A method for automatically generating and allocating elderly care tasks

Abstract

The present invention relates to the technical field of automatic generation and allocation of elderly care tasks, and discloses a method for automatic generation and allocation of elderly care tasks. The overall process includes collecting basic data of the elderly and caregivers, defining original variables, standardizing the data and calculating the characteristic distance between the elderly; then adaptively determining the distance scale parameter, constructing a similarity matrix and extracting connected components to achieve clustering of the elderly's needs; on this basis, a greedy strategy is adopted to select the minimum number of caregivers to cover the service needs, and tasks are accurately allocated according to the ability ratio, and the results are output. This method realizes a closed loop of the entire process from data collection to task allocation, and has end-to-end automation capabilities. Through standardized modeling, similarity calculation and graph clustering, as well as a greedy optimization algorithm, this solution realizes intelligent nursing management in a resource-limited environment.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic generation and allocation of elderly care tasks, and in particular to a method for automatic generation and allocation of elderly care tasks. Background Art

[0002] With the accelerating aging of the population, elderly care institutions and home-based care service systems face multiple challenges, including matching caregiver supply and demand, providing personalized services, and optimizing human resources. Traditionally, the allocation of elderly care tasks relies primarily on the experience of managers and manual scheduling, often grouping and assigning care based on the age of the elderly, bed location, or simple needs (such as basic care and rehabilitation care). This approach presents challenges such as uneven workload, high service quality fluctuations, and inattention to individual differences among elderly individuals.

[0003] In existing technology, some institutions have introduced scheduling software that automatically generates caregiver schedules by setting fixed shifts and upper and lower limits on working hours. However, this type of software only performs linear scheduling based on daily total working hours, without granularly addressing the individual unit's individual needs. It also rarely considers the matching of elderly individuals' behavioral characteristics, health status, and the professional skills of caregivers. Furthermore, some studies have attempted to match elderly individuals with caregivers using fuzzy rules or heuristic algorithms. However, these often rely on manually set weighting parameters, which lack a unified standard and are unable to adapt to data changes, making it difficult to ensure the objectivity and fairness of the allocation scheme. With the widespread adoption of the Internet of Things and wearable devices, elderly individuals' daily activity data, such as step count, sleep duration, and toilet visits, can be collected in real time through smart bracelets or mattress sensors. However, only a few systems currently utilize this multidimensional, dynamic data for care task allocation. Most of these systems simply analyze health indicators or perform risk assessments during the offline analysis phase, failing to deeply explore similarities and differences among elderly individuals and failing to develop a continuous, closed-loop automated scheduling process. On the other hand, graph theory and clustering techniques have been successfully applied in fields such as social network analysis and logistics routing optimization. For example, graph partitioning methods based on community discovery can automatically divide massive nodes into several relatively independent subgroups; threshold selection strategies based on medians or means can achieve static segmentation of graph structures. However, these technologies remain at the academic exploration stage in elderly care allocation and have not yet been closely integrated with nursing workloads, care duration, and real-time needs. There is also no comprehensive algorithm for the entire process, from data collection and feature calculation to connected component extraction and resource allocation.

[0004] To this end, this project aims to propose a method for automatically generating and allocating elderly care tasks. By modeling and clustering the elderly's multidimensional behavioral data and integrating it with the caregiver's competency parameters, a data-driven intelligent task allocation mechanism is constructed. This solution fully integrates data collection, standardization, similarity construction, graph modeling, and greedy optimization strategies. It can scientifically group elderly people based on their similar care needs and, based on this, optimize the matching of personnel and tasks. Summary of the Invention

[0005] The present invention provides a method for automatically generating and allocating elderly care tasks, which helps solve the problems mentioned in the above background technology.

[0006] The present invention provides the following technical solution: a method for automatically generating and allocating elderly care tasks, comprising:

[0007] Collect basic data of the elderly and caregivers, and define original variables, including the elderly's daily behavioral characteristics and caregivers' service ability parameters;

[0008] Standardize the original feature data of the elderly to form a standardized feature vector;

[0009] Calculate the feature distance between any two elderly people based on standardized features;

[0010] Calculate the distance scale parameter based on the feature distance and make adaptive adjustments;

[0011] Construct a similarity matrix based on the distance scale parameter and calculate the similarity threshold;

[0012] Use the similarity matrix to construct an undirected graph and extract connected components;

[0013] A subset of caregivers that meet the needs of the elderly is selected based on a greedy strategy;

[0014] Assign nursing tasks based on the ratio of elderly people's needs to caregivers' abilities, and output the task assignment results.

[0015] Optionally, the basic data of the elderly and caregivers are collected and the original variables are defined, including:

[0016] Assume the total number of elderly people is , the index set of the elderly is ,index Corresponding to An old man, ;

[0017] The total number of caregivers is , the caregiver index set is ,index Corresponding to Caregivers, denoted as ;

[0018] For every elderly person Collect five-dimensional original features ,include:

[0019] , average daily steps in the past 7 days; , average daily sleep duration in the past 7 days; , average number of toilet visits per day in the past 7 days; , average number of meals per day in the past 7 days; , the average number of medications taken per day in the past 7 days; among them, is the original feature index;

[0020] Set up the elderly The duration of care required on that day is , caregiver The available service time for the day is .

[0021] Optionally, the step of normalizing the original feature data of the elderly to form a standardized feature vector specifically includes:

[0022] Calculate the The arithmetic mean of the original features ;

[0023] Calculate the The standard deviation of the original features ;

[0024] Constructing the Old Man The normalized eigenvector of , ;in, For the Dimensional normalization value.

[0025] Optionally, the calculating of the feature distance between any two elderly persons based on the standardized features specifically includes:

[0026] Index of any two elderly people , calculate the Euclidean distance between the two :

[0027] .

[0028] Optionally, the distance scale parameter is calculated based on the feature distance and adaptively adjusted, specifically including:

[0029] Set the overall distance scale parameter to .

[0030] Optionally, constructing a similarity matrix based on the distance scale parameter and calculating a similarity threshold specifically includes:

[0031] Set up the elderly and The demand similarity is , ;

[0032] Construct the similarity set as ;

[0033] Calculating the median threshold ;in, is the median function, Take the median value after sorting.

[0034] Optionally, the method of constructing an undirected graph using a similarity matrix and extracting connected components specifically includes:

[0035] Construct an adjacency indicator function:

[0036] ;in, Is an adjacency indicator variable if and only if the similarity between the two elderly people is not less than the threshold When it is 1, it means that in an undirected graph, the vertex and There is an edge between them; otherwise it is 0, no edge;

[0037] Initialize the unvisited vertex index set , connected component counter ;

[0038] when Repeat the following steps:

[0039] S1100, command ;

[0040] S1200, from Get the minimum index , let the current component vertex set , and from Delete in;

[0041] S1300, set pointer set ;

[0042] S1400, when hour:

[0043] S1410, optional , and change it from Delete in;

[0044] S1420, for all Check, if , then execute:

[0045] S1421, will join in ;

[0046] S1422, will join in ;

[0047] S1423, from Delete ;

[0048] S1500, until hour, That is the The vertex set of connected components, and its edge set is:

[0049] ;

[0050] S1600, repeat steps S1100 to S1500 until ;

[0051] Finally, we get a set of connected components .

[0052] Optionally, the greedy strategy is used to select a subset of caregivers that meet the needs of the elderly, specifically including:

[0053] For each component Perform the following steps:

[0054] S201, calculation of components Total care needs of all elderly people ;

[0055] S202, will Middle Index Press Sort from large to small, and remember the sequence after sorting is ,satisfy ;in, For the Index of available hours for caregivers;

[0056] S203, Order , calculate the cumulative available time ;in, is a counter used to record the number of members currently selected into the caregiver subset;

[0057] like ,make Repeat; otherwise stop;

[0058] S204, take the required quantity Minimum set of caregivers required , ;in, For collection The total available time.

[0059] Optionally, the method allocates care tasks based on the ratio of the elderly's needs to the caregiver's abilities and outputs the task allocation results, specifically including:

[0060] For each Each caregiver calculate: ;in, For the weight middle-aged and elderly people Assigned to a caregiver Length of service;

[0061] verify:

[0062] S301. Elderly needs satisfaction: ;

[0063] S302. Caregiver overload: ;

[0064] For each component Output:

[0065] Elderly group list , Caregiver subset and the allocation matrix .

[0066] The present invention has the following beneficial effects:

[0067] 1. The solution proposes a comprehensive reflection of the elderly's daily lives across five dimensions (walking, sleeping, toileting, dining, and medication). This approach significantly outperforms existing static demand collection methods based solely on questionnaires or subjective ratings. This feature definition approach, based on time-series behavioral data, offers high resolution and quantifiable characteristics, accurately reflecting the actual care intensity needs of the elderly. Most traditional systems rely on tiered care levels, lacking granular data support and making it difficult to achieve differentiated services. This solution constructs a multidimensional, structured input space, providing a solid data foundation for subsequent algorithms.

[0068] 2. By calculating the mean and standard deviation, features across all dimensions are normalized to a zero-mean, unit-variance space, resolving the incomparability between features of different dimensions. This step significantly improves the accuracy and fairness of subsequent distance calculations and is a prerequisite for standardization in the application of data mining and machine learning algorithms. Traditional elderly care task allocation methods often use a scoring system, which is highly subjective and suffers from inconsistent dimensionality. This solution constructs an objectively comparable feature space through mathematical normalization, making model inputs more logical and consistent.

[0069] 3. Euclidean distance is used to measure the comprehensive differences between elderly individuals, and a global root mean square (RMS)-based distance scale adaptation mechanism is further proposed. This mechanism eliminates the need for empirical threshold settings and automatically adjusts the similarity conversion scale based on the actual data distribution, effectively improving the model's versatility and robustness. Existing research often relies on a fixed distance scale for similarity calculation, which is susceptible to interference from outliers. This solution uses a statistical adaptive method to automatically adapt to the differences in datasets in different scenarios, significantly enhancing the system's environmental adaptability.

[0070] 4. Convert the Euclidean distance into a similarity score between 0 and 1, and use the median as the threshold to avoid bias introduced by manually setting parameters. This dynamic threshold method based on the statistical median can achieve efficient clustering with close clustering within and separate clusters between. Traditional methods are often based on empirical thresholds or preset clustering K values ​​and are not adaptive. This method's data-driven segmentation strategy significantly improves the scientific and rationality of elderly grouping.

[0071] 5. Elderly individuals with similarities above a threshold are edge-generated on an undirected graph, and a connected component extraction algorithm is used to automatically identify groups of elderly individuals with similar care needs. This graph-based modeling approach is both efficient and intuitive, clearly defining the units for task allocation. Most existing allocation systems only perform individual assessments, ignoring the group nature of elderly individuals. This solution, focusing on the "group-caregiver" dimension, significantly improves the cohesion and accuracy of task grouping.

[0072] 6. Based on clustering, a greedy algorithm is introduced, gradually covering the needs of caregivers, starting with those with high service capabilities, until a minimum set is reached. This heuristic selection method balances efficiency and resource optimization, making it particularly suitable for resource-constrained environments. Common scheduling systems often rely on rotation or equal sharing strategies, ignoring the matching of task load and human resources. This solution enables dynamic, on-demand scheduling, significantly improving the utilization and flexibility of nursing resources.

[0073] 7. A "proportional" allocation mechanism is introduced to ensure that caregivers' service capacity is proportional to their available hours, and the allocation results are validated for compliance (e.g., ensuring demand is met and overload is avoided). Task output is structured as a "cluster list + caregiver subset + allocation matrix," enabling direct integration with the scheduling system. Traditional systems have a crude allocation mechanism, lacking support for many-to-many mapping and detailed task output. This solution, with its comprehensive mathematical basis and output structure, is the core support for the intelligent upgrade of the nursing scheduling system. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0075] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0076] Example, see Figure 1 , a method for automatically generating and allocating elderly care tasks, comprising:

[0077] Collect basic data of the elderly and caregivers, and define original variables, including the elderly's daily behavioral characteristics and caregivers' service ability parameters;

[0078] Standardize the original feature data of the elderly to form a standardized feature vector;

[0079] Calculate the feature distance between any two elderly people based on standardized features;

[0080] Calculate the distance scale parameter based on the feature distance and make adaptive adjustments;

[0081] Construct a similarity matrix based on the distance scale parameter and calculate the similarity threshold;

[0082] Use the similarity matrix to construct an undirected graph and extract connected components;

[0083] A subset of caregivers that meet the needs of the elderly is selected based on a greedy strategy;

[0084] Assign nursing tasks based on the ratio of elderly people's needs to caregivers' abilities, and output the task assignment results.

[0085] First, behavioral and capability data of the elderly and caregivers are collected to provide basic input for subsequent model calculations. Then, through data processing methods such as feature standardization, similarity measurement, and graph construction, the elderly's demand features are clustered. A greedy algorithm is then used to select nursing resources and accurately allocate nursing tasks based on capability ratios. This method achieves a closed loop from "data collection" to "task output" and possesses end-to-end automation capabilities. Through standardized modeling, graph partitioning, and heuristic optimization, this solution addresses the problems of extensive, irrational, and slow-response task allocation in traditional elderly care. The system can intelligently identify groups with similar nursing needs and optimally allocate caregivers under resource-constrained conditions, achieving dynamic optimization and refined management of nursing tasks, improving service efficiency and fairness, and possessing high practical value.

[0086] The basic data of the elderly and caregivers are collected and the original variables are defined, including:

[0087] Assume the total number of elderly people is , the index set of the elderly is ,index Corresponding to An old man, ; Identify the number of elderly people in the system and their number range to ensure the orderliness of subsequent processing;

[0088] The total number of caregivers is , the caregiver index set is ,index Corresponding to Caregivers, denoted as ; Identify the number of caregivers and their number ranges to support the capacity allocation process;

[0089] For every elderly person Collect five-dimensional original features ,include:

[0090] , average daily steps in the past 7 days; , average daily sleep duration in the past 7 days; , average number of toilet visits per day in the past 7 days; , average number of meals per day in the past 7 days; , the average number of medications taken per day in the past 7 days; among them, Indexes the original features; provides quantitative demand and behavior indicators, and inputs the original data for subsequent similarity calculations;

[0091] Set up the elderly The duration of care required on that day is , caregiver The available service time for the day is ; Quantify the needs of the elderly and the available resources of caregivers, and build allocation boundaries.

[0092] The specific variable structure of "collected raw data" is defined in detail, including the number of elderly people, the number of caregivers, five-dimensional behavioral characteristics, nursing needs and available service time. By establishing a unified data index set and variable system, the standardization of model input data and the orderliness of calculation are ensured. By systematically collecting five core behavioral characteristics, this method establishes a solid data foundation for subsequent similarity calculation and graph clustering from a quantitative level. Compared with the traditional method of dividing nursing needs based on subjective evaluation, this technology has significant advantages such as objectivity, real-time and scalability, laying the foundation for data-driven intelligent scheduling and improving modeling accuracy and system stability.

[0093] The standardization of the original feature data of the elderly to form a standardized feature vector specifically includes:

[0094] Eliminate the influence of different dimensions and map the five-dimensional features into a unified dimensionless space;

[0095] Calculate the The arithmetic mean of the original features ; Obtain the center location of the dimension data for debiasing;

[0096] Calculate the The standard deviation of the original features ;Measure the discreteness of the dimension data for scaling;

[0097] Constructing the Old Man The normalized eigenvector of , ;in, For the Dimensional standardization value; normalize the original features to zero mean and unit variance to ensure that each dimension is equally important.

[0098] The five-dimensional raw features are standardized, converting data of different dimensions into a standard normal distribution space with zero mean and unit variance. This process effectively eliminates the dimensional influence between the feature dimensions, making the subsequent Euclidean distance calculation more reasonable and fair. Through standardization, the problem of "a certain dimension dominating the distance calculation due to its large value" is avoided. This operation ensures the equal weight of each dimension in the calculation, improves the model's ability to identify the true differences in elderly care needs, enhances the scientific nature and discriminative power of the feature space, and is a key technical step in ensuring the accuracy of clustering and allocation.

[0099] The step of calculating the feature distance between any two elderly people based on the standardized features specifically includes:

[0100] The original measure of feature similarity between elders is measured using Euclidean distance;

[0101] Index of any two elderly people , calculate the Euclidean distance between the two :

[0102] ;

[0103] Quantify the overall difference in the standardized features of the two elderly people as the basis for subsequent similarity conversion.

[0104] Euclidean distance is used to calculate pairwise normalized eigenvectors to measure the similarity of care needs between elderly individuals. This serves as the foundation for similarity calculation and subsequent graph construction. As a fundamental form of measuring differences between continuous variables, the introduction of Euclidean distance provides an intuitive and controllable quantitative indicator for systematically constructing a "model of elderly needs differences." This allows elderly care institutions to more accurately determine the similarity of services provided to elderly individuals, addressing the inability of traditional "one-size-fits-all" grouping methods to adapt to individual differences and improving personalized services.

[0105] The distance scale parameter is calculated based on the feature distance and adaptively adjusted, specifically including:

[0106] The scaling factor is defined by the full distance data to ensure that the numerical range of the similarity function is reasonable;

[0107] Set the overall distance scale parameter to ;

[0108] Take the root mean square of all distances and automatically adapt to the data distribution without manual setting.

[0109] The root mean square (RMS) of all distance values ​​is calculated and used as a distance scaling factor to achieve adaptive rescaling of the feature space. This mechanism prevents distance calculation distortion caused by uneven distribution of input data. By adaptively adjusting the similarity scale based on a global RMS, this method avoids the adaptability issues associated with static parameters and improves the versatility and stability of the model. This approach ensures the consistency and effectiveness of the clustering logic, particularly in scenarios with varying nursing homes of varying sizes and data sampling frequencies, significantly enhancing the model's robustness and cross-environmental deployment capabilities.

[0110] The similarity matrix is ​​constructed based on the distance scale parameter, and the similarity threshold is calculated, specifically including:

[0111] Mapping distance to Similarity within the interval, and adaptively select the partition threshold;

[0112] Set up the elderly and The demand similarity is , ; Convert the distance value to similarity. The smaller the distance, the closer the similarity is to 1.

[0113] Construct the similarity set as ;Collect all non-self-loop similarities for statistical analysis;

[0114] Calculating the median threshold ;in, is the median function, Take the median after sorting; select the median of similarity as the division threshold to ensure balance within and between groups.

[0115] Mapping Euclidean distance to The algorithm uses the median of all similarities to determine the partitioning threshold, ensuring high similarity within a cluster and low similarity between clusters. Using the median, rather than an empirical value, as the clustering threshold improves the objectivity and robustness of the partitioning, solving the problem of difficult-to-determine cluster boundaries. This makes it particularly suitable for automatic classification of large-scale data. Compared to static clustering algorithms such as K-means, it does not require a preset number of clusters and can more realistically reflect the natural distribution of care needs among elderly individuals.

[0116] The method of constructing an undirected graph using a similarity matrix and extracting connected components specifically includes:

[0117] Construct a similarity threshold graph and accurately extract connected components;

[0118] Construct an adjacency indicator function:

[0119] ;in, Is an adjacency indicator variable if and only if the similarity between the two elderly people is not less than the threshold When it is 1, it means that in an undirected graph, the vertex and If yes, there is an edge between them; otherwise, it is 0, no edge is connected; determine whether the two elderly people have established an edge connection in the graph;

[0120] Initialize the unvisited vertex index set , connected component counter ; Prepare unvisited node set and connected component counter;

[0121] when Repeat the following steps:

[0122] S1100, command ; Start Connected component extraction;

[0123] S1200, from Get the minimum index , let the current component vertex set , and from Delete; initialize the component vertex set;

[0124] S1300, set pointer set ; Used to iteratively expand vertices;

[0125] S1400, when hour:

[0126] S1410, optional , and change it from Delete in;

[0127] S1420, for all Check, if , then execute:

[0128] S1421, will join in ;

[0129] S1422, will join in ;

[0130] S1423, from Delete ;

[0131] Expand component vertices expansively until the component is fully formed;

[0132] S1500, until hour, That is the The vertex set of connected components, and its edge set is:

[0133] ; Record the legal edges between all vertices in the component;

[0134] S1600, repeat steps S1100 to S1500 until ;Extract all connected components;

[0135] Finally, we get a set of connected components .

[0136] An undirected graph is constructed based on the similarity matrix, and a connected component extraction algorithm is then used to identify groups of elderly people with similar needs. This method transforms data into a structure through graph theory, enhancing information organization. Compared to traditional clustering methods, graph models are more capable of preserving complex relationships and local structural information. The extraction of connected components enables structured grouping of nursing tasks, effectively supporting local optimization and parallel processing of subsequent task allocation, improving the responsiveness of nursing resource scheduling and the scalability of the algorithm. This represents a cutting-edge modeling approach for group management.

[0137] The greedy strategy is used to select a subset of caregivers that meet the needs of the elderly, specifically including:

[0138] In each elderly group, the total demand is met with the least number of caregivers;

[0139] For each component Perform the following steps:

[0140] S201, calculation of components Total care needs of all elderly people ; Calculate the total care needs of all elderly people under this component;

[0141] S202, will Middle Index Press Sort from large to small, and remember the sequence after sorting is ,satisfy ;in, For the The index of the available time caregivers; prepare the basis for greedy selection;

[0142] S203, Order , calculate the cumulative available time ;in, is a counter used to record the number of members currently selected into the caregiver subset;

[0143] like ,make Repeat; otherwise stop;

[0144] Add caregivers with the highest available hours one by one until the demand is covered;

[0145] S204, take the required quantity Minimum set of caregivers required , ;in, For collection The total available time of the component; output the minimum set of caregivers required for this component and their total resources.

[0146] Within each cluster, a greedy strategy is employed to select the minimum number of caregivers needed to cover all residents' needs, prioritizing caregivers with the longest service time to improve resource efficiency. This greedy strategy minimizes resource input while ensuring coverage, reducing the number of caregivers required and ensuring task satisfaction. This "meeting maximum needs with minimal resources" mechanism is particularly well-suited for nursing homes, where labor costs are high and resource allocation is limited, significantly optimizing resource allocation and operating costs.

[0147] The system allocates care tasks based on the ratio of the elderly's needs to the caregiver's abilities and outputs the task allocation results, specifically including:

[0148] Allocate specific hours within each cluster based on the caregiver's capacity ratio and verify that there is no overload or surplus;

[0149] For each Each caregiver calculate: ;in, For the weight middle-aged and elderly people Assigned to a caregiver The service time is proportional to the available time of the caregivers; the needs of the elderly are allocated according to the proportion of the caregivers’ available time to ensure a fair ratio;

[0150] verify:

[0151] S301. Elderly needs satisfaction: ;

[0152] S302. Caregiver overload: ;

[0153] Ensure that allocation is legal, all elderly people receive services, and caregivers are not overloaded;

[0154] For each component Output:

[0155] Elderly group list , Caregiver subset and the allocation matrix ; For direct call and execution by the scheduling system or management personnel.

[0156] Finally, nursing tasks are allocated proportionally to caregivers' available time, and a dual constraint verification mechanism (ensuring that the needs of the elderly are met and that caregivers are not overloaded) is implemented to ensure the fairness and legality of task allocation. This proportional allocation and constraint verification ensure the feasibility of task allocation and the security of resource utilization, effectively preventing caregiver overload and missed elderly care services. The structured allocation matrix output also facilitates direct integration into scheduling systems, forming a closed-loop automated task management process with strong system compatibility and practicality.

[0157] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0158] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for automatically generating and allocating elderly care tasks, characterized in that: include: Collect basic data of the elderly and caregivers, and define original variables, including the elderly's daily behavioral characteristics and caregivers' service ability parameters; Standardize the original feature data of the elderly to form a standardized feature vector; Calculate the feature distance between any two elderly people based on standardized features; Calculate the distance scale parameter based on the feature distance and make adaptive adjustments; Construct a similarity matrix based on the distance scale parameter and calculate the similarity threshold; Use the similarity matrix to construct an undirected graph and extract connected components; A subset of caregivers that meet the needs of the elderly is selected based on a greedy strategy; Assign nursing tasks based on the ratio of elderly people's needs to caregivers' abilities, and output the task assignment results.

2. The method for automatically generating and allocating elderly care tasks according to claim 1, characterized in that: The basic data of the elderly and caregivers are collected and the original variables are defined, including: Assume the total number of elderly people is , the index set of the elderly is ,index Corresponding to An old man, ; The total number of caregivers is , the caregiver index set is ,index Corresponding to Caregivers, denoted as ; For every elderly person Collect five-dimensional original features ,include: , average daily steps in the past 7 days; , average daily sleep duration in the past 7 days; , average number of toilet visits per day in the past 7 days; , average number of meals per day in the past 7 days; , the average number of medications taken per day in the past 7 days; among them, is the original feature index; Set up the elderly The duration of care required on that day is , caregiver The available service time for the day is .

3. The method for automatically generating and allocating elderly care tasks according to claim 2, characterized in that: The standardization of the original feature data of the elderly to form a standardized feature vector specifically includes: Calculate the The arithmetic mean of the original features ; Calculate the The standard deviation of the original features ; Constructing the Old Man The normalized eigenvector of , ;in, For the Dimensional normalization value.

4. The method for automatically generating and allocating elderly care tasks according to claim 3 is characterized in that: The step of calculating the feature distance between any two elderly people based on the standardized features specifically includes: Index of any two elderly people , calculate the Euclidean distance between the two : 。 5. The method for automatically generating and allocating elderly care tasks according to claim 4 is characterized in that: The distance scale parameter is calculated based on the feature distance and adaptively adjusted, specifically including: Set the overall distance scale parameter to .

6. The method for automatically generating and allocating elderly care tasks according to claim 5, characterized in that: The similarity matrix is ​​constructed based on the distance scale parameter, and the similarity threshold is calculated, specifically including: Set up the elderly and The demand similarity is , ; Construct the similarity set as ; Calculating the median threshold ;in, is the median function, Take the median value after sorting.

7. The method for automatically generating and allocating elderly care tasks according to claim 6, characterized in that: The method of constructing an undirected graph using a similarity matrix and extracting connected components specifically includes: Construct an adjacency indicator function: ;in, Is an adjacency indicator variable if and only if the similarity between the two elderly people is not less than the threshold When it is 1, it means that in an undirected graph, the vertex and There is an edge between them; otherwise it is 0, no edge; Initialize the unvisited vertex index set , connected component counter ; when Repeat the following steps: S1100, command ; S1200, from Get the minimum index , let the current component vertex set , and from Delete in; S1300, set pointer set ; S1400, when hour: S1410, optional , and change it from Delete in; S1420, for all Check, if , then execute: S1421, will join in ; S1422, will join in ; S1423, from Delete ; S1500, until hour, That is the The vertex set of connected components, and its edge set is: ； S1600, repeat steps S1100 to S1500 until ; Finally, we get a set of connected components .

8. The method for automatically generating and allocating elderly care tasks according to claim 7, characterized in that: The greedy strategy is used to select a subset of caregivers that meet the needs of the elderly, specifically including: For each component Perform the following steps: S201, calculation of components Total care needs of all elderly people ; S202, will Middle Index Press Sort from large to small, and remember the sequence after sorting is ,satisfy ;in, For the Index of available hours for caregivers; S203, Order , calculate the cumulative available time ;in, is a counter used to record the number of members currently selected into the caregiver subset; like ,make Repeat; otherwise stop; S204, take the required quantity Minimum set of caregivers required , ;in, For collection The total available time.

9. The method for automatically generating and allocating elderly care tasks according to claim 8, characterized in that: The system allocates care tasks based on the ratio of the elderly's needs to the caregiver's abilities and outputs the task allocation results, specifically including: For each Each caregiver calculate: ;in, For the weight middle-aged and elderly people Assigned to a caregiver Length of service; verify: S301. Elderly needs satisfaction: ; S302. Caregiver overload: ; For each component Output: Elderly group list , Caregiver subset and the allocation matrix .

Images