resources based on multi-dimensional profiling includes the following steps: S1. Integrate elderly health records, wearable data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles; synchronize digital caregiver skills, locations, schedules, and equipment resources. Integrating elderly health records, wearable device data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles; synchronizing digital caregiver skills, locations, scheduling, and equipment resources includes the following steps: Furthermore, S11 collects heterogeneous data such as the elderly’s health indicators, daily activities, medication records and care preferences, and cleans, deduplicates and standardizes the data to form a unified and structured raw data pool. S12. Based on the cleaned data, a machine learning model is used to construct a quantifiable personal profile label system from five dimensions: physical health, self-care ability, behavioral patterns, psychological and social aspects, and care needs. S13. Digitally model resources such as caregivers, equipment, and facilities: Define the professional skills, service radius, work schedule, and real-time location of caregivers; Mark the type, status, and geographical location of resources such as rehabilitation equipment, transportation vehicles, and beds; S14. Based on the profile and resource digital model, establish a multi-dimensional matching rule library to associate the disability level with the caregiver's qualifications and the type of chronic disease with the function of rehabilitation equipment. S15. Based on caregivers' historical service data, user reviews, skills certificates, and real-time status, generate a multi-dimensional resource profile for each caregiver, covering their professional competence, service style, adaptability, and workload. S2. Develop a care plan for the elderly, select 3-5 caregivers with matching skills and from nearby areas from the resource pool according to the complexity of the care to form an exclusive service team, and designate a lead caregiver; at the same time, determine the number of elderly people that the lead caregiver can also assist in based on the level of care complexity. Furthermore, a care plan is developed for the elderly. Based on the complexity of the care, 3-5 caregivers with matching skills and from nearby locations are selected from the resource pool to form a dedicated service team, and a lead caregiver is designated. Simultaneously, based on the level of care complexity, the number of elderly people the lead caregiver can also assist with includes the following steps: S21. Based on the multidimensional health profile of the elderly, identify the core nursing needs and risk points; then, in combination with professional nursing standards and available resource types, transform the needs into specific nursing goals, service items, execution frequency and quality standards to form a nursing plan. S22. Analyze the complexity level of the nursing care plan and classify the nursing care plan into low, medium and high levels; search for caregivers who meet the core conditions of qualifications, skills and experience according to the complexity level, and further combine factors such as real-time location, available time and historical service evaluation to screen out a list of candidate caregivers with high skill matching and geographical proximity. Furthermore, analyzing the complexity level of the care plan and classifying it into low, medium, and high levels includes the following steps: S221. Construct an assessment framework that includes daily living activities, cognitive function, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and frequency of nursing care, and clarify the quantitative grading standards for each dimension. S222. Convert the elderly person's profile data into quantifiable indicator data, and input the quantified data into a preset scoring model to calculate the comprehensive score of care complexity; the model formula is: ; in, The calculated overall score; These include scores for activities of daily living, degree of cognitive impairment, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and nursing frequency coefficient. The weighting coefficients are respectively for the scores of activities of daily living, the degree of cognitive impairment, the number of chronic diseases, the need for special nursing skills, emotional and behavioral problems, and the nursing frequency coefficient; It should be noted that for the activities of daily living assessment, the lower the score, the higher the complexity; the degree of cognitive impairment is 0 – none, 1 – mild, 2 – moderate, 3 – severe; the number of chronic diseases is directly counted, such as 0, 1, 2, 3...; special nursing skills requirements are 0 – none, 1 – basic, 2 – professional; emotional and behavioral problems are 0 – none, 1 – occasional, 2 – frequent, such as aggression / wandering; and the nursing frequency coefficient is low frequency (once a week), medium frequency (2-3 times a week), and high frequency (once a day or 24-hour shifts). It was obtained by fitting historical data. S223. Based on the comprehensive scoring results, classify the grades according to the preset threshold range into low, medium and high. S23. Select 3-5 people from the candidate caregivers to form an exclusive service team for the elderly. Based on indicators such as service continuity, past interaction evaluation with the elderly, and matching of professional expertise, designate one caregiver as the main caregiver and the rest as assistant caregivers to form a main-assistant collaborative structure. S24. When the nursing plan is high-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 1; when the nursing plan is medium-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 2; when the nursing plan is low-level, the number of service recipients for which the primary caregiver also serves as an assistant caregiver is automatically limited to no more than 3. S25. A shared workbench is set up for the dedicated caregiver team. The lead caregiver is responsible for maintaining and synchronizing the latest status of the elderly, care points and precautions. The assistant caregiver can obtain updated information in real time when participating in the service, ensuring information continuity and consistent operation during the service process, and reducing the risk of care gaps caused by staff rotation. It should be noted that the system will continuously monitor the service load, satisfaction feedback and actual service effect of the primary caregiver and each auxiliary caregiver. When there is a significant imbalance in workload or a significant change in the elderly person's condition, the system will automatically trigger dynamic adjustments to the team members or concurrent care rules to achieve flexible optimization of resource allocation. S3. Continuously analyze the risk factors in the elderly profile and predict the possible non-standard fluctuations; identify high-risk periods and automatically push the corresponding nursing skills and emergency plans to their primary caregiver and assistant caregiver. Furthermore, continuous analysis of risk factors in elderly profiles is conducted to predict potential non-standard fluctuations. High-risk periods are identified, and corresponding nursing skills and emergency plans are automatically pushed to the primary and secondary caregivers, including the following steps: S31. Continuously collected multi-dimensional profile data of the elderly is correlated and analyzed in real time with the preset medical risk model and behavior pattern library to construct a dynamic risk factor knowledge graph that is exclusive to each individual. S32. By setting threshold rules and machine learning models, identify data patterns that deviate from the baseline or show a worsening trend, including abnormal blood pressure fluctuations for three consecutive days, a sudden drop in daytime activity levels of more than 30%, etc., and automatically determine them as potential risk signals. S33. Based on the identified risk signals, combined with environmental factors such as the elderly’s medical history, age, season and weather, a predictive model is used to infer the non-standard events that may occur during the next care. Based on identified risk signals, and combined with environmental factors such as the elderly person's medical history, age, season, and weather, a predictive model is used to infer potential non-standard events during the next care session. This includes the following steps: Furthermore, S331, search historical data to find the group that is closest to the current elderly profile, and calculate the probability of various non-standard events occurring in this group during the next care. S332. Transform each elderly person's individual nursing record into a set of standardized feature data, which includes static features, dynamic features, and environmental features. Static features include age, gender, basic medical history, and ADL baseline. Dynamic features include recent blood pressure fluctuations, sleep quality, mood score, and medication adherence. Environmental features include season, sudden drop in temperature, and changes in air pressure. S333. Clearly define the non-standard events to be predicted, namely the probability of severe pain, the probability of emotional breakdown, and the probability of needing emergency medical intervention. S334. Based on the collected data, calculate the actual incidence of various non-standard events in this group during the next care session. The formula for the prediction model is: ; in, ; This represents the probability of a non-standard event occurring, with a value ranging from 0 to 1. For the input feature values, These are weighting coefficients, trained using historical big data, representing the degree of influence of each factor on risk. It is a natural constant. The base risk score of the model, i.e., the intercept term, is the baseline probability of an event occurring when all features are 0. S335. The numerical range of the model output is 0-1. The preset risk threshold is P≥0.6. If the value exceeds 0.6, it is automatically marked as high risk. S34. When a high-risk event is predicted, the system will automatically match the corresponding standardized nursing intervention measures, skill operation points and precautions from the contingency plan library to form a structured and executable contingency plan package. S35. The generated contingency plan package will be pushed to the primary caregiver and auxiliary caregiver team of the elderly person via the caregiver's terminal APP, and the caregivers need to confirm receipt online. The pushed content will highlight risk warnings, key time windows, and specific operation instructions; S4. Divide the city into community grids and cluster the needs of the elderly in the same time period and the same grid. Within the appointment time window, the algorithm plans the optimal route connecting multiple points for the dedicated service team caregivers and dynamically inserts work intervals without travel distance as buffers based on the physical and emotional load of the previous task. Furthermore, the city is divided into community grids, and the needs of elderly people within the same grid and at the same time are clustered. Within the appointment time window, the algorithm plans the optimal route connecting multiple points for dedicated service team caregivers, and dynamically inserts work intervals without travel distance as buffers based on the physical and emotional load of the previous task. This includes the following steps: S41. Based on the geographic information system and community administrative boundaries, the urban area is divided into several standardized service grids. The system automatically aggregates all elderly service requests within the same service period and located in the same grid to form a task pool to be scheduled for that grid. S42. From the established dedicated caregiver groups for the elderly, select caregiver members who are currently available and whose current location is in or near the grid, and combine the matching degree of the main and auxiliary caregivers' responsibilities and skills to initially allocate service requests in the task pool to suitable caregivers. S43. Under the premise of meeting the appointment time window of each elderly person, the path optimization algorithm is used to calculate the optimal geographical order for each caregiver to visit all their assigned service points, ensuring the shortest total travel time and generating a preliminary connected service route. S44. Read the type and duration of the previous task completed by the caregiver, and combine the preset physical exertion coefficient and emotional load assessment model to dynamically calculate the pure work buffer interval (excluding travel time) required by the caregiver before starting the next task. S44. The calculated personalized work buffer intervals are inserted between each service node in the preliminary path planning to form a final scheduling scheme that includes the complete sequence of movement-buffering-service, and to clarify the arrival time of the caregiver, the buffer start time and the service start time at each point. It should be noted that during the implementation of the plan, the system tracks the location of caregivers and the progress of tasks in real time. Once it detects that the actual progress deviates from the plan, such as task overtime or traffic delays, it will automatically trigger dynamic rescheduling, re-optimize the path and buffer arrangement of subsequent tasks, and send adjustment notifications to caregivers and managers. S5. Real-time monitoring of caregiver location, task progress and elderly person's condition; dynamic rescheduling in case of emergencies: priority is given to having an auxiliary caregiver from the same group take over, or a mobile caregiver from a nearby grid is dispatched to provide support, and service records are synchronized to minimize service interruption. It should be noted that by monitoring the location of caregivers, task progress, and the condition of the elderly in real time, combined with a dynamic rescheduling mechanism in case of emergencies, it is possible to prioritize the use of auxiliary caregivers in the same group to continue services or to dispatch mobile caregivers from nearby grids for rapid support, and to record services simultaneously. This not only effectively ensures that the elderly, especially the disabled and elderly, receive timely and continuous care in case of emergencies, but also significantly reduces the risk of service interruption and safety hazards. It also optimizes the allocation of caregiver manpower, improves emergency response efficiency and service stability, and retains complete service records for easy management, traceability, and quality control, thus achieving refined, efficient, and safe elderly care as a whole. S6. After the service is completed, collect feedback, satisfaction and core indicators, analyze the scheduling effect, and continuously optimize the profile weight, risk model and path algorithm to achieve intelligent and humanized iterative upgrades of scheduling. It should be noted that after the service is completed, by collecting user feedback, satisfaction evaluations, and core operational indicators, a comprehensive analysis of the scheduling effect is conducted. Based on this analysis, the personnel profile weights, risk prediction models, and order dispatching algorithms are continuously optimized. This allows the scheduling system to continuously improve itself in actual operation, enhancing response speed and resource allocation efficiency, achieving intelligent scheduling decisions, and better aligning with the needs of care scenarios and service recipients. This makes scheduling arrangements more practical and humane, continuously improving service quality and user experience, and promoting efficient iteration and long-term optimization of the overall care service system. A smart scheduling system for elderly care service resources based on multi-dimensional profiles. The system adopts any one of the above-mentioned smart scheduling methods for elderly care service resources based on multi-dimensional profiles, including a dynamic multi-dimensional profile construction module, a primary and secondary caregiver matching module, a risk prediction-based nursing capacity push module, a service route planning module, a dynamic scheduling module, and a feedback evaluation module. The dynamic multidimensional profile building module is responsible for integrating elderly people's health records, wearable device data, behavioral habits and care plans, completing data cleaning, standardization and structuring, and building a dynamically updated multidimensional personal profile of the elderly from five dimensions: physical health, self-care ability, behavioral patterns, psychosocial, and care needs; at the same time, it digitally models caregiver skills, location, scheduling, service capabilities and equipment resources. The primary and secondary caregiver matching module assesses the nursing needs and complexity level based on the elderly's multi-dimensional profile, selects caregivers with matching skills and nearby locations from the resource pool, forms a dedicated service team of 3-5 people for each elderly person and assigns a primary caregiver; and determines the number of elderly people that the primary caregiver can also serve as a secondary caregiver based on the nursing complexity. The risk prediction and nursing capability push module continuously analyzes risk factors in elderly profiles, identifies abnormal signals through risk models and machine learning, and predicts non-standard fluctuations and high-risk events. During high-risk periods, it automatically matches emergency plans and nursing skill points to form standardized plan packages and pushes them to primary and secondary caregivers, thereby improving care foresight and emergency response capabilities. The service route planning module clusters service needs for the same time period by community grid and uses a path optimization algorithm to plan the optimal route connecting multiple points for caregivers within the appointment time window. Combining the physical and emotional load of the caregiver's previous task, it dynamically inserts pure work buffer intervals without routes to generate a complete time sequence plan that includes movement, buffering and service. The dynamic scheduling module monitors the location of caregivers, task progress, and the elderly person's condition in real time. In case of emergencies, it immediately performs dynamic rescheduling, prioritizing the replacement by auxiliary caregivers in the same group or dispatching mobile caregivers from nearby grids for support, and synchronizing service records and service information. The feedback and evaluation module collects user feedback, satisfaction, and core operational metrics after the service is completed, and comprehensively analyzes the scheduling effect and service quality. Based on the analysis results, it continuously optimizes the profile weights, risk models, path algorithms, and scheduling rules, driving the system to continuously iterate and upgrade. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent scheduling of elderly care service resources based on multi-dimensional profiling, characterized in that, The method includes the following steps: S1. Integrate elderly health records, wearable data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles; synchronize digital caregiver skills, locations, schedules, and equipment resources. S2. To develop a care plan for the elderly, select 3-5 caregivers with matching skills and from nearby locations from the resource pool according to the complexity of the care to form an exclusive service team, and designate a lead caregiver; at the same time, determine the number of elderly people that the lead caregiver can also assist in based on the level of care complexity. S3. Continuously analyze the risk factors in the elderly profile and predict the possible non-standard fluctuations; identify high-risk periods and automatically push the corresponding nursing skills and emergency plans to their primary caregiver and assistant caregiver. S4. Divide the city into community grids and cluster the needs of the elderly in the same time period and the same grid. Within the appointment time window, the algorithm plans the optimal route connecting multiple points for the dedicated service team caregivers and dynamically inserts work intervals without travel distance as buffers based on the physical and emotional load of the previous task. S5. Real-time monitoring of caregiver location, task progress and elderly person's condition; dynamic rescheduling in case of emergencies: priority is given to having an auxiliary caregiver from the same group take over, or a mobile caregiver from a nearby grid is dispatched to provide support, and service records are synchronized to minimize service interruption. S6. After the service is completed, collect feedback, satisfaction and core indicators, analyze the scheduling effect, and continuously optimize the profile weight, risk model and path algorithm to achieve intelligent and humanized iterative upgrades of scheduling.

2. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 1, characterized in that, The system integrates elderly health records, wearable data, behavioral habits, and care plans to generate dynamically updated multi-dimensional personal profiles. Synchronizing digital caregiver skills, locations, scheduling, and equipment resources includes the following steps: S11. Collect heterogeneous data such as the elderly’s health indicators, daily activities, medication records and care preferences, and clean, deduplicate and standardize the data to form a unified and structured raw data pool. S12. Based on the cleaned data, a machine learning model is used to construct a quantifiable personal profile label system from five dimensions: physical health, self-care ability, behavioral patterns, psychological and social aspects, and care needs. S13. Digitally model resources such as caregivers, equipment, and facilities: Define the professional skills, service radius, work schedule, and real-time location of caregivers; Mark the type, status, and geographical location of resources such as rehabilitation equipment, transportation vehicles, and beds; S14. Based on the profile and resource digital model, establish a multi-dimensional matching rule library to associate the disability level with the caregiver's qualifications and the type of chronic disease with the function of rehabilitation equipment. S15. Based on caregivers' historical service data, user reviews, skill certificates, and real-time status, generate a multi-dimensional resource profile for each caregiver, covering their professional competence, service style, responsiveness, and workload.

3. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 1, characterized in that, The process of developing a care plan for the elderly involves selecting 3-5 caregivers with matching skills and from nearby locations from a resource pool, based on the complexity of the care, to form a dedicated service team, and designating a lead caregiver. Simultaneously, based on the level of care complexity, the number of elderly people the lead caregiver can also assist with includes the following steps: S21. Based on the multidimensional health profile of the elderly, identify the core nursing needs and risk points; then, in combination with professional nursing standards and available resource types, transform the needs into specific nursing goals, service items, execution frequency and quality standards to form a nursing plan. S22. Analyze the complexity level of the nursing care plan and classify the nursing care plan into low, medium and high levels; search for caregivers who meet the core conditions of qualifications, skills and experience according to the complexity level, and further combine factors such as real-time location, available time and historical service evaluation to screen out a list of candidate caregivers with high skill matching and geographical proximity. S23. Select 3-5 people from the candidate caregivers to form an exclusive service team for the elderly. Based on indicators such as service continuity, past interaction evaluation with the elderly, and matching of professional expertise, designate one caregiver as the main caregiver and the rest as assistant caregivers to form a main-assistant collaborative structure. S24. When the nursing plan is high-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 1; when the nursing plan is medium-level, the number of primary caregivers who also serve as assistant caregivers for other elderly people is automatically limited to no more than 2; when the nursing plan is low-level, the number of service recipients for which the primary caregiver also serves as an assistant caregiver is automatically limited to no more than 3. S25. A shared workbench is set up for the dedicated caregiver team. The lead caregiver is responsible for maintaining and synchronizing the latest status of the elderly, care points and precautions. The assistant caregiver can obtain updated information in real time when participating in the service, ensuring information continuity and consistent operation during the service process, and reducing the risk of care gaps caused by staff rotation.

4. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 3, characterized in that, The analysis of the complexity level of the nursing care plan, classifying the nursing care plan into low, medium, and high levels, includes the following steps: S221. Construct an assessment framework that includes daily living activities, cognitive function, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and frequency of nursing care, and clarify the quantitative grading standards for each dimension. S222. Convert the elderly person's profile data into quantifiable indicator data, and input the quantified data into a preset scoring model to calculate the comprehensive score of care complexity; the model formula is: ; in, The calculated overall score; These include scores for activities of daily living, degree of cognitive impairment, number of chronic diseases, special nursing care needs, emotional and behavioral problems, and nursing frequency coefficient. The weighting coefficients are respectively for the scores of activities of daily living, the degree of cognitive impairment, the number of chronic diseases, the need for special nursing skills, emotional and behavioral problems, and the nursing frequency coefficient; S223. Based on the comprehensive scoring results, classify the grades according to the preset threshold range into low, medium and high.

5. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 1, characterized in that, The continuous analysis of risk factors in elderly profiles to predict potential non-standard fluctuations, and the automatic delivery of corresponding nursing skills and emergency plans to primary and secondary caregivers during high-risk periods, includes the following steps: S31. Continuously collected multi-dimensional profile data of the elderly is correlated and analyzed in real time with the preset medical risk model and behavior pattern library to construct a dynamic risk factor knowledge graph that is exclusive to each individual. S32. By setting threshold rules and machine learning models, identify data patterns that deviate from the baseline or show a worsening trend, including abnormal blood pressure fluctuations for three consecutive days and a sudden drop in daytime activity levels of more than 30%, which are automatically identified as potential risk signals. S33. Based on the identified risk signals, combined with environmental factors such as the elderly’s medical history, age, season and weather, a predictive model is used to infer the non-standard events that may occur during the next care. S34. When a high-risk event is predicted, the system will automatically match the corresponding standardized nursing intervention measures, skill operation points and precautions from the contingency plan library to form a structured and executable contingency plan package. S35. The generated contingency plan package will be pushed to the primary caregiver and auxiliary caregiver team of the elderly through the caregiver terminal APP, and the caregiver needs to confirm receipt online; the pushed content will highlight the risk warning, key time window and specific operation instructions.

6. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 5, characterized in that, The process of using a predictive model to infer potential non-standard events during the next care session, based on identified risk signals and combined with environmental factors such as the elderly person's medical history, age, season, and weather, includes the following steps: S331. Search historical data to find the group that is closest to the current elderly profile, and calculate the probability of various non-standard events occurring in this group during the next care. S332. Transform each elderly person's individual nursing record into a set of standardized feature data, which includes static features, dynamic features, and environmental features. Static features include age, gender, basic medical history, and ADL baseline. Dynamic features include recent blood pressure fluctuations, sleep quality, mood score, and medication adherence. Environmental features include season, sudden drop in temperature, and changes in air pressure. S333. Clearly define the non-standard events to be predicted, namely the probability of severe pain, the probability of emotional breakdown, and the probability of needing emergency medical intervention. S334. Based on the collected data, calculate the actual incidence of various non-standard events in this group during the next care session. The formula for the prediction model is: ; in, ; This represents the probability of a non-standard event occurring, with a value ranging from 0 to 1. For the input feature values, These are weighting coefficients, trained using historical big data, representing the degree of influence of each factor on risk. It is a natural constant. The base risk score of the model, i.e., the intercept term, is the baseline probability of an event occurring when all features are 0. S335. The numerical range of the model output is 0-1. The preset risk threshold is P≥0.

6. If the value exceeds 0.6, it will be automatically marked as high risk.

7. The intelligent scheduling method for elderly care service resources based on multi-dimensional profiling according to claim 1, characterized in that, The process of dividing the city into community grids and clustering the needs of elderly people within the same grid and at the same time period includes the following steps: Within the appointment time window, the algorithm plans the optimal multi-point connection route for dedicated service team caregivers, and dynamically inserts work intervals without travel distance as buffers based on the physical and emotional load of the previous task. S41. Based on the geographic information system and community administrative boundaries, the urban area is divided into several standardized service grids. The system automatically aggregates all elderly service requests within the same service period and located in the same grid to form a task pool to be scheduled for that grid. S42. From the established dedicated caregiver groups for the elderly, select caregiver members who are currently available and whose current location is in or near the grid, and combine the matching degree of the main and auxiliary caregivers' responsibilities and skills to initially allocate service requests in the task pool to suitable caregivers. S43. Under the premise of meeting the appointment time window of each elderly person, the path optimization algorithm is used to calculate the optimal geographical order for each caregiver to visit all their assigned service points, ensuring the shortest total travel time and generating a preliminary connected service route. S44. Read the type and duration of the previous task completed by the caregiver, and combine the preset physical exertion coefficient and emotional load assessment model to dynamically calculate the pure work buffer interval (excluding travel time) required by the caregiver before starting the next task. S44. The calculated personalized work buffer intervals are inserted between each service node in the preliminary path planning to form a final scheduling scheme that includes the complete sequence of movement-buffering-service, and to clarify the arrival time of the caregiver, the buffer start time and the service start time at each point.

8. A smart scheduling system for elderly care service resources based on multi-dimensional profiling, characterized in that, The system adopts a multi-dimensional profile-based intelligent scheduling method for elderly care service resources as described in any one of claims 1-7, including a dynamic multi-dimensional profile construction module, a primary and secondary caregiver matching module, a risk prediction-based nursing capacity push module, a service route planning module, a dynamic scheduling module, and a feedback evaluation module. The dynamic multidimensional profile construction module is responsible for integrating elderly health records, wearable device data, behavioral habits and care plans, completing data cleaning, standardization and structuring, and constructing a dynamically updated multidimensional personal profile of the elderly from five dimensions: physical health, self-care ability, behavioral patterns, psychosocial aspects and care needs; at the same time, it digitally models caregiver skills, location, scheduling, service capabilities and equipment resources. The primary and secondary caregiver matching module assesses the nursing needs and complexity level based on the elderly person's multi-dimensional profile, selects caregivers with matching skills and nearby locations from the resource pool, forms a dedicated service team of 3-5 people for each elderly person and assigns a primary caregiver; and determines the number of elderly people that the primary caregiver can also serve as a secondary caregiver based on the nursing complexity. The risk prediction nursing capability push module continuously analyzes risk factors in the elderly profile, identifies abnormal signals through risk models and machine learning, and predicts non-standard fluctuations and high-risk events. During high-risk periods, emergency plans and nursing skills are automatically matched to form standardized plan packages, which are then pushed to primary and secondary caregivers to improve the foresight of care and emergency response capabilities. The service route planning module clusters service demands for the same time period by community grid, and uses a path optimization algorithm to plan the optimal route connecting multiple points for caregivers within the appointment time window; combined with the physical and emotional load of the caregiver's previous task, it dynamically inserts pure work buffer intervals without routes, and generates a complete time sequence plan including movement, buffering and service. The dynamic scheduling module monitors the location of caregivers, task progress and the elderly person's condition in real time. In case of emergencies, it immediately performs dynamic rescheduling, prioritizing the replacement by auxiliary caregivers in the same group, or dispatching mobile caregivers from nearby grids to provide support, and synchronizing service records and service information. The feedback evaluation module collects user feedback, satisfaction, and core operational metrics after the service is completed, and comprehensively analyzes the scheduling effect and service quality. Based on the analysis results, it continuously optimizes the profile weights, risk models, path algorithms, and scheduling rules, driving the system to continuously iterate and upgrade.

Images