Pension service scheduling system based on artificial intelligence

Abstract

The invention relates to the technical field of health service scheduling, in particular to an old-age care service scheduling system based on artificial intelligence, which comprises an old-age care task interaction influence degree calculation module, an old-age care task path optimization module, an old-age care parallel task scheduling module, an old-age care task intersection flow regulation and control module and an old-age care task time window adjustment module. According to the method, by calculating the spatial proximity, the time interval overlapping ratio and the resource occupation condition of the tasks, the task pairs with high relevance are recognized, efficient collaboration between the tasks is achieved, path adjustment reduces task execution conflicts, the resource utilization rate is increased, the task priority optimization makes the low-interaction-influence task sequence more reasonable, and the task efficiency is improved. Task backlog and scheduling chaos are avoided, flow control balance tasks are distributed in different time periods, the risk of intersection congestion is reduced, execution smoothness is improved, a time window dynamic adjustment mechanism optimizes task arrangement according to real-time data, scheduling adaptability is ensured, and resource allocation accuracy and execution efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of health service scheduling, and in particular to an artificial intelligence-based elderly care service scheduling system. Background Art

[0002] The technical field of elderly care service dispatch system includes the use of information technology and artificial intelligence technology to achieve the reasonable allocation and management of elderly care service resources. The core content of this technical field is to integrate elderly care resources through digital means, optimize service supply and demand matching, and improve service efficiency and accuracy. The overall technical field covers data collection and analysis, service resource management, task scheduling optimization, and service visualization. Through the use of sensor networks, big data processing and artificial intelligence algorithms, this technical field realizes the comprehensive perception and dynamic monitoring of the health status, service needs and resource distribution of the elderly, thereby supporting scientific and reasonable scheduling decisions.

[0003] Among them, health service scheduling refers to the effective configuration of health service resources in an intelligent way in response to the health management needs in elderly care services. The subject of this patent covers specific technical matters such as real-time collection and analysis of health data, classification and prioritization of service requests, construction of resource scheduling models, and optimization of service allocation paths. Taking health data collection as an example, this patent records the health indicators of the elderly through terminals such as wearable devices, and classifies health conditions with the help of data analysis tools. For service requests, the patent generates scheduling priorities based on classification rules, and optimizes the configuration of service resources through scheduling algorithms, and finally completes the dispatch of service personnel or equipment through path planning, thereby realizing intelligent scheduling of health services.

[0004] The scheduling of elderly care services lacks a comprehensive analysis of the dependencies between tasks during the resource matching process, resulting in uneven resource allocation and affecting task connection. Path planning does not fully consider the traffic conditions at task intersections, which is prone to congestion in high-traffic areas, affecting overall execution efficiency. The task priority division method is relatively fixed and cannot be adaptively adjusted to the dynamic changes of tasks, resulting in reduced resource utilization. The scheduling plan fails to dynamically optimize according to the task time window, and resource shortages are prone to occur during task-intensive periods, affecting service continuity and response speed. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose an artificial intelligence-based elderly care service scheduling system.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solution: The elderly care service scheduling system based on artificial intelligence includes:

[0007] The elderly care task interaction impact calculation module obtains the execution location, execution time, and resource occupancy of elderly care, medical follow-up, and health check-up tasks, calculates the spatial proximity, time interval overlap ratio, and resource overlap occupancy time ratio between tasks, and selects task pairs that are lower than the spatial proximity threshold, higher than the time dependency threshold, and higher than the resource dependency threshold, calculates the spatial interaction impact score, time interaction impact score, and resource interaction impact score, and obtains the elderly care task interaction impact matrix through weighted calculation;

[0008] The elderly care task path optimization module extracts high-impact task groups based on the elderly care task interaction influence matrix, screens path overlapping areas, marks intersections, adjusts the execution time of high-interaction-influence tasks, and optimizes the elderly care task path set;

[0009] The elderly care parallel task scheduling module selects low interactive impact task groups based on the elderly care task path set, calculates the intersection travel interval, adjusts the execution order of low priority health check tasks, and optimizes the elderly care parallel scheduling task path set;

[0010] The elderly care task intersection flow control module calculates the intersection flow, selects tasks during high flow periods, adjusts the execution time of low-urgent nursing tasks, and optimizes the elderly care task intersection path set based on the elderly care parallel scheduling task path set;

[0011] The elderly care task time window adjustment module monitors intersection traffic deviations, screens over-threshold tasks, adjusts low-priority health check task time windows, and optimizes the elderly care task dynamic time window set based on the elderly care task intersection path set.

[0012] As a further solution of the present invention, the elderly care task interaction impact matrix includes a spatial interaction impact score, a temporal interaction impact score, and a resource interaction impact score; the elderly care task path set includes a high-impact task group, a path overlapping area, an intersection, and an adjusted high-interaction-impact task execution time; the elderly care parallel scheduling task path set includes a low-interaction-impact task group, an intersection passage interval, and an adjusted low-priority health check task execution order; the elderly care task intersection path set includes intersection traffic, high-traffic period tasks, and an adjusted low-urgency nursing task execution time; the elderly care task dynamic time window set includes intersection traffic deviation, over-threshold tasks, and an adjusted low-priority health check task time window.

[0013] As a further solution of the present invention, the elderly care task interaction impact calculation module includes:

[0014] The task execution information acquisition submodule obtains the execution location, execution time, and resource usage of elderly care, medical follow-up, and health check tasks, extracts the geographic coordinate information of the tasks, calculates the spatial distance between tasks, extracts the start and end time of the tasks based on the execution time and calculates the overlap of the time intervals, obtains the resource usage of the tasks and calculates the resource usage time, and obtains the task execution parameter set;

[0015] The task interaction screening submodule calculates the spatial proximity between tasks based on the task execution parameter set, determines whether it is lower than the spatial proximity threshold, calculates the time interval overlap ratio between tasks, screens the task pairs higher than the time dependency threshold, calculates the resource overlap occupancy time ratio, screens the task pairs higher than the resource dependency threshold, and obtains the task interaction screening result;

[0016] The interaction impact score calculation submodule calculates the spatial interaction impact score based on the task interaction screening result, performs normalization based on the spatial proximity, calculates the time interaction impact score, performs normalization based on the time interval overlap ratio, calculates the resource interaction impact score, performs normalization based on the resource overlap occupancy time ratio, and obtains an interaction impact score set;

[0017] The interaction influence matrix calculation submodule sets the weight coefficient of the differentiated score based on the interaction influence score set, calculates the weighted influence of the task interaction, and constructs the elderly care task interaction influence matrix.

[0018] As a further solution of the present invention, the weighted influence of the calculation task interaction is calculated using the formula:

[0019]

[0020] Calculate the weighted influence value and construct the elderly care task interaction influence matrix;

[0021] Among them, I w Represents the weighted influence of task interaction, W i Represents the weight coefficient of the i-th task interaction, I i represents the impact score of the ith task interaction, n represents the total number of task interactions, Represents the square root of the sum of the squares of all weight coefficients.

[0022] As a further solution of the present invention, the elderly care task path optimization module includes:

[0023] The high-impact task extraction submodule extracts task pairs whose interaction influence exceeds a set threshold based on the elderly care task interaction influence matrix, selects task groups that share execution resources or execution areas, calculates the interaction influence values ​​between tasks within the task group, and sorts them according to the influence values ​​to obtain high-impact task groups;

[0024] The path intersection screening submodule extracts the task execution path based on the high-impact task group, calculates the coordinates of the path intersection points, screens the path overlapping sections, counts the number of path intersection points, classifies them according to the length of the path overlapping sections, marks the intersection points with greater impact, and obtains the path intersection point distribution data;

[0025] The task time adjustment submodule extracts the task execution time corresponding to the path intersection based on the path intersection distribution data, calculates the time adjustment space between tasks, screens the adjustable tasks according to the time interval overlap ratio, adjusts the execution time of the tasks with high interaction impact, and obtains the adjusted task execution time set;

[0026] The path optimization calculation submodule recalculates the task execution path based on the adjusted task execution time set, detects whether the path intersection points are reduced, calculates the change in the total path length, screens the path optimization scheme, and screens the optimal scheme according to the optimized path intersection situation to obtain the optimized elderly care task path set.

[0027] As a further solution of the present invention, the elderly care parallel task scheduling module includes:

[0028] The task screening submodule extracts the interaction impact parameters and task priorities of the task paths based on the set of elderly care task paths, calculates the interaction impact values ​​between the task paths and compares them with the task impact threshold, screens the task paths whose interaction impact values ​​are lower than the task impact threshold, and generates a low interaction impact task group;

[0029] The traffic interval calculation submodule calls the low interaction impact task group, obtains the intersection location data and the task path travel time, calculates the minimum traffic interval between intersection tasks according to the intersection task travel time sequence, and compares the calculated value with the intersection traffic safety threshold to generate the intersection traffic interval;

[0030] The health check task adjustment submodule extracts the execution order data of low-priority health check tasks in the task path based on the low interaction impact task group and the intersection travel interval, adjusts the execution order of the health check tasks according to the execution time of the health check tasks in the task path, the intersection travel interval and the task impact parameters, and generates an optimized elderly care parallel scheduling task path set.

[0031] As a further solution of the present invention, the elderly care task intersection flow control module includes:

[0032] The intersection flow calculation submodule extracts the intersection task traffic data and task flow parameters based on the elderly care parallel scheduling task path set, calculates the task flow value of each intersection in the differentiated time period, and compares it with the intersection flow benchmark value to obtain the intersection flow;

[0033] The high-flow task screening submodule calls the intersection traffic, extracts the task execution time data and the task path number, screens the task paths whose task traffic exceeds the task traffic threshold, and sorts them according to the task execution time to obtain the high-flow period tasks;

[0034] The nursing task time adjustment submodule extracts the execution time data and task type of low-urgency nursing tasks based on the high-flow period tasks, and adjusts the execution time of low-urgency nursing tasks to a time period below the task flow threshold according to the task priority and execution time window, to obtain an optimized elderly care task intersection path set.

[0035] As a further solution of the present invention, the intersection task traffic data and task flow parameters are calculated using the formula:

[0036]

[0037] Calculate the intersection task flow value, combine the calculation result, and compare it with the intersection flow benchmark value to obtain the intersection flow;

[0038] Among them, Q t Represents the intersection task flow value in time period t, V i,t represents the passage rate of task i in time period t, L i represents the path length of task i, N represents the total number of tasks passing through the intersection in time period t, T j,t represents the travel time of task j in time period t, represents the average travel time of all tasks in time period t, and M represents the total number of tasks in time period t.

[0039] As a further solution of the present invention, the pension task time window adjustment module includes:

[0040] The intersection flow deviation monitoring submodule extracts the intersection flow data and time series data based on the elderly care task intersection path set, calculates the mean of the intersection flow in multiple periods, and calculates the deviation between the task flow in each period and the mean, compares the deviation value with the intersection flow deviation threshold, and obtains the intersection flow deviation;

[0041] The over-threshold task screening submodule calls the intersection traffic deviation, extracts task execution time data and task type information, screens task paths whose task traffic deviation values ​​exceed the intersection traffic deviation threshold, and sorts them according to task execution time to obtain over-threshold tasks;

[0042] The health check task time adjustment submodule extracts the execution time data and task impact parameters of the low-priority health check task based on the above-threshold task, and adjusts the execution time of the low-priority health check task to a time period with a lower intersection traffic deviation value according to the task priority and the execution time adjustment range, to obtain a dynamic time window set for optimizing the elderly care tasks.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are:

[0044] In the present invention, by calculating the spatial proximity of tasks, the time interval overlap ratio and resource occupancy, highly correlated task pairs are identified to achieve efficient collaboration between tasks. Path adjustment reduces task execution conflicts and improves resource utilization. Task priority optimization makes the order of low-interaction-impact tasks more reasonable, avoiding task backlogs and scheduling confusion. Flow control balances the distribution of tasks in different time periods, reduces the risk of intersection congestion, and improves execution smoothness. The time window dynamic adjustment mechanism optimizes task scheduling according to real-time data, ensures the adaptability of scheduling, and improves the accuracy and execution efficiency of resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a system flow chart of the present invention;

[0046] Figure 2 This is a system framework diagram of the present invention;

[0047] Figure 3 This is a flow chart of the elderly care task interaction impact calculation module of the present invention;

[0048] Figure 4 This is a flow chart of the elderly care task path optimization module of the present invention;

[0049] Figure 5 This is a flow chart of the elderly care parallel task scheduling module of the present invention;

[0050] Figure 6 This is a flow chart of the traffic control module at the intersection of the elderly care task of the present invention;

[0051] Figure 7 This is a flow chart of the elderly care task time window adjustment module of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0054] Embodiment 1

[0055] See also Figure 1 , the AI-based elderly care service scheduling system includes:

[0056] The elderly care task interaction impact calculation module obtains the execution location, execution time, and resource occupancy of elderly care, medical follow-up, and health check-up tasks, calculates the spatial proximity, time interval overlap ratio, and resource overlap occupancy time ratio between tasks, and selects task pairs that are lower than the spatial proximity threshold, higher than the time dependency threshold, and higher than the resource dependency threshold, calculates the spatial interaction impact score, time interaction impact score, and resource interaction impact score, and obtains the elderly care task interaction impact matrix through weighted calculation;

[0057] The elderly care task path optimization module extracts high-impact task groups based on the elderly care task interaction impact matrix, screens path overlap areas, marks intersections, adjusts the execution time of high-interaction-impact tasks, and optimizes the elderly care task path set;

[0058] The elderly care parallel task scheduling module is based on the elderly care task path set, screens low interactive impact task groups, calculates the intersection travel interval, adjusts the execution order of low priority health check tasks, and optimizes the elderly care parallel scheduling task path set;

[0059] The elderly care task intersection flow control module calculates the intersection flow, screens tasks during high-flow periods, adjusts the execution time of low-urgency nursing tasks, and optimizes the elderly care task intersection path set based on the elderly care parallel scheduling task path set;

[0060] The elderly care task time window adjustment module is based on the elderly care task intersection path set, monitors intersection traffic deviation, filters out over-threshold tasks, adjusts low-priority health check task time windows, and optimizes the elderly care task dynamic time window set.

[0061] The elderly care task interaction impact matrix includes the spatial interaction impact score, the temporal interaction impact score, and the resource interaction impact score. The elderly care task path set includes high-impact task groups, path overlapping areas, intersections, and adjusted high-interaction-impact task execution times. The elderly care parallel scheduling task path set includes low-interaction-impact task groups, intersection travel intervals, and adjusted low-priority health check task execution order. The elderly care task intersection path set includes intersection traffic, high-traffic period tasks, and adjusted low-urgency nursing task execution times. The elderly care task dynamic time window set includes intersection traffic deviations, over-threshold tasks, and adjusted low-priority health check task time windows.

[0062] See also Figure 2 and Figure 3 ,The elderly care task interaction impact calculation module includes:

[0063] The task execution information acquisition submodule obtains the execution location, execution time, and resource usage of elderly care, medical follow-up, and health check tasks, extracts the geographic coordinate information of the tasks, calculates the spatial distance between tasks, extracts the start and end time of the tasks based on the execution time and calculates the overlap of the time intervals, obtains the resource usage of the tasks and calculates the resource usage time, and obtains the task execution parameter set;

[0064] The task execution information acquisition submodule collects the execution location, execution time and resource usage from the elderly care, medical follow-up and health check tasks. First, the geographic coordinate information of the task is obtained, and the latitude and longitude coordinates of the task are recorded using a global positioning system (GPS) device or a geographic information system (GIS). For example, the coordinates of the elderly care task are (30.2672°N, 97.7431°W), the coordinates of the medical follow-up task are (34.0522°N, 118.2437°W), and the coordinates of the health check task are (40.7128°N, 74.0060°W). Then, the spatial distance between the tasks is calculated, and the Haversine formula is used to calculate the shortest distance between two points on the earth's surface. The formula is:

[0065] Among them, r is the radius of the earth, which is about 6371 kilometers, Δφ and Δλ are the differences between the latitude and longitude of the two points, respectively, φ1 and φ2 are the latitudes of the two points. Next, the start and end time of the task is extracted based on the execution time, and the start and end time of each task is recorded. For example, the execution time of the elderly care task is from 10:00 to 12:00 on January 31, 2025, the medical follow-up task is from 11:00 to 13:00 on January 31, 2025, and the health check task is from 14:00 to 16:00 on January 31, 2025. The overlap of the time intervals is calculated to determine the overlapping part of the task time periods, for example, the time of the elderly care task and the medical follow-up task The overlap is from 11:00 to 12:00, a total of 1 hour. Obtain the resource usage of the task and record the type and quantity of resources used for each task. For example, the elderly care task uses 1 nurse, the medical follow-up task uses 1 doctor and 1 vehicle, and the health check task uses 2 nurses and 1 electrocardiograph. Calculate the resource occupancy time and multiply it by the number of resources used and the task duration. For example, the vehicle resource occupancy time of the medical follow-up task is 1 vehicle × 2 hours = 2 hours. Finally, summarize the above information to obtain a set of task execution parameters, including the geographical coordinates, spatial distance, time interval overlap, resource usage, and resource occupancy time of the task.

[0066] The task interaction screening submodule calculates the spatial proximity between tasks based on the task execution parameter set, determines whether it is lower than the spatial proximity threshold, calculates the time interval overlap ratio between tasks, screens out task pairs higher than the time dependency threshold, calculates the resource overlap occupancy time ratio, screens out task pairs higher than the resource dependency threshold, and obtains the task interaction screening result;

[0067] The task interaction screening submodule calculates the spatial proximity between tasks based on the task execution parameter set. First, the spatial distance between tasks is calculated. The Haversine formula is used to calculate the shortest distance between two points on the earth's surface. The formula is:

[0068] Among them, r is the radius of the earth, which is about 6371 kilometers. Δφ and Δλ are the differences between the latitude and longitude of the two points, respectively. φ1 and φ2 are the latitudes of the two points. The calculated distance is used to evaluate the spatial proximity between tasks and determine whether it is lower than the preset spatial proximity threshold. For example, the spatial proximity threshold is set to 5 kilometers. If the distance between the two tasks is less than 5 kilometers, they are considered to have spatial proximity. Next, the time interval overlap ratio between the tasks is calculated. First, determine the overlapping part of the time intervals of the two tasks. For example, the execution time of task A is from 10:00 to 12:00, and the execution time of task B is from 11:00 to 13:00. The overlapping time is from 11:00 to 12:00, a total of 1 hour. The time interval overlap ratio is calculated as the overlapping time divided by the duration of the shorter task. For example, task A lasts for 2 hours and the overlapping time is 1 hour, then the overlapping ratio is 1 / 2=0.5. The task pairs above the time dependence threshold are screened. For example, the time dependence threshold is set to 0.3. If the time interval overlap ratio of the two tasks is greater than 0.3, then They are considered to have time dependence. Finally, the resource overlapping occupancy time ratio is calculated. First, determine whether the resource types used by the two tasks are the same. If they are the same, calculate the resource overlapping occupancy time. For example, task A and task B both use the same vehicle. The usage time of task A is from 10:00 to 12:00, and the usage time of task B is from 11:00 to 13:00. Then the resource overlapping occupancy time is from 11:00 to 12:00, for a total of 1 hour. The resource overlapping occupancy time ratio is calculated as the overlapping time divided by the total resource occupancy time. For example, the total occupancy time of the vehicle is 2 hours of task A plus 2 hours of task B, minus 1 hour of overlap, for a total of 3 hours. Then the resource overlapping occupancy time ratio is 1 / 3≈0.33. Task pairs above the resource dependence threshold are screened. For example, the resource dependence threshold is set to 0.2. If the resource overlapping occupancy time ratio of two tasks is greater than 0.2, they are considered to have resource dependence. After the above screening process, the task interaction screening results are obtained, including task pairs with spatial proximity, time dependence and resource dependence.

[0069] The interaction impact score calculation submodule calculates the spatial interaction impact score based on the task interaction screening results, performs normalization based on the spatial proximity, calculates the time interaction impact score, performs normalization based on the time interval overlap ratio, calculates the resource interaction impact score, performs normalization based on the resource overlap occupancy time ratio, and obtains the interaction impact score set;

[0070] The interaction impact score calculation submodule calculates the spatial interaction impact score based on the task interaction screening results. First, the task pairs with spatial proximity in the screening results are obtained, and the spatial distance between them is calculated. The Haversine formula is used to calculate the shortest distance between two points on the earth's surface. The formula is:

[0071]

[0072] Among them, r is the radius of the earth, which is about 6371 kilometers, Δφ and Δλ are the differences in latitude and longitude of the two points, φ1 and φ2

[0073] is the latitude of the two points. The calculated spatial distance is normalized to be between 0 and 1. For example, using the minimum -

[0074] Maximum Normalization Method:

[0075] Among them, d max and d min are the maximum and minimum spatial distances between tasks, S norm

[0076] Represents the normalized spatial interaction impact score. For example, if the maximum distance is set to 10 kilometers and the minimum distance is 1 kilometer,

[0077]

[0078] Next, calculate the time interaction impact score, obtain the task pairs with time dependency in the screening results, and calculate their time interval overlap

[0079] Among them, t overlap is the time overlap between tasks, t min

[0080] The duration of the shorter task, for example, Task A takes 2 hours to execute and Task B takes 3 hours to execute, and their overlap time is

[0081]

[0082] Then, the resource interaction impact score is calculated, the task pairs with resource dependencies in the screening results are obtained, and the resource overlap occupancy is calculated.

[0083] Among them, t resource_overlap is the overlapping occupation time of resources, t total_resource

[0084] is the total resource occupancy time. For example, if Task A and Task B share a vehicle, Task A takes 2 hours and Task B takes 3 hours.

[0085] Finally, we get the interaction impact score set, including spatial interaction impact score, temporal interaction impact score and resource interaction impact score.

[0086] The interaction influence matrix calculation submodule sets the weight coefficient of the differentiated score based on the interaction influence score set, calculates the weighted influence of task interaction, and constructs the elderly care task interaction influence matrix.

[0087] Calculate the weighted influence of task interactions using the formula:

[0088]

[0089] Calculate the weighted influence value and construct the elderly care task interaction influence matrix;

[0090] Among them, I w Represents the weighted influence of task interaction, W i Represents the weight coefficient of the i-th task interaction, I i represents the impact score of the ith task interaction, n represents the total number of task interactions, Represents the square root of the sum of the squares of all weight coefficients.

[0091] formula:

[0092]

[0093] Detailed explanation of the formula and the process of formula calculation and derivation:

[0094] In this formula, I w represents the weighted influence of task interaction. In order to calculate I w , the following parameters need to be determined:

[0095] Impact score of task interaction I i :Evaluate the impact of each task interaction and get the corresponding score. The score can be obtained through expert evaluation, historical data analysis or questionnaire survey. For example, for a task interaction, the expert evaluates its impact score as 7.

[0096] Weight coefficient W i :According to the importance or credibility of each task interaction, a corresponding weight coefficient is assigned. The weight coefficient can be determined based on statistical analysis, expert scoring or hierarchical analysis method. For example, the importance of a task interaction is evaluated as 0.8.

[0097] Assume that there are three task interactions, and their impact scores and weight coefficients are as follows:

[0098] Task interaction 1: I1=7, W1=0.8;

[0099] Task interaction 2: I2=5, W2=0.6;

[0100] Task interaction 3: I3=9, W3=0.9;

[0101] Substituting these values ​​into the formula, the calculation process is as follows:

[0102] Calculate the absolute value of the weighted influence value of each task interaction and sum them:

[0103]

[0104] Calculate the square root of the sum of the squares of the weight coefficients:

[0105]

[0106] Calculate weighted influence I w :

[0107]

[0108] The result shows that after comprehensively considering the impact scores of each task interaction and the corresponding weight coefficients, the weighted impact calculated is 12.42. This value reflects the overall weighted impact of task interactions under the current evaluation system.

[0109] See also Figure 2 and Figure 4 , the elderly care task path optimization module includes:

[0110] The high-impact task extraction submodule extracts task pairs whose interaction influence exceeds the set threshold based on the elderly care task interaction influence matrix, selects task groups that share execution resources or execution areas, calculates the interaction influence values ​​between tasks within the task group, and sorts them according to the influence values ​​to obtain high-impact task groups;

[0111] The elderly care task interaction influence matrix is ​​used to calculate the interaction influence value between tasks. First, extract the task pairs whose interaction influence exceeds the set threshold, set the threshold T, and filter all the interaction influence values ​​I ij Satisfy I ij ≥T task pairs (i, j), secondly, screen the task groups that share execution resources or execution areas, first check whether the execution positions of the task pairs overlap, and use the spatial coordinates to calculate the overlap ratio of the task execution areas, using the formula:

[0112] Among them, A overlap is the overlapping area of ​​the task execution area, A min is the area of ​​the smaller task area. For example, the execution areas of Task A and Task B are 100 square meters and 150 square meters respectively, and their overlapping area is 50 square meters. The overlapping ratio is calculated as:

[0113] Then, check the shared resources of the tasks. For example, if both Task A and Task B use the same medical inspection vehicle, they are considered to share resources. After screening the task groups that share resources or execution areas, calculate the interaction impact value between tasks within the task group. The calculation formula is: I group =∑ i,j∈G I ij

[0114] Among them, G is the task set of the task group. For example, the task group includes tasks A, B and C, and their interaction influence is as follows: I AB =0.7, I AC =0.6, I BC =0.5, then the interaction effect value within the task group is calculated as follows:

[0115] I group =0.7+0.6+0.5=1.8;

[0116] According to the influence value sorting, the interaction influence values ​​of the task groups are arranged in descending order to obtain the high-impact task groups. The results are used for subsequent path optimization and task adjustment.

[0117] The path intersection screening submodule extracts the task execution path based on the high-impact task group, calculates the coordinates of the path intersection points, screens the path overlapping sections, counts the number of path intersection points, classifies them according to the length of the path overlapping sections, marks the intersection points with greater impact, and obtains the path intersection distribution data;

[0118] The path cross-screening of the high-impact task group uses the task execution path data for analysis. First, the task execution path is extracted, the task trajectory is recorded, and the sequence of geographic coordinate points of the task execution trajectory is obtained using the GPS device. For example, the path coordinate points of task A are (x1, y1), (x2, y2)…(x n ,y n ), the path coordinate points of task B are (x′1, y′1), (x′2, y′2)…(x′ m ,y′ m ), then calculate the coordinates of the path intersection point, use the intersection calculation method of the two path trajectories to check whether the adjacent line segments intersect, and assume that the two line segments are:

[0119] L1=(x i ,y i )→(x i+1 ,y i+1 );

[0120] L2=(x′ j ,y′ j )→(x′ j+1 ,y′ j+1 );

[0121] If two line segments meet the intersection condition:

[0122] (x i -x i+1 )(y′ j -y′ j+1 )-(y i -y i+1 )(x′ j -x′ j+1 )≠0;

[0123] Then calculate the intersection coordinates (x c ,y c ), count the number of path intersections, set the path overlap length threshold, and calculate the path overlap length L overlap And classify them. For example, if the number of path intersections between tasks A and B is 3, and the lengths of the path overlapping sections are 2.5km, 1.8km and 3.2km respectively, they are classified and the intersections with greater influence (such as those with a length of more than 2km) are marked, and finally the path intersection distribution data is obtained.

[0124] The task time adjustment submodule extracts the task execution time corresponding to the path intersection based on the path intersection distribution data, calculates the time adjustment space between tasks, screens the adjustable tasks according to the time interval overlap ratio, adjusts the execution time of the tasks with high interaction impact, and obtains the adjusted task execution time set;

[0125] Based on the path intersection distribution data, the task time adjustment submodule adjusts the execution time of the task. First, the task execution time corresponding to the path intersection is extracted. For example, the execution time of task A is 10:00-12:00, and the execution time of task B is 11:00-13:00. The intersection appears at 11:15. To calculate the time adjustment space between tasks, first calculate the time interval overlap ratio:

[0126]

[0127] Among them, t overlap is the task overlap time, t min is the execution time of the shorter task. For example, if the time overlap between tasks A and B is 45 minutes and task A lasts for 2 hours, the calculation result is:

[0128]

[0129] Filter the adjustable tasks according to the time interval overlap ratio and set the adjustment threshold value. For example, if the threshold value is 0.3, the task pair meets the adjustment conditions and adjusts the execution time of the tasks with high interaction impact. The execution time of Task A can be changed to 9:30-11:30 to reduce the impact of path intersection and obtain the adjusted task execution time set.

[0130] The path optimization calculation submodule recalculates the task execution path based on the adjusted task execution time set, detects whether the path intersections are reduced, calculates the change in the total path length, screens the path optimization scheme, and selects the optimal scheme according to the optimized path intersection situation to obtain the optimized elderly care task path set.

[0131] The path optimization calculation submodule recalculates the task execution path based on the adjusted task execution time set. First, update the adjusted task execution trajectory, re-extract the task path data, and calculate the path intersections. The same method is used to calculate the intersection coordinates and detect whether the path intersections are reduced. For example, the number of path intersections before adjustment is 5, and it is reduced to 2 after adjustment. The change in the total path length is calculated. Assuming that the total path length before adjustment is 20km and after adjustment is 18.5km, the path change is:

[0132] ΔL=L before -L after =20-18.5=1.5km;

[0133] Screen the path optimization scheme, compare the reduction of path intersections and the change of path length under different adjustment strategies, and determine the optimal scheme. For example, among the three adjustment strategies, Scheme 1 reduces 2 intersections, Scheme 2 reduces 3 intersections, and Scheme 3 reduces 4 intersections, but the path increases by 2km. Finally, Scheme 2 is selected to obtain the optimized elderly care task path set.

[0134] See also Figure 2 and Figure 5 , the pension parallel task scheduling module includes:

[0135] The task screening submodule extracts the interaction impact parameters and task priorities of the task paths based on the set of elderly care task paths, calculates the interaction impact values ​​between the task paths and compares them with the task impact threshold, screens the task paths whose interaction impact values ​​are lower than the task impact threshold, and generates a low interaction impact task group;

[0136] The task screening submodule analyzes the set of elderly care task paths. First, the interaction impact parameters of the task paths are extracted, including task execution time, number of task path intersections, shared resources, and spatial proximity. The interaction impact value between task paths is calculated using the following formula:

[0137] I path =αS overlap +βT overlap +γR overlap ;

[0138] Where: S overlapis the spatial overlap of the task paths, which is calculated as the ratio of the length of the path intersection area to the total length of the shorter task path; T overlap is the task time overlap, that is, the ratio of the time overlap to the shorter task execution time; R overlap is the task resource overlap, which is calculated as the ratio of the shared resource usage time to the total task execution time; α, β, and γ are weighting coefficients, which are set according to the task importance, resource usage, and actual execution requirements.

[0139] Then, the calculated interaction impact value is compared with the task impact threshold, and the task paths whose interaction impact value is lower than the task impact threshold are screened. For example, if the task impact threshold is set to 0.4, if the interaction impact value I between tasks A and B is path =0.35, then the task pair meets the screening criteria, and finally a low interaction impact task group is generated, including a set of all task paths that meet the screening criteria.

[0140] The travel interval calculation submodule calls the low interaction impact task group to obtain the intersection location data and the task path travel time, calculates the minimum travel interval between intersection tasks according to the intersection task travel time sequence, and compares the calculated value with the intersection travel safety threshold to generate the intersection travel interval;

[0141] The travel interval calculation submodule calls the low interaction impact task group for calculation. First, the intersection location data is obtained, and the intersection coordinates of each task path are recorded using a GPS device. For example, the intersection coordinates of tasks A and B are (x1, y1) and (x2, y2), respectively. At the same time, the travel time of the task path is obtained, and the time series of each task arriving at the intersection is recorded. For example, task A arrives at the intersections (x1, y1) and (x2, y2) at 10:05 and 10:20 respectively, and task B arrives at the same intersection at 10:08 and 10:22.

[0142] Then, the minimum travel interval between tasks at the intersection is calculated using the following formula:

[0143] T gap =min(T B,i -T A,i );

[0144] Among them, T B,i and T A,i They represent the arrival time of task B and task A at the intersection i. For example, for the intersection (x1, y1), we have:

[0145] T gap =10:08-10:05=3 minutes;

[0146] Then, the calculated value is compared with the intersection safety threshold. For example, if the safety threshold is set to 2 minutes, if the calculated minimum interval is 3 minutes, the safe passage requirement is met. If it is less than 2 minutes, the task execution time needs to be adjusted. Finally, the intersection interval data is generated, including the task interval information of each intersection.

[0147] The health check task adjustment submodule extracts the execution order data of low-priority health check tasks in the task path based on the low-interaction-impact task group and the intersection travel interval. It adjusts the execution order of the health check tasks according to the execution time of the health check tasks in the task path, the intersection travel interval and the task impact parameters, and generates an optimized elderly care parallel scheduling task path set.

[0148] The health check task adjustment submodule is analyzed based on the low interaction impact task group and the intersection travel interval. First, the execution order data of the low priority health check tasks in the task path is extracted to obtain the path position of the health check tasks in the elderly care tasks. For example, task A performs elderly care, medical follow-up and health check in sequence, and task B performs medical follow-up, health check and elderly care in sequence. Then, according to the execution time of the health check tasks in the task path, combined with the intersection travel interval and task impact parameters, the execution order of the health check tasks is adjusted. The adjustment method is as follows:

[0149] Check the time window of the health check task and calculate the time overlap between the task and other tasks. For example, the health check time of task A is 11:00-11:30, and the health check time of task B is 10:50-11:20, so the overlap time is 20 minutes.

[0150] Adjust the order of tasks based on the traffic interval. If the execution time of the health check task causes the intersection traffic interval to be lower than the safety threshold, adjust the execution order of Task A so that its health check time is adjusted from 11:00-11:30 to 11:20-11:50.

[0151] Update the task path, recalculate the path intersection position, and record the adjusted execution order to ensure that the traffic safety threshold meets the requirements, and finally generate an optimized elderly care parallel scheduling task path set.

[0152] See also Figure 2 and Figure 6 ,The traffic control module for the elderly care task intersection includes:

[0153] The intersection traffic calculation submodule extracts the intersection task traffic data and task traffic parameters based on the elderly care parallel scheduling task path set, calculates the task traffic value of each intersection in the differentiated time period, and compares it with the intersection traffic benchmark value to obtain the intersection traffic;

[0154] The high-traffic task screening submodule calls the intersection traffic, extracts the task execution time data and task path number, screens the task paths whose task traffic exceeds the task traffic threshold, and sorts them according to the task execution time to obtain the high-traffic period tasks;

[0155] The high-flow task screening submodule first calls the intersection traffic data, extracts the task execution time data and task path number of each task path, records the time all tasks pass through the intersection, and uses GPS trajectory data and task scheduling system to record the travel path and passing time of each task. For example, the passing time of task A at the intersection (x?, y?) is 10:10, the passing time of task B at the same point is 10:12, and task C passes at 10:14. After that, the flow of each task path is counted, that is, the number of tasks passing through the intersection per unit time, and the task flow F is calculated using the following formula:

[0156] Where: N is the number of tasks passing through the intersection per unit time, and T is the time window size, for example, set to 10 minutes.

[0157] If N = 6 tasks pass through the intersection in a certain time period, and the time window is 10 minutes, the task flow is calculated as: Tasks / minute;

[0158] Then, it is compared with the set task flow threshold. For example, if the task flow threshold is set to 0.5 tasks / minute, the task flow in this time period exceeds the threshold and needs to be adjusted. The task paths whose task flow exceeds the threshold are screened out and sorted according to the task execution time. For example, tasks A, B, and C pass at 10:10, 10:12, and 10:14 respectively, then a high-flow period task set is formed after sorting.

[0159] The nursing task time adjustment submodule extracts the execution time data and task type of low-urgency nursing tasks based on the high-flow period tasks. According to the task priority and execution time window, the execution time of low-urgency nursing tasks is adjusted to the time period below the task flow threshold to obtain the optimized elderly care task intersection path set.

[0160] The nursing task time adjustment submodule analyzes the task set during high-flow periods. First, the execution time data and task types of low-urgency nursing tasks are extracted. Nursing tasks are divided into high urgency (such as emergency patient care), medium urgency (such as regular follow-up), and low urgency (such as routine examinations) according to their urgency. Low-urgency tasks are selected from them. For example, if task A is an emergency nursing task, task B is a regular follow-up, and task C is a routine examination, then task C is classified as a low-urgency nursing task. After that, the execution time of low-urgency tasks is adjusted according to the task priority and execution time window, and they are arranged to the time period below the task flow threshold. For example:

[0161] Check the current task flow, count the number of tasks per unit time, and calculate the task flow F;

[0162] Identify the time window with lower traffic, for example, the task traffic is below the threshold between 10:30 and 10:40;

[0163] Adjust the time of low-urgency tasks. For example, adjust the execution time of Task C from 10:14 to 10:35, so that the task traffic is lower than the threshold.

[0164] Finally, the nursing task time adjustment is completed and the optimized elderly care task intersection path set is obtained.

[0165] The intersection task traffic data and task flow parameters are calculated using the formula:

[0166]

[0167] Calculate the intersection task flow value, combine the calculation result, and compare it with the intersection flow benchmark value to obtain the intersection flow;

[0168] Among them, Q t Represents the intersection task flow value in time period t, V i,t represents the passage rate of task i in time period t, L i represents the path length of task i, N represents the total number of tasks passing through the intersection in time period t, T j,t represents the travel time of task j in time period t, represents the average travel time of all tasks in time period t, and M represents the total number of tasks in time period t.

[0169] formula:

[0170]

[0171] Detailed explanation of the formula and the process of formula calculation and derivation:

[0172] In the above formula, Q trepresents the intersection task flow value in time period t. In order to calculate Q t , you need to get the following parameters:

[0173] Mission Passing Rate V i,t : represents the speed of task i in time period t (unit: m / s). This parameter can be obtained by real-time monitoring of the speed of vehicles passing through the intersection through traffic flow detection equipment (such as induction coils, video detection, etc.). For example, the average speed of a vehicle in time period t is 15 m / s.

[0174] Path length L i : represents the path length of task i (unit: meter). This parameter can be obtained through map measurement or vehicle odometer. For example, the path length of a task is 500 meters.

[0175] Total number of tasks N: indicates the total number of tasks that pass through the intersection within time period t. This parameter can be obtained by counting the number of vehicles passing through the intersection through traffic flow detection equipment. For example, in time period t, 100 vehicles pass through the intersection.

[0176] Task transit time T j,t : represents the travel time of task j in time period t (in seconds). This parameter can be calculated by the time difference between a vehicle entering and leaving the intersection. For example, the travel time of a vehicle is 40 seconds.

[0177] Average travel time Indicates the average travel time of all tasks in time period t (in seconds). This parameter is obtained by averaging the travel time of all tasks. For example, the average travel time of all vehicles in time period t is 35 seconds.

[0178] Total number of tasks M: represents the total number of tasks in time period t. Usually, M is equal to N. For example, in time period t, 100 vehicles pass through the intersection, so M = 100.

[0179] Substitute the above parameters into the formula for calculation:

[0180] Calculate the numerator part:

[0181]

[0182] Calculate the denominator:

[0183] Calculate the absolute value of the difference between each task's transit time and the average transit time:

[0184]

[0185] Sum over all tasks:

[0186]

[0187] Calculate the square root:

[0188]

[0189] Calculate Q t :

[0190]

[0191] The result shows that in time period t, the mission flow value of the intersection is about 33,545.5 square meters per second to the power of 1.5. This value reflects the traffic flow of the intersection during this time period and can be used to evaluate the traffic efficiency of the intersection.

[0192] See also Figure 2 and Figure 7 , the pension task time window adjustment module includes:

[0193] The intersection traffic deviation monitoring submodule extracts the intersection task traffic data and time series data based on the elderly care task intersection path set, calculates the mean of the intersection task traffic in multiple periods, and calculates the deviation between the task traffic in each period and the mean, compares the deviation value with the intersection traffic deviation threshold, and obtains the intersection traffic deviation;

[0194] The intersection flow deviation monitoring submodule is calculated based on the intersection path set of the elderly care task. First, the intersection task flow data and time series data are extracted, and the task scheduling system is used to record the travel time of each task, and the task flow in each time period is calculated. For example, in the 9:00-9:10 time period, tasks A, B and C pass through the intersection, and the task flow in this time period is recorded as 3. Then, the average of the intersection multi-time task flow is calculated, and the average is calculated using the data of the past N time periods:

[0195]

[0196] Among them, F i represents the task flow in the i-th time period. Assuming that the task flow in the past five time periods is 3, 4, 2, 5, and 3 respectively, the mean is calculated as follows:

[0197]

[0198] Then, calculate the deviation of task flow from the mean value in each period:

[0199]

[0200] For example, if the task flow rate F6 in the current period is 6, then calculate its deviation:

[0201] D6=|6-3.4|=2.6;

[0202] Finally, the deviation value is compared with the intersection traffic deviation threshold. For example, the deviation threshold is set to 2. If the calculated deviation value is greater than the threshold, it is determined that there is traffic anomaly during this period, and finally the intersection traffic deviation data is obtained.

[0203] The over-threshold task screening submodule calls the intersection traffic deviation, extracts the task execution time data and task type information, screens the task paths whose task traffic deviation values ​​exceed the intersection traffic deviation threshold, and sorts them according to the task execution time to obtain over-threshold tasks;

[0204] The over-threshold task screening submodule calls the intersection traffic deviation data, extracts the task execution time data and task type information, and screens the task paths whose task traffic deviation values ​​exceed the intersection traffic deviation threshold. First, all traffic deviation values ​​D are screened out. i The time period exceeds the set threshold T. For example, the threshold T is set to 2. If the flow deviation value D6=2.6 in a certain time period exceeds 2, the tasks in this time period need to be further analyzed. Then, the task execution time data of this time period is extracted. For example, tasks A, B and C are executed in the time period of 10:00-10:10. Then the execution time of these tasks is recorded, and the task type information is extracted. For example, task A is a health check, task B is elderly care, and task C is a medical follow-up. Finally, the tasks are sorted according to the execution time. For example, task A starts at 10:02, task B starts at 10:05, and task C starts at 10:08. After sorting, the set of tasks exceeding the threshold is obtained.

[0205] The health check task time adjustment submodule extracts the execution time data and task impact parameters of low-priority health check tasks based on the over-threshold tasks. According to the task priority and execution time adjustment range, the execution time of the low-priority health check tasks is adjusted to the time period with lower intersection traffic deviation value, and the dynamic time window set of optimized elderly care tasks is obtained.

[0206] The health check task time adjustment submodule is calculated based on the over-threshold task. First, the execution time data and task impact parameters of the low-priority health check task are extracted, and the priority of the health check task in the task type is screened. Usually, the emergency check task has a higher priority, and the regular physical examination task has a lower priority. For example, Task A is an emergency check and Task B is a regular physical examination, then Task B is identified as a low-priority task. Next, the adjustable time window is calculated to determine the time period with lower task traffic. For example, in the time period of 10:30-10:40, the task traffic is low, then this time period is suitable as an adjustment window, and then the low-priority health check task is adjusted to this time period to reduce the traffic deviation value, and finally obtain the optimized dynamic time window set of elderly care tasks.

[0207] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. The artificial intelligence-based elderly care service scheduling system is characterized by: The system comprises: The elderly care task interaction impact calculation module obtains the execution location, execution time, and resource occupancy of elderly care, medical follow-up, and health check-up tasks, calculates the spatial proximity, time interval overlap ratio, and resource overlap occupancy time ratio between tasks, and selects task pairs that are lower than the spatial proximity threshold, higher than the time dependency threshold, and higher than the resource dependency threshold, calculates the spatial interaction impact score, time interaction impact score, and resource interaction impact score, and obtains the elderly care task interaction impact matrix through weighted calculation; The elderly care task path optimization module extracts high-impact task groups based on the elderly care task interaction influence matrix, screens path overlapping areas, marks intersections, adjusts the execution time of high-interaction-influence tasks, and optimizes the elderly care task path set; The elderly care parallel task scheduling module selects low interactive impact task groups based on the elderly care task path set, calculates the intersection travel interval, adjusts the execution order of low priority health check tasks, and optimizes the elderly care parallel scheduling task path set; The elderly care task intersection flow control module calculates the intersection flow, selects tasks during high flow periods, adjusts the execution time of low-urgent nursing tasks, and optimizes the elderly care task intersection path set based on the elderly care parallel scheduling task path set; The elderly care task time window adjustment module monitors intersection traffic deviations, screens over-threshold tasks, adjusts low-priority health check task time windows, and optimizes the elderly care task dynamic time window set based on the elderly care task intersection path set.

2. The artificial intelligence-based elderly care service scheduling system according to claim 1 is characterized in that: The elderly care task interaction impact matrix includes a spatial interaction impact score, a temporal interaction impact score, and a resource interaction impact score. The elderly care task path set includes a high-impact task group, a path overlapping area, an intersection, and an adjusted high-interaction-impact task execution time. The elderly care parallel scheduling task path set includes a low-interaction-impact task group, an intersection travel interval, and an adjusted low-priority health check task execution order. The elderly care task intersection path set includes intersection traffic, high-traffic period tasks, and an adjusted low-urgency nursing task execution time. The elderly care task dynamic time window set includes intersection traffic deviation, over-threshold tasks, and an adjusted low-priority health check task time window.

3. The artificial intelligence-based elderly care service scheduling system according to claim 2 is characterized in that: The elderly care task interaction impact calculation module includes: The task execution information acquisition submodule obtains the execution location, execution time, and resource usage of elderly care, medical follow-up, and health check tasks, extracts the geographic coordinate information of the tasks, calculates the spatial distance between tasks, extracts the start and end time of the tasks based on the execution time and calculates the overlap of the time intervals, obtains the resource usage of the tasks and calculates the resource usage time, and obtains the task execution parameter set; The task interaction screening submodule calculates the spatial proximity between tasks based on the task execution parameter set, determines whether it is lower than the spatial proximity threshold, calculates the time interval overlap ratio between tasks, screens the task pairs higher than the time dependency threshold, calculates the resource overlap occupancy time ratio, screens the task pairs higher than the resource dependency threshold, and obtains the task interaction screening result; The interaction impact score calculation submodule calculates the spatial interaction impact score based on the task interaction screening result, performs normalization based on the spatial proximity, calculates the time interaction impact score, performs normalization based on the time interval overlap ratio, calculates the resource interaction impact score, performs normalization based on the resource overlap occupancy time ratio, and obtains an interaction impact score set; The interaction influence matrix calculation submodule sets the weight coefficient of the differentiated score based on the interaction influence score set, calculates the weighted influence of the task interaction, and constructs the elderly care task interaction influence matrix.

4. The artificial intelligence-based elderly care service scheduling system according to claim 3 is characterized in that: The weighted influence of the calculation task interaction is calculated using the formula: Calculate the weighted influence value and construct the elderly care task interaction influence matrix; Among them, I w Represents the weighted influence of task interaction, W i Represents the weight coefficient of the i-th task interaction, I i represents the impact score of the ith task interaction, n represents the total number of task interactions, Represents the square root of the sum of the squares of all weight coefficients.

5. The artificial intelligence-based elderly care service scheduling system according to claim 4 is characterized in that: The elderly care task path optimization module includes: The high-impact task extraction submodule extracts task pairs whose interaction influence exceeds a set threshold based on the elderly care task interaction influence matrix, selects task groups that share execution resources or execution areas, calculates the interaction influence values ​​between tasks within the task group, and sorts them according to the influence values ​​to obtain high-impact task groups; The path intersection screening submodule extracts the task execution path based on the high-impact task group, calculates the coordinates of the path intersection points, screens the path overlapping sections, counts the number of path intersection points, classifies them according to the length of the path overlapping sections, marks the intersection points with greater impact, and obtains the path intersection point distribution data; The task time adjustment submodule extracts the task execution time corresponding to the path intersection based on the path intersection distribution data, calculates the time adjustment space between tasks, screens the adjustable tasks according to the time interval overlap ratio, adjusts the execution time of the tasks with high interaction impact, and obtains the adjusted task execution time set; The path optimization calculation submodule recalculates the task execution path based on the adjusted task execution time set, detects whether the path intersection points are reduced, calculates the change in the total path length, screens the path optimization scheme, and screens the optimal scheme according to the optimized path intersection situation to obtain the optimized elderly care task path set.

6. The artificial intelligence-based elderly care service scheduling system according to claim 5 is characterized in that: The pension parallel task scheduling module includes: The task screening submodule extracts the interaction impact parameters and task priorities of the task paths based on the set of elderly care task paths, calculates the interaction impact values ​​between the task paths and compares them with the task impact threshold, screens the task paths whose interaction impact values ​​are lower than the task impact threshold, and generates a low interaction impact task group; The traffic interval calculation submodule calls the low interaction impact task group, obtains the intersection location data and the task path travel time, calculates the minimum traffic interval between intersection tasks according to the intersection task travel time sequence, and compares the calculated value with the intersection traffic safety threshold to generate the intersection traffic interval; The health check task adjustment submodule extracts the execution order data of low-priority health check tasks in the task path based on the low interaction impact task group and the intersection travel interval, adjusts the execution order of the health check tasks according to the execution time of the health check tasks in the task path, the intersection travel interval and the task impact parameters, and generates an optimized elderly care parallel scheduling task path set.

7. The artificial intelligence-based elderly care service scheduling system according to claim 6 is characterized in that: The elderly care task intersection flow control module includes: The intersection flow calculation submodule extracts the intersection task traffic data and task flow parameters based on the elderly care parallel scheduling task path set, calculates the task flow value of each intersection in the differentiated time period, and compares it with the intersection flow benchmark value to obtain the intersection flow; The high-flow task screening submodule calls the intersection traffic, extracts the task execution time data and the task path number, screens the task paths whose task traffic exceeds the task traffic threshold, and sorts them according to the task execution time to obtain the high-flow period tasks; The nursing task time adjustment submodule extracts the execution time data and task type of low-urgency nursing tasks based on the high-flow period tasks, and adjusts the execution time of low-urgency nursing tasks to a time period below the task flow threshold according to the task priority and execution time window, to obtain an optimized elderly care task intersection path set.

8. The artificial intelligence-based elderly care service scheduling system according to claim 7 is characterized in that: The intersection task traffic data and task flow parameters are calculated using the formula: Calculate the intersection task flow value, combine the calculation result, and compare it with the intersection flow benchmark value to obtain the intersection flow; Among them, Q t Represents the intersection task flow value in time period t, V i,t represents the passage rate of task i in time period t, L i represents the path length of task i, N represents the total number of tasks passing through the intersection in time period t, T j,t represents the travel time of task j in time period t, represents the average travel time of all tasks in time period t, and M represents the total number of tasks in time period t.

9. The artificial intelligence-based elderly care service scheduling system according to claim 8 is characterized in that: The pension task time window adjustment module includes: The intersection flow deviation monitoring submodule extracts the intersection flow data and time series data based on the elderly care task intersection path set, calculates the mean of the intersection flow in multiple periods, and calculates the deviation between the task flow in each period and the mean, compares the deviation value with the intersection flow deviation threshold, and obtains the intersection flow deviation; The over-threshold task screening submodule calls the intersection traffic deviation, extracts task execution time data and task type information, screens task paths whose task traffic deviation values ​​exceed the intersection traffic deviation threshold, and sorts them according to task execution time to obtain over-threshold tasks; The health check task time adjustment submodule extracts the execution time data and task impact parameters of the low-priority health check task based on the above-threshold task, and adjusts the execution time of the low-priority health check task to a time period with a lower intersection traffic deviation value according to the task priority and the execution time adjustment range, to obtain a dynamic time window set for optimizing the elderly care tasks.

Images