Co-tenancy pension service system and method based on multi-dimensional matching and service resource collaborative scheduling

Abstract

The embodiment of the application discloses a kind of based on multi-dimension matching and service resource collaborative scheduling's co-renting old-age service system and method.System includes: co-renting matching subsystem, through multi-dimension matching score engine calculates four-dimensional score of nursing efficiency complement, social compatibility, geographical position adaptation and residence environment matching, and based on co-renting feedback data self-adapting optimization weight coefficient, weighted recommendation optimal combination;Service resource collaborative scheduling and supervision subsystem, maintain configurable artificial, robot and other heterogeneous resource pool, generate collaborative scheduling, realize quality supervision without discrimination, and integrate environmental safety and old person state monitoring multi-level alarm and scheduling linkage;Intelligent cost allocation subsystem, according to service provider type and actual use condition fine cost allocation.

Description

Technical Field

[0001] This invention relates to the field of smart elderly care and resource scheduling technology, specifically to a co-living elderly care service system and method based on multi-dimensional matching and collaborative scheduling of service resources. Background Technology

[0002] The co-living elderly care (or "group living for the elderly") model has emerged because it can significantly reduce the cost of individual elderly care, but its large-scale promotion faces the following fundamental pain points caused by the shortcomings of existing technologies:

[0003] Pain point 1: The matching criteria for roommates are too limited, resulting in poor group stability.

[0004] Existing matching technologies rely solely on similarity calculations of interests or consider only simple conditions such as geographical location, completely ignoring the comprehensive feasibility of co-living for elderly care. On one hand, ignoring whether peak periods of care demand overlap may lead to inefficiency in shared services and increased shared costs; on the other hand, ignoring personality compatibility may cause interpersonal conflicts. Furthermore, significant conflicts in living environment preferences (such as requirements for quietness, lighting, and accessibility) will also affect long-term stable residence. Moreover, the weights of each dimension in existing matching schemes rely on human experience and cannot adapt to the differentiated needs of different regions and populations. Currently, there is no matching method that can simultaneously quantify multiple dimensions such as "complementarity of care service efficiency," "social compatibility," "geographical location suitability," and "living environment suitability," and possess the ability to adaptively optimize weights.

[0005] Pain Point 2: Inflexible service resource allocation and a lack of systematic guarantees for service quality and security.

[0006] Current elderly care service scheduling solutions only target human caregivers, and quality supervision relies entirely on post-event subjective user evaluations. As nursing robots and intelligent automated equipment mature, future co-living elderly care scenarios will inevitably introduce diverse service providers. Existing technologies cannot uniformly schedule, collaboratively manage, and standardize the quality supervision of heterogeneous service resources such as human caregivers, nursing robots, and automated equipment, leading to resource silos. Furthermore, key service parameters (such as bathing water temperature, the number and intensity of rehabilitation movements) lack real-time data collection and objective evaluation methods, resulting in frequent and difficult-to-trace service quality disputes. In addition, elderly people in co-living elderly care face safety risks such as falls, sudden illnesses, fires, and gas leaks. Existing systems lack technical solutions to integrate environmental safety monitoring and elderly person status monitoring into service scheduling, leading to untimely responses to safety incidents and severely impacting the feasibility and trustworthiness of the co-living model.

[0007] Pain Point 3: The cost allocation is too broad and cannot fairly cover the usage costs of different types of service resources.

[0008] Existing cost-sharing tools for shared housing can only handle fixed rent payments based on area or shared daily necessities per person. Variable costs such as water, electricity, and gas are generally shared equally, resulting in poor fairness. Caregiver service fees are simply shared based on the length of service, failing to consider the different costs incurred by elderly people with varying degrees of disability. More importantly, when introducing new service resources such as care robots, how to fairly distribute their depreciation, energy, and other operating costs among shared housing members is a completely unresolved issue in current technology.

[0009] The aforementioned pain points are interconnected and none of them can be solved by existing single technical solutions, constituting the core technical obstacles hindering the large-scale development of the co-living elderly care model. Summary of the Invention

[0010] The purpose of this invention is to provide a shared elderly care service system and method based on multi-dimensional matching and collaborative scheduling of service resources. By introducing a multi-dimensional joint optimization matching engine covering complementary nursing efficiency, social compatibility, geographical location adaptability, and living environment matching, coupled with a configurable heterogeneous service resource pool and collaborative scheduling, it covers unified quality supervision and safety risk monitoring of different service providers, as well as a refined cost-sharing model. This systematically solves the problems of single matching dimensions, rigid resource scheduling, lack of objective quality and safety guarantees, and coarse cost-sharing, providing a technical closed-loop solution for shared elderly care that adapts to current needs and can be smoothly expanded.

[0011] To achieve the above objectives, in a first aspect, embodiments of the present invention provide a shared elderly care service system based on multi-dimensional matching and collaborative scheduling of service resources, comprising:

[0012] The co-living matching subsystem is equipped with a multi-dimensional matching scoring engine, which calculates the complementarity score of nursing efficiency, the social compatibility score, the geographical location suitability score, and the living environment matching score based on the user's nursing needs time-series vector, social feature vector, geographical location information, and living environment preference vector, respectively. It then generates a comprehensive matching score through normalized weighted summation to recommend co-living combinations.

[0013] The service resource collaborative scheduling and supervision subsystem is equipped with a service resource management module, a collaborative scheduling module, a unified service quality assessment module, and a security and risk monitoring module. It is used to realize unified scheduling, collaborative scheduling, quantitative quality supervision of service processes, and security monitoring of the shared rental environment for different types of service resources.

[0014] The intelligent cost allocation subsystem is equipped with a service fee allocation unit, which is used to calculate the service fees that each user should share based on the duration of service received by each user from different service providers and the corresponding cost parameters, and generate a comprehensive allocation bill.

[0015] The data storage unit is used to store user profiles, demand vectors, scheduling schemes, quality assessment data, security logs, metering data, matching weight parameters, allocation rules and settlement records, providing unified data support for each subsystem.

[0016] As a specific implementation of the present invention, the nursing efficiency complementarity calculation is specifically as follows: extract the time slots of serial exclusive service items in the time sequence vector of each user's nursing needs, calculate the time overlap between all user pairs in the combination on this type of time slot as the demand conflict degree, the lower the conflict degree, the higher the nursing efficiency complementarity score, and introduce the nursing demand level difference coefficient.

[0017] The social compatibility calculation specifically involves: calculating the personality compatibility index between each user pair based on a preset personality dimension compatibility matrix, calculating the interest compatibility index based on the overlap of interest tags, and then weighting and summing the two to generate a social compatibility score.

[0018] The location fit calculation specifically involves: calculating the geographical distance between each user's current residence and the degree of overlap in their preferences for the target shared rental area. The closer the distance and the higher the degree of overlap in preferences, the higher the location fit score.

[0019] The calculation of the living environment matching degree specifically involves calculating the similarity between the living environment preference vectors of each user. The higher the similarity, the higher the living environment matching degree score. The living environment preference vector includes scores for room orientation, floor, quietness, lighting conditions, and accessibility facilities.

[0020] In a preferred embodiment of the present invention, the multi-dimensional matching scoring engine further includes a matching weight optimization module, used for:

[0021] The stability data of the formed co-tenancy combination within a preset period is collected. The stability data is used as a supervision signal, and the scores of each dimension of the co-tenancy combination are used as input features. The gradient descent method or Bayesian optimization algorithm is used to iteratively update the weight coefficients of the normalized weighted summation.

[0022] As a specific implementation of the present invention, the service resource management module is used to maintain a service resource pool containing at least one type of service provider; the service provider type includes at least one of human caregivers, nursing robots and intelligent automated equipment; the service resource pool can be configured to contain only human caregivers and operate in a pure human service mode, or to contain multiple types and operate in a collaborative service mode.

[0023] The collaborative scheduling module is used to allocate the service needs of the shared rental unit to suitable service providers based on the service providers' capability parameters, and generate a conflict-free joint service schedule.

[0024] The unified service quality assessment module is used to collect key operational parameters during the service execution process through smart devices, compare them with preset unified quality standard thresholds applicable to different service provider types, and generate service process compliance and quality assessment results.

[0025] The safety and risk monitoring module is used to detect safety events and emergency states in real time through environmental safety sensors and elderly status sensing devices, generate alarms, and when an alarm is triggered, the collaborative scheduling module suspends the currently affected services, reassesses the availability of service resources, and generates an adjusted service schedule to prioritize the response to alarm events.

[0026] Furthermore, as a specific implementation method, the unified service quality assessment module includes:

[0027] The perception data acquisition interface is used to connect services performed by human caregivers with intelligent nursing equipment or vision sensors deployed at the service site. For services performed by nursing robots or automated equipment, execution parameters are directly obtained through their control interface.

[0028] The service compliance assessment unit is used to provide a unified key quality indicator threshold for all services, and the same threshold applies to different types of service providers.

[0029] Furthermore, as a specific implementation, the security and risk monitoring module includes:

[0030] An environmental safety sensing unit is connected to environmental safety sensors deployed in shared housing to detect at least one environmental safety event, including smoke, gas leaks, water immersion, and illegal intrusion, and generate environmental safety alarms in real time.

[0031] An elderly condition monitoring unit is connected to at least one of a fall detector, a vital signs monitoring device, and an emergency call device deployed in a shared housing unit, and is used to detect at least one emergency state of elderly people falling, abnormal heart rate, or prolonged stillness and generate an emergency rescue alarm.

[0032] The multi-level alarm and linkage response unit is configured to push alarm information to the caregiver terminal, the preset emergency contact terminal for children, and / or the emergency service platform, and to link with the collaborative scheduling module.

[0033] As a specific implementation, in the service fee sharing unit, the service fee for human caregivers is shared according to the weighted proportion of the duration each user receives the human service, multiplied by a preset nursing need level weighting coefficient; the service fee for nursing robots or intelligent automated equipment is shared according to the proportion of the actual time each user occupies the equipment.

[0034] Furthermore, as a preferred implementation of the present invention, the system also includes a dynamic management module for co-tenants, which is used to automatically trigger the rescheduling of services for the remaining members, the settlement of fees for the departing members, and the invocation of the multi-dimensional matching and scoring engine to recommend new members when a member is detected to have left.

[0035] Secondly, embodiments of the present invention also provide a method for co-living elderly care services based on multi-dimensional matching and collaborative scheduling of service resources, including:

[0036] Collect users' nursing needs time-series vector, social feature vector, geographical location information, and residential environment preference vector; calculate scores for nursing efficiency complementarity, social compatibility, geographical location adaptability, and residential environment matching; generate a comprehensive matching score by normalizing and weighting the sum; and recommend co-living combinations; and iteratively optimize the weight coefficients of the weighted sum based on the stability data of co-living feedback.

[0037] After establishing a co-tenancy relationship, the service requirements of the co-tenancy unit are allocated to suitable service providers based on at least one type of service provider and their capability parameters available in the service resource pool, thereby generating a conflict-free joint service schedule.

[0038] During service execution, operation parameters are collected through smart devices and compared with preset unified quality standard thresholds to generate service quality assessment results and trigger necessary warnings. At the same time, safety events and emergency states are monitored in real time through environmental safety sensors and elderly status sensing devices, alarms are generated, and when an alarm is triggered, the currently affected service is suspended, the availability of service resources is reassessed, and an adjusted service schedule is generated.

[0039] During the settlement period, the service fee allocation is calculated based on the duration of service received by each user from different service providers and the corresponding cost parameters, and a comprehensive cost bill is generated by combining the allocation of variable usage.

[0040] Implementing the embodiments of the present invention has the following beneficial effects:

[0041] 1. Multidimensional matching significantly improves the stability and economic feasibility of co-living: For the first time, four dimensions—complementary care efficiency, social compatibility, geographical location suitability, and living environment matching—are incorporated into a unified matching model. An innovative weighted adaptive optimization mechanism based on real-world co-living feedback is introduced to ensure optimal service costs, interpersonal harmony, relocation convenience, and living comfort for co-living groups from the outset. Calculations show that the average care cost per person can be reduced by 15%-25% after matching, and the breakup rate due to conflict can be reduced by more than 70%.

[0042] 2. Collaborative scheduling and seamless expansion of heterogeneous service resources: The configurable design of the service resource pool allows the system to currently operate as a purely manual co-living elderly care management platform. In the future, it can be seamlessly upgraded to a human-machine collaborative or automated service mode simply by registering new devices, thus protecting users' long-term investment.

[0043] 3. Unified service quality supervision without discrimination: Regardless of whether the service is performed by humans, robots or automated equipment, parameters are collected through a unified sensor interface and compared with the same standard thresholds to ensure the objectivity of quality evaluation and solve the problem of difficulty in obtaining evidence in service disputes.

[0044] 4. Comprehensive safety and risk protection: The system integrates environmental safety monitoring and elderly status monitoring, and innovatively links safety alarms with service scheduling to ensure that the elderly can receive assistance immediately in case of falls, sudden illnesses or dangerous environments, without affecting the basic services of other elderly people, effectively reducing the safety risks of co-living elderly care.

[0045] 5. Refined cost allocation across the entire project: For the first time, equipment usage costs are included in the allocation based on the duration of use. Combined with intelligent metering and behavioral estimation models to handle variable consumption and nursing level weighted processing of labor costs, this constitutes the most comprehensive cost-sharing scheme for co-living elderly care.

[0046] 6. System-level combinatorial innovation and continuous evolution capability: It organically integrates multi-dimensional matching, heterogeneous resource collaborative scheduling, unified quality perception and security monitoring, and comprehensive cost allocation to construct a closed-loop data flow throughout the entire process: "matching results drive scheduling → scheduling determines monitoring strategies → monitoring and security data are jointly input into the cost allocation model → feedback from the co-working group optimizes matching weights." Its synergistic effects and adaptive evolution capabilities are not publicly disclosed in existing technologies, demonstrating outstanding substantive characteristics and significant progress. Attached Figure Description

[0047] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below.

[0048] Figure 1 This is an overall interactive architecture diagram of the co-living elderly care service system based on multi-dimensional matching and service resource collaborative scheduling provided in this embodiment of the invention;

[0049] Figure 2 This is a flowchart of the multi-dimensional matching scoring engine calculation process (including weight adaptive optimization).

[0050] Figure 3 This is a schematic diagram of the task classification and resource allocation logic of the collaborative scheduling module;

[0051] Figure 4It is a unified service quality assessment and security monitoring module monitoring logic diagram;

[0052] Figure 5 This is a data flow diagram of the intelligent cost allocation subsystem;

[0053] Figure 6 This is a flowchart of a co-living elderly care service method based on multi-dimensional matching and collaborative scheduling of service resources provided in an embodiment of the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0056] Please refer to Figure 1 This invention provides a shared elderly care service system based on multi-dimensional matching and collaborative scheduling of service resources. Deployed on a cloud server, it connects user terminals, service execution terminals, various intelligent sensing devices, and intelligent metering devices via a communication network. The system includes a shared accommodation matching subsystem, a service resource collaborative scheduling and monitoring subsystem, a cost intelligent allocation subsystem, and a data storage unit. Each part is described in detail below.

[0057] (a) Co-renting Matching Subsystem

[0058] This subsystem incorporates a multi-dimensional matching and scoring engine to comprehensively match high-stability, low-cost roommates based on four dimensions: care efficiency, social relationships, geographical location, and living environment. The calculation process of the multi-dimensional matching and scoring engine is as follows: Figure 2 As shown.

[0059] More specifically, the multi-dimensional matching scoring engine includes:

[0060] (1) The multidimensional user profile acquisition module is configured to acquire the nursing needs time sequence vector, social feature vector, geographical location information and living environment preference vector of elderly users.

[0061] The nursing needs time-series vector is defined as follows: a day is divided into several continuous time slots according to a preset granularity, and each time slot is associated with a service item code, an item type label, and a required duration. The item type label includes at least: a "parallel service item" label indicating that it can be provided to multiple elderly people at the same time, and a "serial exclusive service item" label indicating that it can only be provided to one elderly person at a time.

[0062] The social feature vector includes: scores for five dimensions—openness, conscientiousness, extraversion, agreeableness, and neuroticism—obtained through an online simplified Big Five personality test, as well as a set of interest tags selected and weighted by the user.

[0063] The geographic location information includes: the latitude and longitude of the current residence, the desired area for shared accommodation, the maximum acceptable relocation distance, and a tag indicating the need for specific nearby facilities.

[0064] The residential environment preference vector includes ratings for preferences across multiple dimensions, such as room orientation, floor level, quietness, lighting conditions, and accessibility facilities.

[0065] (2) Multi-dimensional matching score calculation module

[0066] This module includes a nursing efficiency complementarity calculation unit, a social compatibility calculation unit, a geographical location suitability calculation unit, a living environment matching unit, and a comprehensive scoring and recommendation unit.

[0067] The nursing efficiency complementarity calculation unit extracts all time slots labeled "serial exclusive service items" from each user's time-series vector, calculates the time overlap between all user pairs within the combination on these time slots, and sums them up as the total demand conflict degree. The lower the demand conflict degree, the higher the nursing efficiency complementarity score. Preferably, the conflict degree C = Σ_iΣ_{j>i} Overlap(SerialSlots_i, SerialSlots_j), and the complementarity score S_efficiency = 1 / (1+ C). A nursing demand level difference coefficient is also introduced to encourage matching users with different disability levels to balance service load.

[0068] The social compatibility calculation unit is used to calculate the personality compatibility index between each user pair within a combination based on a preset personality dimension compatibility matrix; calculate the interest compatibility index based on the overlap of interest tags and the interest weights set by the user; and generate a social compatibility score by weighted summation of the two.

[0069] The location fit calculation unit is used to calculate the geographical distance between each user's current residence, the degree of overlap in each user's preferences for the target shared accommodation area, and the accessibility to important nearby facilities based on each user's location information. The closer the geographical distance and the higher the degree of preference overlap, the higher the location fit score.

[0070] The residential environment matching calculation unit is used to calculate the similarity between the residential environment preference vectors of each user. The higher the similarity, the higher the residential environment matching score. Similarity calculation can use cosine similarity or Euclidean distance.

[0071] The comprehensive scoring and recommendation unit normalizes the four scores mentioned above, then sums them according to weighted coefficients to obtain a comprehensive matching score. It outputs a list of recommended roommate combinations in descending order of score and provides detailed score breakdowns for each dimension for user reference. The weighted coefficients are initially preset values ​​and can be updated by the matching weight optimization module.

[0072] (3) Matching weight optimization module

[0073] This module adaptively adjusts the weighting coefficients. It collects stability data of existing co-tenancy groups over a preset period, including at least whether a group has disbanded, member satisfaction scores, and the number of disputes. Using this stability data to construct a composite monitoring signal, and employing the four-dimensional scores of the co-tenancy group as input features, the module iteratively updates the weighting coefficients using gradient descent or Bayesian optimization algorithms to maximize the correlation between the overall matching score and co-tenancy stability. This module is not a simple application of general machine learning algorithms; its specially constructed composite monitoring signal provides a quantitative definition of the specific technical indicator of "co-tenancy group stability," solving the technical challenge of existing technologies' inability to quantitatively evaluate and adaptively improve the long-term effectiveness of matching schemes.

[0074] It should be noted that, based on the above four dimensions, those skilled in the art can add auxiliary dimensions such as economic budget matching degree, work-rest time compatibility degree, hygiene habit matching degree, and dietary preference compatibility degree according to actual needs, and these should all fall within the protection scope of this invention. In the initial stage when data accumulation is insufficient, the system can use preset fixed weights for matching, and the matching weight optimization module will be activated after sufficient data has been accumulated.

[0075] (II) Service Resource Collaborative Scheduling and Supervision Subsystem

[0076] This subsystem is used to achieve unified scheduling, collaborative scheduling, quantitative quality supervision of service processes, and security monitoring of shared rental environments for different types of service resources. It includes:

[0077] (1) Service Resource Management Module

[0078] This module is used to maintain the pool of available service resources. Service provider types include human caregivers, nursing robots, and intelligent automated devices with service execution capabilities. Each service provider carries a type identifier, capability parameters, and cost parameters. The service resource pool can be configured to contain only human caregivers, enabling the system to operate in a purely human service mode; or it can be configured to contain multiple types, enabling the system to operate in a collaborative service mode.

[0079] (2) Collaborative scheduling module

[0080] This module merges the time-series vectors of care needs from all users within a shared rental unit. Using a service item classifier, it assigns all service items to suitable service provider types based on the capability parameters of each provider. The scheduling engine optimizes by minimizing overall service cost and maximizing service coverage. Based on each service provider's available time window, capability constraints, and cost weights, it uses a genetic algorithm or mixed-integer programming algorithm to generate a conflict-free joint service schedule that meets all service needs, explicitly identifying the task executor type for each time period. It should be noted that the task classification and resource allocation logic of the collaborative scheduling module is as follows: Figure 3 As shown.

[0081] (3) Unified Service Quality Assessment Module (Please refer to) Figure 4 )

[0082] This module is used to establish a non-discriminatory quality supervision framework covering different types of service providers, and may include:

[0083] Perception data acquisition interface: For services performed by human caregivers, it connects to intelligent nursing equipment or vision sensors deployed at the service site to collect operating parameters; for services performed by nursing robots or automated equipment, it directly obtains process data such as motion parameters, force control parameters, and temperature parameters of the actuator through its control interface.

[0084] Service Compliance Assessment Unit: This unit has pre-set thresholds for uniform key quality indicators for all services, applicable to all service providers regardless of their type. For example, the water temperature range for assisted bathing services is 37.0℃ to 40.0℃, and the time spent within this range must account for no less than 90% of the total time. The peak torque for rehabilitation training assistance must not exceed a preset safety limit, and the number of completed movements must not be less than 90% of the prescription requirements. This unit compares the collected parameters with the thresholds in real time to generate a quantitative compliance score. When a parameter deviates from the threshold, an early warning command is immediately generated and sent to the corresponding service execution terminal.

[0085] Comprehensive Quality Scoring and Reporting Unit: This unit integrates objective compliance scores with real-time satisfaction scores collected from seniors via voice interaction terminals, weights these scores to generate a final service quality score, and periodically compiles and generates performance reports for each service provider.

[0086] (4) Safety and Risk Monitoring Module (Please refer to) Figure 4 )

[0087] This module is used to detect the safety of the shared living environment and any abnormal conditions of the elderly in real time, and is linked to scheduling. It can include:

[0088] Environmental safety sensing unit: Connects to environmental safety sensors such as smoke detectors, gas leak detectors, water immersion sensors, door magnetic sensors, and infrared intrusion detectors deployed in shared housing, and is used to detect environmental safety events in real time and generate environmental safety alarms.

[0089] Elderly Status Monitoring Unit: Connects to status sensing devices such as millimeter-wave radar fall detectors, wearable vital sign monitoring bracelets, bedside emergency call buttons, and bathroom pull-cord alarms deployed in shared housing. It is used to analyze in real time whether the elderly person has fallen, has an abnormal heart rate, or has been stationary for a long time, and to generate emergency rescue alarms.

[0090] Multi-level alarm and linkage response unit: Equipped with multi-level alarm strategies, level 1 alarms are pushed to the caregiver's smart wristband and voice broadcasting equipment in the shared housing; level 2 alarms are pushed to the preset emergency contact terminal for children; level 3 alarms automatically dial emergency services and send the house's location information. The core innovation of this unit lies in its deep linkage with the collaborative scheduling module: when an alarm is triggered, the collaborative scheduling module suspends the currently affected services, reassesses the availability of service resources, and generates an adjusted service schedule to prioritize the response to the alarm event. For example, when an elderly person experiences a fall alarm, the system automatically adjusts the elderly person's non-emergency services such as rehabilitation training or bathing assistance for the next day to rest and observation, and re-optimizes the service schedules of the remaining elderly people, ensuring that caregiver resources can still meet the basic needs of all elderly people in emergency situations.

[0091] Security Log and Risk Analysis Unit: Records all security events and generates security logs; based on historical security data, it uses anomaly detection algorithms (such as the Isolation Forest algorithm) to identify high-frequency risk periods and risk types, and generates risk warning prompts.

[0092] (III) Intelligent Cost Allocation Subsystem (Please refer to) Figure 5 )

[0093] This subsystem enables a comprehensive and fair allocation of costs across the entire project, including fixed expenses, variable costs, and the usage costs of various service resources.

[0094] 1. Smart metering network access module

[0095] This module is equipped with an interface for communication with smart metering hardware such as smart meters / water meters, smart sockets, and small flow meters, used to obtain water, electricity, and gas consumption data for each member or area, including:

[0096] Variable usage allocation calculation unit: Direct proportional allocation is used based on usage data; when fine-grained independent metering data is lacking, an estimation model based on behavioral characteristics is activated. Electricity cost estimates are calculated based on room area coefficient and air conditioning operating time weights; water cost estimates are calculated based on per capita basic usage, individual bathing frequency, and single water usage.

[0097] Service fee allocation unit: The service fee for human caregivers is allocated based on the weighted proportion of the duration of human care received by each user, multiplied by the nursing need level weighting coefficient (0.8 for mild, 1.0 for moderate, and 1.2 for severe); the usage fee for nursing robots or automated equipment is allocated based on the proportion of the actual time each user occupies the equipment.

[0098] 2. Comprehensive Bill Generation Module

[0099] This module combines rent, property management fees, variable usage allocation, and various service fee allocations to generate an electronic comprehensive bill that includes details of resource usage and a complete allocation calculation process, and pushes it to each user's terminal and their pre-set emergency contact terminal for their children.

[0100] (iv) Dynamic Management Module for Co-tenants

[0101] This module is used to automatically trigger the following when a member leaves a shared unit: rescheduling the service schedule of the remaining members; settling the overall fees of the departing members; and, if new members need to be added, starting a multi-dimensional matching and scoring engine to recommend new members.

[0102] (v) Data storage unit

[0103] This unit stores user profiles, demand vectors, scheduling schemes, quality assessment data, security logs, measurement data, matching weight parameters, allocation rules, and settlement records, providing unified data support for each subsystem.

[0104] Preferably, the system provided by this invention may further include business support modules such as an electronic protocol management module, a deposit and payment integration module, an insurance service matching module, a government subsidy calculation module, and a user credit evaluation module, as well as enhancement technology modules such as a digital twin visualization module, a voice interaction module, an edge computing gateway, a device health management module, and a multi-unit resource sharing module. The inclusion of these modules does not deviate from the core architecture of this invention.

[0105] As can be seen from the above description, implementing the embodiments of the present invention has the following beneficial effects:

[0106] 1. Multidimensional matching significantly improves the stability and economic feasibility of co-living: For the first time, four dimensions—complementary care efficiency, social compatibility, geographical location suitability, and living environment matching—are incorporated into a unified matching model. An innovative weighted adaptive optimization mechanism based on real-world co-living feedback is introduced to ensure optimal service costs, interpersonal harmony, relocation convenience, and living comfort for co-living groups from the outset. Calculations show that the average care cost per person can be reduced by 15%-25% after matching, and the breakup rate due to conflict can be reduced by more than 70%.

[0107] 2. Collaborative scheduling and seamless expansion of heterogeneous service resources: The configurable design of the service resource pool allows the system to currently operate as a purely manual co-living elderly care management platform. In the future, it can be seamlessly upgraded to a human-machine collaborative or automated service mode simply by registering new devices, thus protecting users' long-term investment.

[0108] 3. Unified service quality supervision without discrimination: Regardless of whether the service is performed by humans, robots or automated equipment, parameters are collected through a unified sensor interface and compared with the same standard thresholds to ensure the objectivity of quality evaluation and solve the problem of difficulty in obtaining evidence in service disputes.

[0109] 4. Comprehensive safety and risk protection: The system integrates environmental safety monitoring and elderly status monitoring, and innovatively links safety alarms with service scheduling to ensure that the elderly can receive assistance immediately in case of falls, sudden illnesses or dangerous environments, without affecting the basic services of other elderly people, effectively reducing the safety risks of co-living elderly care.

[0110] 5. Refined cost allocation across the entire project: For the first time, equipment usage costs are included in the allocation based on the duration of use. Combined with intelligent metering and behavioral estimation models to handle variable consumption and nursing level weighted processing of labor costs, this constitutes the most comprehensive cost-sharing scheme for co-living elderly care.

[0111] 6. System-level combinatorial innovation and continuous evolution capability: It organically integrates multi-dimensional matching, heterogeneous resource collaborative scheduling, unified quality perception and security monitoring, and comprehensive cost allocation to construct a closed-loop data flow throughout the entire process: "matching results drive scheduling → scheduling determines monitoring strategies → monitoring and security data are jointly input into the cost allocation model → feedback from the co-working group optimizes matching weights." Its synergistic effects and adaptive evolution capabilities are not publicly disclosed in existing technologies, demonstrating outstanding substantive characteristics and significant progress.

[0112] Based on the same inventive concept, embodiments of the present invention also provide a method for co-living elderly care services based on multi-dimensional matching and collaborative scheduling of service resources, including:

[0113] Collect users' nursing needs time-series vector, social feature vector, geographical location information, and residential environment preference vector; calculate scores for nursing efficiency complementarity, social compatibility, geographical location adaptability, and residential environment matching; generate a comprehensive matching score by normalizing and weighting the sum; and recommend co-living combinations; and iteratively optimize the weight coefficients of the weighted sum based on the stability data of co-living feedback.

[0114] After establishing a co-tenancy relationship, the service requirements of the co-tenancy unit are allocated to suitable service providers based on at least one type of service provider and their capability parameters available in the service resource pool, thereby generating a conflict-free joint service schedule.

[0115] During service execution, operation parameters are collected through smart devices and compared with preset unified quality standard thresholds to generate service quality assessment results and trigger necessary warnings. At the same time, safety events and emergency states are monitored in real time through environmental safety sensors and elderly status sensing devices, alarms are generated, and when an alarm is triggered, the currently affected service is suspended, the availability of service resources is reassessed, and an adjusted service schedule is generated.

[0116] During the settlement period, the service fee allocation is calculated based on the duration of service received by each user from different service providers and the corresponding cost parameters, and a comprehensive cost bill is generated by combining the allocation of variable usage.

[0117] Optionally, such as Figure 6 As shown, the above method flow may include:

[0118] S1, multi-dimensional matching and recommendation of roommate groups.

[0119] In practice, it is mainly used to collect data on nursing needs, social interactions, geographical location, and living environment, calculate scores for the four dimensions based on this data, and then optimize the system.

[0120] S2 generates conflict-free joint service schedules, identifies available service providers, classifies service items, and performs cost-optimized scheduling.

[0121] S3, Service Quality Assessment and Security Monitoring: Unified threshold comparison generates compliant environment security and status detection; when an alarm is triggered, service is suspended and a reassessment and new scheduling are performed.

[0122] S4 generates a comprehensive cost bill: it weights and distributes labor costs, distributes equipment costs by duration, and allocates variable usage, and pushes the bill to the user's and their children's terminals.

[0123] Based on the results of step S1, the feedback from S4 is used as the co-rental feedback data. The two are then normalized, weighted, summed, and the weights are adaptively optimized to finally recommend a co-rental combination.

[0124] To better understand the solution provided by this invention, specific examples are given below:

[0125] Example: Four-dimensional matching, security monitoring, and end-to-end operation

[0126] Elderly residents A (76 years old, mildly disabled, residing in Community A, requiring meal assistance at 8:00 AM and bath assistance at 2:00 PM), B (81 years old, moderately disabled, residing in Community B, requiring meal assistance at 7:00 AM and rehabilitation at 10:00 AM), C (79 years old, mildly disabled, residing in Community C, requiring meal assistance at 8:00 AM and a walk at 4:00 PM), and D (78 years old, moderately disabled, residing in Community D, requiring meal assistance at 9:00 AM and bath assistance at 11:00 AM) all expressed interest in sharing a rental apartment and completed their information entry on the user terminal with the assistance of their respective children.

[0127] The system collects time-series vectors of each individual's nursing needs, social feature vectors, geographic location information, and residential environment preference vectors. The multi-dimensional matching score calculation module calculates the following scores: nursing efficiency complementarity score 0.91 (conflict degree ∑Overlap≈0.1, complementarity = 1 / 1.1≈0.91), social compatibility score 0.83, geographic location suitability score 0.94, and residential environment suitability score 0.88. The system then uses the weighting coefficients from the previous optimization (0.32, 0.26, 0.22, 0.20), resulting in a weighted overall score of 0.89, recommending co-renting. The co-renting relationship is established after confirmation by all four individuals.

[0128] The service resource collaborative scheduling and monitoring subsystem is currently configured with human caregivers. The collaborative scheduling module generates a conflict-free daytime schedule for one caregiver. At 3:00 AM one day, an elderly person, B, falls in their bedroom. The millimeter-wave radar fall detector detects the event, and after confirmation by the elderly person's status monitoring unit, the multi-level alarm and linkage response unit simultaneously sends a strong vibration alarm and location information to the night shift caregiver's smart bracelet and pushes a notification to their children's devices. The collaborative scheduling module immediately suspends B's rehabilitation training for the next day, reassesses the availability of caregiver resources, and generates an adjusted schedule: B will rest and be observed for the entire next day, and basic services such as meal assistance and walks for the other three elderly people will be rescheduled to ensure no impact. The caregiver quickly handles the situation, and after confirming that no medical treatment is needed, the safety log is analyzed using the isolated forest algorithm, and no abnormal patterns are found.

[0129] The service quality assessment module collects data in real time through sensors in smart bathtubs and rehabilitation equipment, comparing it against thresholds to score compliance and satisfaction. Month-end fees are allocated based on area, water and electricity usage, and weighted caregiver hours, with a transparent comprehensive bill pushed to the elderly and their children's devices.

[0130] Three months later, the system's matching weight optimization module was activated: Satisfaction data from the shared apartment unit, which had been running stably for three months (average satisfaction score of 4.5 for four people, no signs of dissolution), was collected, along with feedback data from other shared apartments during the same period. The weight coefficients were updated using gradient descent. The new round of weight adjustments was (0.30, 0.25, 0.25, 0.20), with a slight increase in the weight of geographical location to reflect the higher preference of elderly residents in the area for convenient neighborhood access. The updated weights will be applied to subsequent matching calculations.

[0131] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A shared elderly care service system based on multi-dimensional matching and collaborative scheduling of service resources, characterized in that, include: The co-living matching subsystem is equipped with a multi-dimensional matching scoring engine, which calculates the complementarity score of nursing efficiency, the social compatibility score, the geographical location suitability score, and the living environment matching score based on the user's nursing needs time-series vector, social feature vector, geographical location information, and living environment preference vector, respectively. It then generates a comprehensive matching score through normalized weighted summation to recommend co-living combinations. The service resource collaborative scheduling and supervision subsystem is equipped with a service resource management module, a collaborative scheduling module, a unified service quality assessment module, and a security and risk monitoring module. It is used to realize unified scheduling, collaborative scheduling, quantitative quality supervision of service processes, and security monitoring of the shared rental environment for different types of service resources. The intelligent cost allocation subsystem is equipped with a service fee allocation unit, which is used to calculate the service fees that each user should share based on the duration of service received by each user from different service providers and the corresponding cost parameters, and generate a comprehensive allocation bill. The data storage unit is used to store user profiles, demand vectors, scheduling schemes, quality assessment data, security logs, metering data, matching weight parameters, allocation rules and settlement records, providing unified data support for each subsystem.

2. The system as described in claim 1, characterized in that, The nursing efficiency complementarity calculation is specifically as follows: extract the time slots of serial exclusive service items in the time sequence vector of each user's nursing needs, calculate the time overlap between all user pairs in the combination on this type of time slot as the demand conflict degree. The lower the conflict degree, the higher the nursing efficiency complementarity score, and introduce the nursing demand level difference coefficient. The social compatibility calculation specifically involves: calculating the personality compatibility index between each user pair based on a preset personality dimension compatibility matrix, calculating the interest compatibility index based on the overlap of interest tags, and then weighting and summing the two to generate a social compatibility score. The location fit calculation specifically involves: calculating the geographical distance between each user's current residence and the degree of overlap in their preferences for the target shared rental area. The closer the distance and the higher the degree of overlap in preferences, the higher the location fit score. The calculation of the living environment matching degree specifically involves calculating the similarity between the living environment preference vectors of each user. The higher the similarity, the higher the living environment matching degree score. The living environment preference vector includes scores for room orientation, floor, quietness, lighting conditions, and accessibility facilities.

3. The system as described in claim 1, characterized in that, The multi-dimensional matching scoring engine also includes a matching weight optimization module, used for: The stability data of the formed co-tenancy combination within a preset period is collected. The stability data is used as a supervision signal, and the scores of each dimension of the co-tenancy combination are used as input features. The gradient descent method or Bayesian optimization algorithm is used to iteratively update the weight coefficients of the normalized weighted summation.

4. The system as described in claim 1, characterized in that, The service resource management module is used to maintain a service resource pool containing at least one type of service provider; the service provider type includes at least one of human caregivers, nursing robots, and intelligent automated equipment; the service resource pool can be configured to contain only human caregivers and operate in a purely human service mode, or to contain multiple types and operate in a collaborative service mode; The collaborative scheduling module is used to allocate the service needs of the shared rental unit to suitable service providers based on the service providers' capability parameters, and generate a conflict-free joint service schedule. The unified service quality assessment module is used to collect key operational parameters during the service execution process through smart devices, compare them with preset unified quality standard thresholds applicable to different service provider types, and generate service process compliance and quality assessment results. The safety and risk monitoring module is used to detect safety events and emergency states in real time through environmental safety sensors and elderly status sensing devices, generate alarms, and when an alarm is triggered, the collaborative scheduling module suspends the currently affected services, reassesses the availability of service resources, and generates an adjusted service schedule to prioritize the response to alarm events.

5. The system as described in claim 4, characterized in that, The unified service quality assessment module includes: The perception data acquisition interface is used to connect services performed by human caregivers with intelligent nursing equipment or vision sensors deployed at the service site. For services performed by nursing robots or automated equipment, execution parameters are directly obtained through their control interface. The service compliance assessment unit is used to provide a unified key quality indicator threshold for all services, and the same threshold applies to different types of service providers.

6. The system as described in claim 4, characterized in that, The security and risk monitoring module includes: An environmental safety sensing unit is connected to environmental safety sensors deployed in shared housing to detect at least one environmental safety event, including smoke, gas leaks, water immersion, and illegal intrusion, and generate environmental safety alarms in real time. An elderly condition monitoring unit is connected to at least one of a fall detector, a vital signs monitoring device, and an emergency call device deployed in a shared housing unit, and is used to detect at least one emergency state of elderly people falling, abnormal heart rate, or prolonged stillness and generate an emergency rescue alarm. The multi-level alarm and linkage response unit is configured to push alarm information to the caregiver terminal, the preset emergency contact terminal for children, and / or the emergency service platform, and to link with the collaborative scheduling module.

7. The system as described in claim 1, characterized in that, In the service fee sharing unit, the service fee for human caregivers is shared according to the weighted proportion of the duration of the human service received by each user multiplied by a preset nursing need level weighting coefficient; the service fee for nursing robots or intelligent automated equipment is shared according to the proportion of the actual time each user occupies the service of the equipment.

8. The system as described in claim 1, characterized in that, It also includes a co-tenancy member dynamic management module, which is used to automatically trigger the rescheduling of services for the remaining members, the settlement of fees for the departing members, and the invocation of the multi-dimensional matching and scoring engine to match and recommend new members when a member is detected to have left.

9. A method for co-living elderly care services based on multi-dimensional matching and collaborative scheduling of service resources, characterized in that, include: Collect users' nursing needs time-series vector, social feature vector, geographical location information, and residential environment preference vector; calculate scores for nursing efficiency complementarity, social compatibility, geographical location adaptability, and residential environment matching; generate a comprehensive matching score by normalizing and weighting the sum; and recommend co-living combinations; and iteratively optimize the weight coefficients of the weighted sum based on the stability data of co-living feedback. After establishing a co-tenancy relationship, the service requirements of the co-tenancy unit are allocated to suitable service providers based on at least one type of service provider and their capability parameters available in the service resource pool, thereby generating a conflict-free joint service schedule. During service execution, operation parameters are collected through smart devices, compared with preset unified quality standard thresholds, and service quality assessment results are generated and necessary warnings are triggered. Simultaneously, environmental safety sensors and elderly status sensing devices are used to monitor safety events and emergency states in real time, generate alarms, and when an alarm is triggered, the currently affected services are suspended, the availability of service resources is reassessed, and an adjusted service schedule is generated. During the settlement period, the service fee allocation is calculated based on the duration of service received by each user from different service providers and the corresponding cost parameters, and a comprehensive cost bill is generated by combining the allocation of variable usage.

10. The method as described in claim 9, characterized in that, The nursing efficiency complementarity calculation is specifically as follows: extract the time slots of serial exclusive service items in the time sequence vector of each user's nursing needs, calculate the time overlap between all user pairs in the combination on this type of time slot as the demand conflict degree. The lower the conflict degree, the higher the nursing efficiency complementarity score, and introduce the nursing demand level difference coefficient. The social compatibility calculation specifically involves: calculating the personality compatibility index between each user pair based on a preset personality dimension compatibility matrix, calculating the interest compatibility index based on the overlap of interest tags, and then weighting and summing the two to generate a social compatibility score. The location fit calculation specifically involves: calculating the geographical distance between each user's current residence and the degree of overlap in their preferences for the target shared rental area. The closer the distance and the higher the degree of overlap in preferences, the higher the location fit score. The calculation of the living environment matching degree specifically involves calculating the similarity between the living environment preference vectors of each user. The higher the similarity, the higher the living environment matching degree score. The living environment preference vector includes scores for room orientation, floor, quietness, lighting conditions, and accessibility facilities.

Images