A Multi-Robot Cloud Scheduling and Coordination Method and System

Abstract

This invention relates to the field of robot scheduling technology, and more particularly to a multi-robot cloud scheduling and collaborative method and system. The method includes the following steps: acquiring building elevator group data, and dividing the building elevator service groups based on the building elevator group data to obtain building elevator service group data; acquiring robot target path data; segmenting the robot target path data and building elevator service group data into path resource segments to obtain split path data; acquiring current transfer floor data, and estimating transfer waiting time based on the current transfer floor data and split path data to obtain transfer waiting time estimation data; constructing a transfer map based on the transfer waiting time estimation data to obtain transfer map data; obtaining path conflict data based on the transfer map data; and obtaining robot scheduling data based on the path conflict data to perform multi-robot collaborative scheduling operations. This invention can improve task completion efficiency and scheduling intelligence in multi-robot cross-floor collaborative scenarios.

Description

Technical Field

[0001] This invention relates to the field of robot scheduling technology, and in particular to a multi-robot cloud scheduling and coordination method and system. Background Technology

[0002] With the increasing prevalence of service robots in enclosed environments such as buildings, hospitals, and hotels, the core issue has become how to efficiently schedule multiple robots and coordinate their use of shared resources (such as elevators). For example, in high-rise buildings, to improve elevator efficiency and implement zoned management, elevator systems are often divided into high-zone elevators and low-zone elevators, with the service floors covered by the two groups not overlapping. While this zoned management model improves operational efficiency, it also brings the unavoidable problem of "cross-elevator transfers" in robot task paths. For instance, a robot may need to take an elevator from the low zone to a transfer floor (such as the 10th floor) and then wait for the high-zone elevator to complete the transfer before reaching its target high-floor. Current technologies lack effective solutions for this cross-elevator scheduling requirement. Existing methods often assume that all elevators are uniformly scheduled resources, failing to fully consider the service range boundaries of elevators and the capacity limitations of transfer points. This leads to problems such as robot congestion at transfer floors, amplified transfer delays, and frequent scheduling conflicts. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes a multi-robot cloud scheduling and coordination method and system to solve at least one of the above-mentioned technical problems.

[0004] This application provides a multi-robot cloud scheduling and coordination method, applied to a cloud platform, the method comprising:

[0005] S1. Obtain building elevator group data, and divide the building elevator service groups according to the building elevator group data to obtain building elevator service group data.

[0006] S2. Obtain robot target path data; perform path resource segmentation on robot target path data and building elevator service group data to obtain split path data;

[0007] S3. Obtain the current transfer floor data, and estimate the transfer waiting time based on the current transfer floor data and the split path data to obtain the transfer waiting time estimate data;

[0008] S4. Construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; calculate the graph structure conflict based on the transfer map data to obtain path conflict data; generate a schedule based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations.

[0009] This invention divides the service areas of elevator groups within a building and cuts path resource segments based on the robot's target path, enabling accurate identification of service boundaries across floors for multiple elevator groups. By estimating waiting times at transfer floors, it dynamically assesses transfer risks and waiting times, avoiding scheduling delays caused by elevator group separation. By constructing a transfer graph and performing graph structure conflict calculations, it effectively identifies scheduling bottlenecks and resource competition relationships between path segments. Based on conflict data, it generates scheduling schemes, enabling collaborative scheduling of multiple robots in scenarios where high- and low-floor elevators are not interconnected, improving overall operational efficiency and cross-floor task completion rates.

[0010] Optionally, the building elevator service group division includes:

[0011] The elevator service relationship between floors is extracted from the building elevator group data to obtain the floor elevator matrix data.

[0012] Elevator operation behavior data is acquired, and service overlap frequency is calculated based on the elevator operation behavior data to obtain service overlap frequency data.

[0013] Based on the floor elevator matrix data and service overlap frequency data, a floor elevator connectivity graph is constructed to obtain building elevator service group data.

[0014] This invention extracts elevator service relationships between floors to form a floor elevator matrix, which clearly depicts the service coverage of each elevator to different floors, providing structured input for scheduling. By combining elevator operation behavior data with operation trajectories and scheduling records to calculate service overlap frequency, the degree of coordination and redundancy between different elevators in floor services can be quantified, avoiding scheduling rigidity caused by static partitioning. By fusing matrix data and service overlap frequency data to construct a floor elevator connectivity graph, not only is the visibility and flexibility of cross-elevator group scheduling in a building improved, but service discontinuity areas or high overlap risk areas can also be effectively identified, providing data support for constructing reasonable building elevator service groups.

[0015] Optionally, the construction of the floor elevator connectivity graph includes:

[0016] Based on the floor elevator matrix data and service overlap frequency data, structural coverage feature extraction and elevator collaborative scheduling feature extraction are performed to obtain structural coverage feature data and elevator collaborative scheduling feature data, respectively.

[0017] Based on structural coverage feature data and elevator collaborative scheduling feature data, a coupling graph is constructed from the floor elevator matrix data to obtain the coupling graph data.

[0018] Fault zone detection is performed on the coupled graph data to obtain authorized control graph data. Fault zone detection includes floor jump fault detection, floor no access record detection, and authorization restriction detection.

[0019] Apply time-period service mode switching constraints to the authorized control chart data to obtain constraint chart data;

[0020] Connectivity branch extraction is performed based on constraint graph data to obtain building elevator service group data.

[0021] This invention integrates floor elevator matrix data and service overlap frequency data to extract structural coverage features and elevator collaborative scheduling features. This not only identifies the service range of each elevator at the spatial level but also reflects their collaborative capabilities in actual operation, thus achieving comprehensive modeling of service capabilities and scheduling correlation. Constructing a coupling graph helps reveal the coupling relationships and resource accessibility between elevator services. By detecting fault zones in the coupling graph data, it can accurately identify service blind spots or inaccessible areas in the building caused by factors such as physical structure, elevator grouping, and access permissions, enhancing the system's adaptability to complex scenarios. Through time-based service mode switching constraints, elevator authorization and path feasibility can be dynamically adjusted at different times to meet the needs of changing strategies during morning and evening peak hours or at night. By extracting connected components from the constraint graph, it achieves fine division of multi-elevator service blocks, providing highly reliable structural support and scheduling foundation for cross-group path scheduling and robot cloud collaboration.

[0022] Optionally, the path resource segment cutting includes:

[0023] The robot's target path data and the building elevator service group data are divided into road segments to obtain path segmentation data;

[0024] Based on the path segmentation data, service attributes and transit points are extracted to obtain service attribute data and transit point data, respectively.

[0025] Based on service attribute data and transfer point data, the path partitioning data is processed to perform transfer constraint dependency processing to obtain path dependency data;

[0026] Path dependency graphs are constructed based on path dependency data to obtain split path data.

[0027] This invention divides the path into segments by combining robot target path data with building elevator service group data, enabling precise differentiation of segments across elevator groups and floors, thus improving scheduling granularity. Extracting service attributes and transfer point information from path segments helps identify the service elevator group and potential transfer nodes corresponding to each path segment, introducing spatial-resource dual semantics into path planning. By handling transfer constraint dependencies through service attributes and transfer points, the resource continuity logic and temporal sequence constraints between path segments can be explicitly modeled, making it suitable for path switching management in scenarios with separate high- and low-level elevators. Finally, constructing a path dependency graph not only improves the structural clarity of path planning but also provides an operable graphical model support for scheduling sequence reasoning and multi-robot collaborative allocation, significantly enhancing the path feasibility and robustness of the scheduling system under building structures.

[0028] Optionally, the transit wait estimation includes:

[0029] Based on the current transfer floor data and split path data, the arrival times of multiple robots are superimposed to calculate the arrival superimposed data.

[0030] Based on the arrival and overlay data, kernel density is calculated to obtain local clustering data;

[0031] Based on the current transfer floor data, the throughput and the opening gate cycle are calculated to obtain the throughput data and the opening gate cycle data respectively.

[0032] Based on the output volume data and the opening gate cycle data, the transfer waiting time of the local aggregation volume data is estimated to obtain the transfer waiting time estimate data.

[0033] This invention captures potential path concentration trends by superimposing arrival times of multiple robots onto current transfer floor data and split path data, accurately simulating the probability of multiple robots arriving simultaneously at transfer points. Kernel density calculation is used to smooth the arrival distribution, generating local clustering data that effectively measures congestion intensity at transfer nodes within a specific time window. Combining actual throughput of transfer floors with elevator door opening / closing cycle data dynamically reflects the processing capacity and traffic efficiency of transfer elevators at the real-world physical operation level. Weighted estimation of the local clustering data using the aforementioned throughput and cycle parameters quantifies the waiting delay of the current transfer node under different scheduling schemes, improving the scheduling system's congestion perception and response capabilities across elevator group scheduling paths. This provides data support for path conflict avoidance and task rearrangement, effectively enhancing the timeliness and stability of multi-robot collaborative scheduling.

[0034] Optionally, the transfer map construction includes:

[0035] Based on the current transfer floor data and the split path data, a path segment transfer map is constructed to obtain the path segment transfer map data;

[0036] Based on the estimated transit waiting data, the transit graph data of the path segment is weighted to obtain weighted graph data.

[0037] The weighted graph data is directed and scheduled to obtain the transfer graph data.

[0038] This invention constructs a path segment transfer graph based on transfer floor data and split path data, which abstracts the transfer relationship between robots in different service segments into a structured graph model, clearly representing the connection structure between path segments and transfer nodes. By introducing transfer waiting estimation data to assign edge weights to the graph structure, dynamic factors such as transfer time and congestion level can be integrated into the graph model, making the graph structure not only connected but also schedulable. By performing directed scheduling on the weighted graph data, transfer graph data oriented towards execution order is generated, ensuring that the transfer order between path segments meets the constraints of time dependency and resource reachability. This improves the system's ability to analyze the structure of multiple tasks in cross-elevator group paths, enabling the scheduling engine to have the data processing capability of "path-transfer-scheduling cost".

[0039] Optionally, the graph structure conflict calculation includes:

[0040] Path occupancy is mapped based on the transfer map data to obtain path occupancy data;

[0041] Conflict pairs are extracted based on path occupancy data to obtain conflict pair data.

[0042] Based on the conflict data, the path overlap tension is calculated to obtain the path conflict data.

[0043] This invention maps path occupancy data to explicitly represent the spatial and temporal occupancy relationships of multiple robots in key resources such as shared passages and elevator nodes, forming a high-precision resource scheduling view. Based on path occupancy data, conflict pair information can be extracted, enabling rapid location of potential scheduling conflict areas and critical path intersections, providing a foundation for conflict intervention. By introducing a path overlap tension calculation method, combining the spatial overlap and temporal concurrency of path segments, the degree of competition for shared resources by different paths is quantified, thereby generating path conflict data and enabling measurable and rankable conflict severity. This invention enhances the ability to identify and assess potential bottlenecks and congestion points between multiple robot paths, providing a precise conflict prediction and preprocessing mechanism for the scheduling system, effectively avoiding task delays and system deadlocks, and significantly improving the coordination and success rate of multi-robot scheduling.

[0044] Optionally, the path overlap tension calculation includes:

[0045] Based on the conflict data, non-rigid geometric overlap and time occupancy intersection are calculated to obtain fuzzy boundary overlap data and time occupancy intersection data respectively.

[0046] Based on the fuzzy boundary overlap data and the time occupation intersection data, the shared resource node competition data is processed to obtain node competition data.

[0047] Based on node contention data, resource occupancy tension is propagated to obtain occupancy propagation matrix data;

[0048] Based on the occupancy propagation matrix data, high-priority task scheduling gravity control is performed to obtain path conflict data.

[0049] This invention calculates the non-rigid geometric overlap and temporal overlap of conflict data, enabling more accurate capture of potential overlaps between robot paths in fuzzy boundary regions within high-rise building environments with diverse indoor corridor styles. Combined with task execution time window overlap, it outputs more granular fuzzy conflict features. The invention also handles shared resource node contention, identifying resource contention trends among multiple robots at bottleneck nodes such as elevator doors and corridor corners. By introducing a resource occupancy tension propagation mechanism, local competition is extended to the overall path structure, generating a propagation matrix that reflects the global scheduling pressure diffusion effect caused by local competition between paths. Furthermore, a high-priority task scheduling gravity control strategy dynamically balances the attractiveness of high-priority tasks to resources with the overall system tension, enhancing the priority guarantee capability for task execution. The overall process enhances the modeling accuracy and response flexibility for complex scheduling conflicts, possessing strong predictive and feedforward control capabilities, effectively improving the coordination efficiency and service quality of multi-robot scheduling systems in high-concurrency, resource-constrained scenarios.

[0050] Optionally, the schedule generation includes:

[0051] Based on the path conflict data, the core conflict groups are identified to obtain the core conflict group data.

[0052] Based on the conflict core group data, path-level scheduling is sorted to obtain scheduling sorting data;

[0053] Robot scheduling data is generated based on the scheduling order to enable multi-robot collaborative scheduling operations.

[0054] This invention identifies core conflict groups through path conflict data, accurately capturing robot path sets that form high-density competitive relationships within specific time windows or spatial nodes, and identifying key conflict sources affecting overall scheduling bottlenecks. Based on this, path-level scheduling is sorted according to multiple factors such as conflict intensity, task priority, and resource occupation duration, ensuring that high-priority tasks avoid low-priority tasks while maintaining overall resource utilization efficiency. Specific robot scheduling data is generated based on the sorting results, enabling staggered path execution, proactive conflict avoidance, and dynamic resource reuse, thereby improving the stability and system throughput of multi-robot collaborative scheduling operations.

[0055] Optionally, this application also provides a multi-robot cloud scheduling and collaboration system for executing the multi-robot cloud scheduling and collaboration method described above, the multi-robot cloud scheduling and collaboration system comprising:

[0056] The elevator service group division module is used to acquire building elevator group data and divide the building elevator service groups based on the building elevator group data to obtain building elevator service group data.

[0057] The path segment structure modeling module is used to acquire robot target path data; it also performs path resource segmentation on the robot target path data and building elevator service group data to obtain split path data.

[0058] The transfer waiting prediction module is used to obtain the current transfer floor data and estimate the transfer waiting based on the current transfer floor data and the split path data, thus obtaining the transfer waiting estimate data.

[0059] The conflict-aware scheduling generation module is used to construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; to calculate the graph structure conflict based on the transfer map data to obtain path conflict data; and to generate scheduling data based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations.

[0060] The purpose of this invention is as follows: S1, by dividing building elevator group data into service groups, introducing floor elevator connectivity and service overlap frequency modeling, effectively identifying elevator service boundaries, and supporting the subsequent reasonable division of path structure; S2, by combining the robot target path with the elevator service range to cut resource segments, and modeling the service attributes and transfer dependencies of the path, forming structured split path data; S3, by analyzing the convergence behavior of transfer floors between different path segments, calculating local aggregation density, throughput capacity, and elevator periodic response, thereby accurately estimating transfer waiting time delay; S4, by dividing the robot task path into path segment units, and using these as nodes and transfer floors as connecting edges to construct an initial transfer graph; and assigning values ​​to the edge weights in the graph based on the transfer waiting estimate, forming a weighted directed graph reflecting scheduling overhead. In this graph structure, the system identifies path pairs with overlapping time windows and spatial path intersections as potential conflict pairs; calculates the overlap tension function between conflicting paths to assess the severity of their scheduling conflicts; and then extracts the core conflict groups and performs local sorting optimization based on path priority and resource tension, thereby achieving coordinated scheduling and control of multiple robot tasks at transfer points. Attached Figure Description

[0061] Other features, objects, and advantages of this application will become more apparent from the following detailed description of the non-limiting embodiments, taken with reference to the accompanying drawings:

[0062] Figure 1 A flowchart illustrating the steps of a multi-robot cloud scheduling and collaboration method according to an embodiment is shown.

[0063] Figure 2 A flowchart illustrating the steps of a building elevator service group division method according to an embodiment is shown;

[0064] Figure 3 A flowchart illustrating the steps of a path resource segment cutting method according to an embodiment is shown.

[0065] Figure 4 A flowchart illustrating the steps of a transit waiting estimation method according to one embodiment is shown.

[0066] Figure 5 A flowchart illustrating the steps of a transfer map construction method according to one embodiment is shown.

[0067] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0068] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0069] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. Functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0070] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0071] Please see Figures 1 to 5 This application provides a multi-robot cloud scheduling and coordination method, applied to a cloud platform, the method comprising:

[0072] S1. Obtain building elevator group data, and divide the building elevator service groups according to the building elevator group data to obtain building elevator service group data.

[0073] Specifically, basic information about all elevators in the building is obtained, including the floor range served by each elevator, the controller number to which the elevator belongs, and the group control logic parameters; a floor-elevator service relationship matrix is ​​constructed, where the horizontal axis represents the floor number and the vertical axis represents the elevator number; let this matrix be M, and its dimension be n. m, where n represents the total number of floors and m represents the number of elevators. If the j-th elevator serves the i-th floor, then... =1, otherwise =0; Based on historical elevator operation logs, count the number of times any two elevators simultaneously serve the same floor, and calculate the service overlap frequency accordingly to generate an overlap matrix R between elevators, where Indicates the first With the The elevator has a frequency index of service overlap in all service floors; the floors are regarded as nodes in the graph model, and the relationship between two floors with elevator service overlap is regarded as an edge, and an elevator-floor coupled graph model is constructed; "fault zone" is identified in the graph structure, that is, the area with the following situations is detected: (1) floor skip service, that is, some floors cannot be directly reached by any single elevator; (2) isolated floors without elevator coverage service; (3) some restricted access floors under access control; based on the constructed coupled graph, all connected subgraphs are extracted by graph traversal algorithms (such as depth-first traversal, disjoint set, etc.). Each connected subgraph corresponds to a set of physically reachable and schedulable elevator service areas, and finally multiple elevator service groups are formed. The elevators in each service group can serve all the floors to which they belong. The service groups are connected through floor transfers; each elevator service group is numbered and its key attribute information is output, including: the set of floor numbers covered, the set of elevator numbers to which it belongs, and possible transfer floors (i.e. floors that belong to two or more service groups at the same time).

[0074] S2. Obtain robot target path data; perform path resource segmentation on robot target path data and building elevator service group data to obtain split path data;

[0075] Specifically, the current task path data for each robot to be scheduled is obtained, including the coordinates of the start and end points of the task. This path is mapped onto the building's spatial structure, clarifying the floors and physical locations of the start and end points, thus forming a complete navigation path for the robot within the building. Based on the building's elevator service group data, it is determined whether the robot's path crosses multiple elevator service group areas. Potential transfer floors are identified at the boundary between the path and elevator services, serving as path cutting points. If the path is continuously located within the same elevator service group, no path segment is cut, and it is retained as a single path segment. If the path crosses different elevator service groups, a transfer segment is inserted at the service boundary, marked as a connecting path requiring a transfer. Each path segment is numbered according to its start and end floors, and the elevator group number on which the path depends is indicated. Service attribute extraction: For each divided path segment, its service attribute information is extracted, specifically including the elevator group number it belongs to; the start and end floor numbers it crosses; the estimated path segment running time window, such as the expected time range for entering and leaving the path segment; and resource type markings, such as "elevator segment" or "corridor segment". Establish scheduling dependencies in the path segment graph. If the execution of a certain path segment requires the completion of the previous segment as a prerequisite (e.g., the previous segment reaches the transfer floor via elevator), then establish a directed dependency edge between them to represent the scheduling constraints of the order of execution.

[0076] S3. Obtain the current transfer floor data, and estimate the transfer waiting time based on the current transfer floor data and the split path data to obtain the transfer waiting time estimate data;

[0077] Specifically, based on the predetermined execution order and planned start and end times of each robot path segment, the expected arrival time of each robot at a specific transfer floor is calculated, forming a time series set. Based on the expected arrival times of all robots at that transfer floor, the arrival frequency distribution per unit time is statistically obtained, forming a preliminary arrival density function. The arrival density function is smoothed using a kernel density estimation algorithm to generate a continuous clustering strength function, which represents the expected arrival density of robots at any given time at that transfer floor. For the elevator used at the target transfer floor, typical operating cycle parameters are extracted, including door opening time, passenger start time, running time, arrival dwell time, and door closing time. The average time required for each complete service cycle of the elevator is calculated, and based on this, the maximum number of robot services the elevator can complete per unit time is estimated, denoted as the elevator's maximum processing capacity. Based on the above clustering strength and elevator throughput capacity, a queuing model is evaluated at any given time. If the clustering strength at a certain time is higher than the elevator's maximum throughput capacity, the queue backlog is calculated, and the corresponding waiting time is estimated accordingly. The estimation function for waiting time is: divide the number of robots exceeding the throughput limit by the elevator's processing capacity, and multiply by the time required for each service cycle to obtain the estimated transfer waiting time for the robots at that moment. ,in Estimate the transit waiting time. For a moment The aggregation intensity value, This represents the elevator's maximum throughput capacity. The time taken for the complete service cycle of the elevator. Based on the transfer floors and scheduling time windows involved in each segment of the path resource segment diagram, the estimated transfer waiting time is allocated to the corresponding path segment.

[0078] S4. Construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; calculate the graph structure conflict based on the transfer map data to obtain path conflict data; generate a schedule based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations.

[0079] Specifically, each path segment is treated as a node in the graph. If two path segments have an elevator transfer relationship (i.e., they are logically dependent on each other and involve different service groups), a directed edge is introduced between the corresponding nodes. The edge weight is defined as the estimated transfer waiting time between the two path segments, and it incorporates resource congestion indicators (such as the resource occupancy rate of the target transfer node during that time period) to form the scheduling cost. Additional attribute information is added to each node, including the elevator service group number to which the path segment belongs, the elevator identifier, and the estimated usage time window. Furthermore, considering the multi-entry and multi-exit structure of the building, it supports the construction of an asymmetric graph structure with multiple in-degree / out-degree nodes.

[0080] For each path segment, a resource occupancy model in the spatiotemporal dimensions is established. This model includes the location segment corresponding to the path segment in the floor space (such as corridors, waiting areas, elevator car interiors, etc.) and the time window of the path segment within the execution cycle, forming a three-dimensional spatial-temporal resource occupancy representation. For any two path segments, it is determined whether they intersect in the time window and overlap or are adjacent in the spatial region (e.g., sharing an elevator door or passing through the same floor corridor). If both are true, they are identified as a conflict pair. For each conflict path pair, the spatial overlap and temporal intersection are calculated, and a conflict tension function is established accordingly. This function can be expressed as the tension value equal to the spatial overlap multiplied by the weight coefficient α, plus the temporal intersection multiplied by the weight coefficient β, i.e. ,in This represents the path conflict tension value. This is the spatial overlap weighting coefficient. Spatial overlap, representing path segments and The degree of spatial overlap, such as the shared proportion of corridor sections and elevator waiting areas. The time intersection degree weighting coefficient, The time intersection degree represents the path segment. and The degree of overlap in execution time windows, For path segment numbering, Number another path segment. Construct a path overlap tension matrix from the tension values ​​of all conflict pairs. Use a graph-based clustering algorithm (e.g., density clustering or spectral clustering) to extract conflict-dense subgraphs from the path overlap tension matrix, thereby identifying a set of path segments with high conflict coupling and forming several core conflict groups.

[0081] For all scheduled path segments, a scheduling scoring function is constructed based on factors such as task priority indicators, the total tension value of their respective conflict groups, and required waiting time. The path segments are then topologically sorted according to the scoring function, determining the execution order of each path segment while ensuring the correctness of dependencies. Based on the sorting results and the transfer graph structure, a scheduling time window accurate to the path segment level is generated for each robot, and an elevator resource scheduling plan is simultaneously constructed. This plan is distributed to the floor intelligent terminals, elevator control system, and robot execution units via the cloud platform, ensuring consistency of multi-agent behavior. During actual scheduling execution, if unexpected events occur, such as temporary elevator malfunctions, robot deviations from the path, or changes in the floor environment, the system automatically triggers a dynamic scheduling recalculation mechanism to adjust the scheduling plan in real time to restore scheduling stability and coordination.

[0082] Optionally, the building elevator service group division includes:

[0083] S11. Extract the elevator service relationship of each floor based on the building elevator group data to obtain the floor elevator matrix data;

[0084] Specifically, the system extracts elevator group configuration information from the building control platform or its supporting equipment information sources. This information can come from the Building Information Modeling (BIM) system, building automation control system logs, property elevator operation configuration database, or control logic parameters provided by the elevator manufacturer. Key attributes of each elevator include, but are not limited to, elevator number (used to uniquely identify an elevator); group control number (indicating whether the elevator belongs to a certain control group); service floor list (listing the floors covered by the elevator, e.g., serving floors 1 to 10, or only serving floors 1, 3, 5, and 7); service time period (if time-limited service rules exist, the service range of the elevator is recorded at different time periods, such as direct service during peak hours); assuming the building contains L floors, numbered from floor 1 to floor L; and there are E elevators, numbered from elevator 1 to elevator E, a floor-elevator service matrix can be constructed. For floor i and elevator j, if elevator j serves floor i, then "1" is entered in the corresponding position to indicate a service relationship; if elevator j does not serve floor i, then "0" is entered to indicate no service relationship. The matrix described above has L rows and E columns, where each row represents the service status of a floor, and each column represents the service coverage of an elevator. Additional label information is attached to each service cell in the matrix, including but not limited to the following: service priority label, indicating the elevator's priority for that floor (e.g., "high priority" indicates frequent stops or primary service); operation status label, recording the current or historical operation status, such as "available," "under maintenance," or "busy"; and control logic label, indicating whether the service mode is "express elevator," "normal stop elevator," or "restricted stop elevator," etc.

[0085] S12. Obtain elevator operation behavior data, and calculate the service overlap frequency based on the elevator operation behavior data to obtain service overlap frequency data.

[0086] Specifically, the system extracts historical operation log data from the building elevator control platform (such as elevator control systems from brands circulating in the market). The time period for log recording can be set according to actual business needs (such as the past day, the past week, or the past month). Each log record includes, but is not limited to, the elevator number, indicating which elevator performed the operation; the operation timestamp, the specific stopping time, accurate to the second or minute; the elevator direction of travel, such as up, down, or empty return; the stopping floor number, the floor reached during this operation; the full load flag, indicating whether the elevator is fully loaded and whether there is a transport task during this stop; and the request type, such as triggered by a passenger's external call (floor button) or internal call (car button), etc. Service overlap frequency is the frequency at which any two elevators stop at the same floor approximately simultaneously. For any pair of elevators, elevator 1 and elevator 2, within a set observation period (e.g., 24 hours), the number of events in which they stop simultaneously on each floor is counted. Simultaneous stopping is defined as the time difference between the two elevators stopping at a certain floor not exceeding a set threshold (e.g., 60 seconds). The sum of the statistical results for all floors is then divided by the total observation time (units can be minutes or hours). The service overlap frequency between elevator 1 and elevator 2 is the ratio of the total number of approximately simultaneous stopping events of any two elevators on all floors to the total observation time. After completing the above overlap frequency statistics, the system can construct a service overlap frequency matrix between elevator pairs. This matrix is ​​symmetric, with the number of rows and columns equal to the total number of elevators. The value in the i-th row and j-th column of the matrix represents the service overlap frequency between the i-th elevator and the j-th elevator. The system sets an overlap frequency threshold (e.g., 0.002 times / minute) to filter out elevator pairs with extremely low service overlap. The retained high-overlap elevator pairs will be used as key connection edges in the subsequent service connectivity graph construction.

[0087] S13. Construct a floor elevator connectivity graph based on the floor elevator matrix data and service overlap frequency data to obtain building elevator service group data.

[0088] Specifically, an undirected weighted graph is constructed based on the floor elevator service matrix and the service overlap frequency matrix. Each elevator corresponds to a node in the graph, and the basic attributes of the node include elevator number, group control number, and service floor range. Two elevators have a common service record on at least one floor, that is, they are both marked as "1" in the floor elevator matrix for that floor. The service overlap frequency of the two elevators is higher than a preset threshold (e.g., 0.001 times / minute). Based on the basic edge weights, the following feature attributes are further extracted and added to each edge, such as the number of shared floors (the total number of floors jointly served by the two elevators); the degree of overlap in service areas (by calculating the intersection and difference of the service floors served by the two elevators); the similarity score of the running trajectory (calculated based on the cosine similarity of the embedding vectors of the running trajectory or stopping pattern); and the group control logic consistency mark (determining whether the two elevators are scheduled by the same group control system, which can be represented by a Boolean value mark (yes / no). When there are consecutive floors that do not overlap in an elevator group (e.g., one group serves floors 1–10, and another group serves floors 12–20), it is defined as a service discontinuity. If two elevators belong to different group control systems and their service areas are not continuous, the discontinuity determination is enhanced. The service connectivity graph is segmented to divide areas with weak connectivity and obvious discontinuities into different subgraphs.

[0089] Optionally, the construction of the floor elevator connectivity graph includes:

[0090] Based on the floor elevator matrix data and service overlap frequency data, structural coverage feature extraction and elevator collaborative scheduling feature extraction are performed to obtain structural coverage feature data and elevator collaborative scheduling feature data, respectively.

[0091] Specifically, the structural coverage features include for each elevator Extract its service floor set Its coverage area width (i.e., coverage span) is defined as ,in It is a function with maximum value. For elevator Service floor collection, The function is the minimum value; calculate the elevator. Floor coverage density It is defined as the ratio of the number of floors it serves to the coverage area: ,in To achieve dense floor coverage, For elevator Service floor collection, To determine the service area span; to determine whether each elevator serves the ground floor (e.g., floor 1) or the top floor (e.g., floor L) to identify the coverage capacity of key vertical boundaries; and to identify whether there are cross-fault structures in the service area, i.e. whether there are continuous gaps in the service floors, as a potential indicator of structural faults.

[0092] Elevator collaborative scheduling feature extraction includes for any pair of elevators Extract the intersection of its service floor set. Calculate the number of overlapping service floors ;for The standard deviation of the floor location is calculated to determine whether the overlap is concentrated in a certain floor segment (such as the middle or lower floors). The smaller the standard deviation, the more concentrated the overlap. The elevator control system's operation log is also analyzed. and The frequency of simultaneous service requests on shared floors. Determine whether elevator pairs belong to the same group control system (e.g., shared scheduling logic or belonging to the same controller), represented by a Boolean value (yes / no).

[0093] Based on structural coverage feature data and elevator collaborative scheduling feature data, a coupling graph is constructed from the floor elevator matrix data to obtain the coupling graph data.

[0094] Specifically, each elevator is considered as a node in the graph, and for any elevator pair... If their service floor sets have an intersection (i.e., the number of overlapping floors) ), then at the corresponding node and Establish an undirected edge between them Assign a correlation weight to this edge. The weight value is composed of the following three weighted combinations: ,in Let the edge weights be the weights between elevator nodes. This is weighted data for the number of overlapping floors. The number of overlapping floors. For elevator Service floors collection, For elevator Service floors collection, For coordinated scheduling frequency weight data, For coordinated frequency scheduling, To control the structural consistency weighting coefficient, To control structural consistency indicators. This refers to the elevator node sequence item. This is another elevator node sequence item. Attach the following attribute information to each graph node: service coverage density, service area span, boundary coverage status, and fault identification auxiliary information.

[0095] Fault zone detection is performed on the coupled graph data to obtain authorized control graph data. Fault zone detection includes floor jump fault detection, floor no access record detection, and authorization restriction detection.

[0096] Specifically, for each elevator The corresponding service floor set If there exists any pair of adjacent floors satisfy If the elevator node is not found to have a floor-jumping problem, then the elevator is considered to have a floor-jumping problem. If there is an edge in the graph originating from the elevator node, and the elevator corresponding to the target node also exhibits jumping behavior, then the edge weight is reduced. ,in These are the edge weights after the layer decay. This is the jump penalty coefficient, with a value ranging from 0.5 to 0.8. The original edge weights.

[0097] All floors in the building The system checks whether a floor is covered by at least one elevator. It did not appear in any of the elevator service sets, that is... , For elevator If the service floor set is determined, then a "physical discontinuity" is identified on that floor; for all service floors adjacent to this discontinuity (such as...), the service floor set is determined. or Add fault labels to elevator nodes to restrict their participation in the service group clustering process of the graph, and prevent incorrect connectivity judgments due to missing local services.

[0098] Extract the access control configuration of building elevators, including but not limited to floor access permissions for different roles such as employee elevators, freight elevators, and visitor elevators. Data sources can be the building control system configuration table or elevator controller parameter sets. For floors with access restrictions, mark their corresponding elevator-floor service pairs as "logical faults," meaning only robot tasks with specific permissions are allowed to traverse them. This structure forms "semi-fault" dividing zones in the graph, requiring additional policy judgment during path generation. Add the identified fault regions as attribute layers to the original coupled graph, labeling them as unreachable edges, nodes not participating in clustering, or logical boundaries requiring special policy processing.

[0099] Apply time-period service mode switching constraints to the authorized control chart data to obtain constraint chart data;

[0100] Specifically, scheduling time periods are divided based on actual building operation patterns and robot operation density. For example, morning peak hours: 8:00 AM to 10:00 AM; off-peak hours: 10:00 AM to 4:00 PM; evening peak hours: 4:00 PM to 6:30 PM; and low-frequency nighttime hours: 7:00 PM to 6:00 AM the next day. During different time periods, some elevators operate in different modes, including directional restrictions (e.g., only upward movement during morning peak hours, only downward movement during evening peak hours); service suspension (e.g., shutting down some non-main elevators at night for energy conservation); and service priority adjustments (e.g., prioritizing freight or visitor requests). [This is related to the authorization control chart.] Each edge in Define the corresponding set of active time periods. ;like This indicates that the route is available during morning and evening rush hours; if a certain side is accessible all day, then... If an edge is unreachable or disabled during a certain time period, it will be blocked or its direction restricted during that time period. If only one-way traffic is supported during a certain time period (e.g., an elevator only goes up), then the direction of the corresponding edge in the diagram must be set to "one-way edge," meaning it only allows traffic from... arrive If a path is suspended for a certain period of time, the edge will be set as "conditionally unreachable edge" or completely removed during that period; the edge attribute records the "period dependency label".

[0101] Connectivity branch extraction is performed based on constraint graph data to obtain building elevator service group data.

[0102] Specifically, based on the time-stamped and direction-constrained graph structure generated in the previous step (denoted as the time-constrained graph), connectivity analysis is performed for each preset scheduling time period (such as morning peak, off-peak, and evening peak). Edges in the graph that are only "valid" within the current time period are considered usable edges. Within this time period, all strongly connected subgraphs (i.e., bidirectional reachable paths exist between any two nodes) are extracted from the graph. Each strongly connected subgraph represents an elevator service unit with complete reachability within this time period. For each extracted connected subgraph, its service overlap index (such as the number of shared floors, the number of elevators, etc.) is calculated. If the service overlap of a subgraph is lower than a preset threshold, it is judged as a redundant service group or an auxiliary path. Redundant subgraphs can be merged into adjacent main service groups or marked as "auxiliary service groups" for low-frequency scheduling or abnormal fallback path selection. During merging, consistency in time period and direction is maintained to avoid introducing cross-time period disconnected structures.

[0103] Optionally, the path resource segment cutting includes:

[0104] S21. Divide the robot target path data and building elevator service group data into road segments to obtain path segmentation data;

[0105] Specifically, the complete path sequence of the robot from the starting floor to the target floor is denoted as the floor node sequence; the building elevator service group data represents a set composed of multiple elevator service units; each node in the path is traversed. Determine the service group to which it belongs. If floor nodes Belongs to service group Service floor set (i.e. If the node is selected, then it will be assigned to the appropriate category. The corresponding path segment. For adjacent node pairs in the path ( ) to perform analysis, if they belong to different service groups (i.e. , ,and If the node pair is not specified, then the location between them is taken as the transfer breakpoint. The entire path is divided into several segments according to the transfer breakpoints, and each segment is a separate segment. Includes the starting and ending floors; the elevator service group number to which the route belongs; such as the route direction (up / down), route length (expressed as floor difference or spatial distance), and estimated time window, etc.

[0106] S22. Based on the path division data, extract service attributes and transit points to obtain service attribute data and transit point data respectively;

[0107] Specifically, for each path segment in the path segmentation data, the following service attribute features are extracted from the building elevator management system, historical operation records, or static configuration data to construct service attribute data, including: elevator group identifier: recording the elevator service group ID that this path segment depends on; elevator type label: including classification information such as freight elevator, passenger elevator, high-speed elevator, and low-speed elevator; control logic information: distinguishing whether the elevator is in group control scheduling or independent operation mode; access performance indicators: statistically analyzing the average travel time, peak waiting time, failure rate, and other time-related indicators of this elevator group in historical data; robot accessibility: indicating whether the path segment allows robot access, for example, some elevators are restricted to manual use, or are limited by load, sensors, etc. The boundary between path segments is the potential transfer point. The system extracts transfer information from the connection points of adjacent path segments to generate a transfer point dataset, including location identifiers, floor numbers corresponding to the transfer points, service group connection pairs (representing the two elevator service group pairs connected by the transfer point, used to identify the source and destination of the transfer path), transfer condition evaluation indicators (such as whether physical connectivity is available, such as whether they are on the same transfer platform or whether there are inter-floor passages), transfer passage length (meters), minimum width of the transfer area (meters), historical transfer congestion index (which can be defined as the number of pedestrians / robots per unit time), and whether continuous robot path recognition is supported (such as barrier-free platforms, gateless barriers, etc.).

[0108] S23. Based on the service attribute data and transfer point data, perform transfer constraint dependency processing on the path division data to obtain path dependency data;

[0109] Specifically, dependency relationships are established, including time dependency marking, spatial mutual exclusion dependency marking (if the physical area where the transfer point is located has a narrow structure (such as a single channel or narrow platform), the system prohibits two or more robots from occupying it simultaneously, exhibiting exclusivity; the judgment criteria include the spatial width index of the transfer point, the transfer congestion index, etc.), and resource interlock dependency handling (when two path segments need to share a set of resources (such as elevators, access control, elevator scheduling slots, etc.), they must be executed serially in a mutually exclusive scheduling manner). For each type of valid dependency relationship, the system constructs directed dependency edges between path segments.

[0110] S24. Construct a path dependency graph based on the path dependency data to obtain the split path data.

[0111] Specifically, based on path dependency , build a line from point to The system uses directed edges; each edge is labeled with a dependency type such as time dependency, resource lock dependency, or spatial conflict; scheduling cost such as waiting time estimation and resource contention risk; dynamic waiting time estimation: if there is a congestion probability at the transfer point, the expected waiting time is estimated based on the current robot density and scheduling load; when there is a path segment traveling between high and low floors (such as high→low→high), the system performs graph traversal detection to identify whether a resource occupancy deadlock loop has been formed; for the detected potential deadlock path, the conflict loop path is recorded and a penalty weight is added or path avoidance processing is performed.

[0112] Optionally, the transit wait estimation includes:

[0113] S31. Based on the current transfer floor data and split path data, perform multi-robot arrival time superposition calculation to obtain arrival superposition data;

[0114] Specifically, the statistical time window length is set to... (e.g., 5 seconds) The time axis is discretized to form a sequence of time segments. For each transfer floor Statistics within a time period The estimated number of robots to reach this floor is defined by the arrival superposition function. ,in For at any time to Reaching the floor during the period The total number of robots Assign a number to the robot. This is an indicator function. If the condition in the expression is true (i.e., the arrival time of robot r falls within the current time window), this term takes the value 1; otherwise, it takes the value 0. For the robot's arrival timestamp, For the transfer floor number, This is the starting point of the current time slice. Let be the length of the time window. Perform the above statistical operations on all time segments t to form the multi-robot arrival density sequence for each floor, denoted as . This sequence reflects the transit floors during different time periods. The robot arrives at the density dynamic evolution process.

[0115] S32. Calculate the kernel density based on the arrival superimposed data to obtain the local aggregation data;

[0116] Specifically, given the transfer floor Arrival time series Its kernel density function is defined as follows: ,in For at any time When the robot arrives at the transfer floor The local aggregation amount, This represents the total number of robots that reached this floor. The bandwidth parameter for kernel density estimation represents the smoothing window size (e.g., h = 5 seconds). For robot sequence items, For kernel function, , Pi is a constant. The base of the natural logarithm, To standardize time deviation, Estimate the time for the target. For: Robot Arrive at the floor Timestamp for each transfer floor. Output its local aggregation function over continuous time.

[0117] S33. Calculate the throughput and the opening gate cycle based on the current transfer floor data, and obtain the throughput data and the opening gate cycle data respectively.

[0118] Specifically, transfer throughput refers to the maximum number of entities (e.g., robots) that can pass through a certain transfer floor under ideal scheduling conditions per unit time, which is limited by the capacity and service cycle of the elevators or access control systems connected to that floor. For transfer floors... If it is from (Elevator quantity marker) [Number of elevators or other transfer devices] provide services, and the maximum load capacity of each device is [Number of elevators or other transfer devices]. The average time for an elevator to complete one service cycle is Then the maximum throughput of that floor It can be represented as ,in The maximum throughput of the floor. This refers to the maximum number of people or robots that can be carried in a single elevator at one time. To serve the floor The number of elevators, The complete service cycle duration includes the process of opening, entering / exiting, closing, and moving to the next floor. The door opening / closing cycle is estimated as the time it takes for the elevator to perform one complete door opening / closing operation, defined as the sum of the door opening time, passenger (or robot) entry / exit time, and door closing time. Parameters can be set empirically, obtained from historical operation log statistics, or real-time sensor data.

[0119] S34. Based on the output data and the opening gate cycle data, estimate the transfer waiting time of the local aggregation data to obtain the transfer waiting estimation data.

[0120] Specifically, when the current cluster density exceeds the elevator's throughput capacity, the robot needs to queue for multiple service cycles, causing the waiting time to increase non-linearly. (Floor) At any moment The estimated waiting time is ,in For floors At any moment The estimated waiting time As an adjustment factor, for example, a larger value of 1.5 is set for peak periods (such as 08:00–09:00), and 0.7 is set for off-peak periods. This represents the kernel density clustering value within the current time window. This represents the maximum number of robots that can be processed per unit of time. This is the time required for one round of opening, entering, exiting, and closing of the door. If... This indicates that the current floor is overloaded; to avoid short-term peak interference, the continuous time window is... Perform moving average or Gaussian weighted smoothing; output is floor level. An estimated sequence of waiting time distribution within a future scheduling period.

[0121] Optionally, the transfer map construction includes:

[0122] Based on the current transfer floor data and the split path data, a path segment transfer map is constructed to obtain the path segment transfer map data;

[0123] Specifically, a heterogeneous two-layer graph structure is constructed, which consists of two types of nodes, including path segment nodes, each path segment... Abstracted as a path node Transfer resource nodes, transfer elevators on each floor. Abstracted as a transit node For each path segment If its path requires passing through an elevator Achieve from floor arrive For services like this, edges are created in the graph as follows: for example, an edge points from a path segment node to the elevator node it depends on. This indicates that elevators need to be dispatched for this route segment. ; Point from the elevator node to the next path segment node This indicates that elevator transfer resources are available for use in the next path segment. If multiple elevators need to work together for a transfer (such as cross-zone transfers), a multi-hop dependency chain is constructed to achieve cascading connections between elevator nodes. The nodes in the graph are heterogeneous, and the edges contain directionality and functional meaning (dependency, transfer). It is expanded into an attributed graph structure, with edges carrying additional information such as estimated time, floor displacement distance, and congestion weight. The graph structure is the initial scheduling topology for task path dependencies and transfer resource interactions.

[0124] Based on the estimated transit waiting data, the transit graph data of the path segment is weighted to obtain weighted graph data.

[0125] Specifically, for each directed edge in the graph, it is identified as belonging to one of the following two categories: edge from a path segment node to a transfer node, indicating that the path segment depends on a certain elevator service; edge from a transfer node to the next path segment node, indicating the continuation of the path after the transfer resource completes its service. For each identified edge, its scheduling cost is calculated as the edge weight. The edge weight includes a static time component, including elevator door opening time, elevator passenger boarding and alighting time, elevator door closing and departure preparation time, etc., collectively referred to as the basic transfer time, the value of which can be obtained from historical operating data or preset standards; a dynamic waiting component, which is the estimated waiting time from the current transfer floor, representing the queuing waiting time caused by the current transfer elevator being in a busy state; and a delay adjustment factor, which is an adjustment coefficient (between 1 and 2) introduced to reflect the delay amplification effect of high-load or high-congestion floors. The scheduling cost (i.e., edge weight) of each edge is expressed as: ,in The scheduling cost of an edge (i.e., the edge weight). This is the static time consumption portion. As a delay adjustment factor, This is the dynamic waiting part. All edge weights are positive real numbers; the larger the value, the higher the expected scheduling cost. If a transit floor is in a low-load state during the current period, its corresponding edge weight tends to the basic static time consumption. If the edge connects to a highly congested floor or a single-channel transit node, the dynamic part increases significantly, reflecting a potential scheduling bottleneck. The output is a weighted path segment transit graph.

[0126] The weighted graph data is directed and scheduled to obtain the transfer graph data.

[0127] Specifically, for all the structural patterns of "path segment node → transit node → path segment node" in the graph, the scheduling order dependency between path segments is extracted; if a path segment exists... With path segment All rely on a certain transit resource And the scheduling logic requires Must precede If execution occurs (e.g., due to time window order, resource exclusivity, or priority differences), then a directed edge is constructed. This represents the execution constraints. All transit nodes (such as elevator nodes) are logically collapsed into edge constraints between path segments; that is, they are not retained as explicit nodes, but their scheduling impact is reflected in the edge weights and dependencies between path segments. The scheduling delay and queuing impact introduced by transit resources are retained. The system simplifies the graph, such as merging continuous and uninterrupted path segments into single nodes; it identifies and marks possible dependency loop structures in the graph to prevent deadlocks in task scheduling; and it identifies all strongly connected subgraphs as independent schedulable units.

[0128] Optionally, the graph structure conflict calculation includes:

[0129] Path occupancy is mapped based on the transfer map data to obtain path occupancy data;

[0130] Specifically, for each path segment, construct its resource consumption description tuple, denoted as . , of which each Represents path segment The usage status of the j-th resource node during execution is represented by a triple: ,in The identifier for the occupied resource node includes, but is not limited to, elevator shaft number, door zone buffer zone identifier, floor corridor number, etc. This refers to the point in time when the path segment begins to occupy the resource. The time point at which the resource is released for the path segment. The path segment includes: elevator car resources (representing the area occupied by the robot or passenger moving vertically within the elevator); elevator door buffer zones (representing the area where the robot briefly pauses before and after entering or exiting the elevator doors); and floor corridors (representing the path area occupied by the robot moving horizontally between floors). Time interval. It can be calculated from the estimated start and end times of the scheduling path; if it includes waiting or congestion delays, the corresponding resource occupancy window should include the estimated waiting time period.

[0131] Conflict pairs are extracted based on path occupancy data to obtain conflict pair data.

[0132] Specifically, for all path segment pairs ,in This involves comparing each pair of paths one by one, extracting the set of resource nodes they occupy and their corresponding time windows. For each pair of path segments... A set of conflict pairs is formed if all of the following conflict determination conditions are met: There exists a resource node l such that this node appears in the same location as the resource node l. and In the resource usage description, that is: , This refers to the set of resources occupied by a path segment. This represents the occupancy status of resource nodes for a path segment; the occupancy time on the same resource node overlaps, i.e. ,in For path segment The start and end times on this resource, For path segment The start and end times for the resource are defined. Based on the resource's shareability and scheduling strategy, conflict types are categorized and labeled. If the conflicting resource is strictly mutually exclusive (e.g., an elevator car), it is marked as "strong conflict"; if the conflicting resource is partially shareable (e.g., a floor corridor), it is marked as "weak conflict"; other resource types can be further defined according to the actual system rules. For path segment pairs that meet the above conditions, a conflict pair tuple is constructed. ,in This is a conflict type label.

[0133] Based on the conflict data, the path overlap tension is calculated to obtain the path conflict data.

[0134] Specifically, for each set of conflict pairs Based on its shared resource set The path conflict tension function is defined as follows: ,in For path conflict tension data, For shared resource nodes, For path segment and A collection of shared resources This is the overlap ratio weighting coefficient. For resource nodes The overlap ratio of path segment space occupancy, such as the ratio of overlapping length to total resource length in corridor resources. The overlap ratio weighting coefficient is used. For resource nodes The overlap ratio of occupied time intervals is calculated as the ratio of the length of the temporal intersection of two path segments at the resource node to the length of any single segment. For resources with fuzzy physical boundaries or buffer zones, a Gaussian kernel function is used to model the conflict influence radius; spatial overlap and temporal overlap are smoothed by embedding them into their respective kernel functions. The tension values ​​of all path segment pairs form the conflict tension matrix. Each element Represents path segment and The intensity of the conflict between them.

[0135] Optionally, the path overlap tension calculation includes:

[0136] Based on the conflict data, non-rigid geometric overlap and time occupancy intersection are calculated to obtain fuzzy boundary overlap data and time occupancy intersection data respectively.

[0137] Specifically, in a three-dimensional environment, the path segment , The occupied space is constructed as a fuzzy envelope region, and its spatial influence range is defined using a fuzzy function (such as a Gaussian function). In the shared resource domain... In areas such as corridors and elevator shafts, construct spatial influence functions for each path segment: ,in, Let represent a spatial location point, and let the function value represent the intensity of the influence of the path segment on that location, which is a function of the distance from the center trajectory. The overlap integral is the non-rigid geometric overlap of the two path segments at the resource node, which is the integral of the product of their fuzzy occupancy functions over the spatial domain. ,in For fuzzy boundary overlap, For spatial location points, For shared resource space domain, For path segment Spatial influence function, , It is a natural exponential function. For path segment The central trajectory position, The kernel function bandwidth parameter represents the radius of path influence. For path segment The spatial influence function is similar in form to the former. This value characterizes the degree of spatial overlap or interference between two path segments; a larger value indicates more severe spatial overlap. The temporal intersection degree is the ratio of the intersection length of two time windows to their respective shortest durations: ,in The intersection degree of time occupation. It is a function with maximum value. It is a minimum value function. For path segment The initial occupancy time, For path segment The time taken to finish For path segment The initial occupancy time, For path segment The time taken to complete the process. If the time intervals do not overlap, the result is 0; if they completely overlap, the result is 1.

[0138] Based on the fuzzy boundary overlap data and the time occupation intersection data, the shared resource node competition data is processed to obtain node competition data.

[0139] Specifically, for each conflict pair Shared resource nodes in Calculate the unit resource contention density on this node, defined as follows: ,in The intensity of competition for node resources per unit. For path segment and In resources The degree of overlap of the fuzzy boundary space on the surface. For path segment and In resources The time occupied by the intersection degree, For resource nodes The type weights are calculated (e.g., elevator car = 1.0, elevator door buffer zone = 0.7, floor corridor = 0.5, elevator door opening / closing mechanism = 0.6, transfer junction = 0.4, general waiting area = 0.3). The competition intensity of each path segment across all its shared resource nodes is summed to obtain its node competition score. ,in For path segment pairs Total resource competition pressure score For a single resource node, For a set of shared resource nodes, The intensity of competition for node resources.

[0140] Based on node contention data, resource occupancy tension is propagated to obtain occupancy propagation matrix data;

[0141] Specifically, the path segment With resource nodes Establish a bidirectional connection; if the path segment is paired with... There are shared resource nodes Then it is treated as an edge in the bipartite graph; each edge The edge weights are set as the local competition intensity. That is, the path segment is at the resource node The impact of unit competition on the upper part. Introducing the graph propagation function. Weighting the collision intensity is used to control propagation attenuation (e.g.: ,in (for each path segment) Statistical analysis of its relationship with all other path segments The cumulative propagation tension value between them is defined as follows: ,in For path segment Resource utilization tension value, For path order item , For path order item , The propagation function result for the competition intensity represents the degree to which the conflict intensity contributes to the tension. To indicate path segment and Whether there is an association in the scheduling graph (such as shared resources or execution order dependencies). If there is an association, the value is 1; otherwise, the value is 0. This represents the global path segment conflict intensity. The output is the path segment tension propagation matrix. Each element represents the cumulative resource conflict pressure that a path segment experiences in the current scheduling structure.

[0142] Based on the occupancy propagation matrix data, high-priority task scheduling gravity control is performed to obtain path conflict data.

[0143] Specifically, for each path segment The system reads the scheduling priority value of the corresponding task. This value can come from the task definition library, manual configuration, or intelligent classification results; a higher priority value indicates a more important task, and the more resource allocation to the scheduling path should receive preferential weight. A gravity regulation model is used to adjust the tension value. The tension value is obtained by performing priority weighting correction. ,in For path segment Conflict tension after considering task priorities For path segment conflict tension, This is the priority gravity adjustment coefficient, which controls the sensitivity of the scheduling strategy to the task level. If set... Then, a task with priority 3 will receive an additional 60% scheduling tension on top of the tension propagation. This represents the priority level value of the task corresponding to the path segment. Each path segment... Its corresponding adjusted tension value Combine and construct a set of path conflict data. This set will serve as the basic input for subsequent path segment sorting priority determination, conflict avoidance strategy adjustment, and multi-task scheduling optimization.

[0144] Optionally, the schedule generation includes:

[0145] S41. Identify conflict core groups based on path conflict data to obtain conflict core group data;

[0146] Specifically, based on path conflict data Construct a conflict graph , where the node set For all path segments participating in scheduling; edge set For path pairs with conflicting relationships; each edge The edge weight is the corresponding tension value The completed conflict graph The algorithm uses graph clustering to divide conflicting and coupled paths into groups and identify areas of high local tension. Each node forms a community. The algorithm attempts to move the current node into the community of its neighbors. If the local modularity gain is maximized, the nodes are merged. The modularity function is based on edge weights and automatically clusters high-tension paths into the same conflict community. The algorithm terminates when the modularity no longer increases significantly and outputs several core conflict groups.

[0147] S42. Based on the conflict core group data, perform path-level scheduling sorting to obtain scheduling sorting data;

[0148] Specifically, for each path segment Calculate its scheduling priority Specifically ,in To prioritize data scheduling, This is the priority score weighting coefficient, with a value of 0.5. Assign a priority score to the task corresponding to the path (e.g., set the priority of a medicine delivery task to 5, and a routine inspection task to 2). This is the total conflict tension weighting coefficient, with a value of 0.3. The total conflict tension between the path segment and other paths within the current group. This is the weighting coefficient for the cumulative waiting time estimate, with a value of 0.5. This is an estimate of the cumulative waiting time for this path segment within the current transit structure. For each set of paths within a group, this is calculated based on the score. Sort the paths from highest to lowest. If the scores are similar (the difference is less than a set threshold), additional sorting criteria can be introduced, such as total path length; whether the path crosses floors (cross-floor paths take priority); task initiation time (first-come, first-served), etc. This results in a sorted path scheduling sequence corresponding to each conflict core group.

[0149] S43. Generate robot scheduling data based on the scheduling order to perform multi-robot collaborative scheduling operations.

[0150] Specifically, for each path segment, a planned execution time window is assigned, satisfying the following conditions: It should not overlap with other path segments in the same group in terms of resource usage time (avoiding conflict); the robot's state before allocation should be considered (e.g., whether it is still executing a previous task, whether it is in an elevator, etc.); if the path segment involves elevator transfer operations, it must be confirmed that elevator resources are available for scheduling within the estimated time window to avoid duplicate elevator usage; and the estimated path time (e.g., floor travel time, waiting time, elevator usage delay) should be considered for reasonable allocation. For each robot, a scheduling execution plan is generated based on its assigned path segment and corresponding time window; this plan should include a structured set of operation instructions, with instruction types including START: start executing the scheduling task; MOVE_TO: move to the target location; WAIT: wait due to resource (e.g., elevator, corridor) unavailable; ENTER_ELEVATOR / EXIT_ELEVATOR: elevator entry / exit actions; FINISH: complete the path segment task. Each instruction should include the operation type; start and end times; location identifier or resource ID; supplementary reason (such as waiting for resource occupation to end) or additional information (such as target floor, robot number, etc.).

[0151] Optionally, this application also provides a multi-robot cloud scheduling and collaboration system for executing the multi-robot cloud scheduling and collaboration method described above, the multi-robot cloud scheduling and collaboration system comprising:

[0152] The elevator service group division module is used to acquire building elevator group data and divide the building elevator service groups based on the building elevator group data to obtain building elevator service group data.

[0153] The path segment structure modeling module is used to acquire robot target path data; it also performs path resource segmentation on the robot target path data and building elevator service group data to obtain split path data.

[0154] The transfer waiting prediction module is used to obtain the current transfer floor data and estimate the transfer waiting based on the current transfer floor data and the split path data, thus obtaining the transfer waiting estimate data.

[0155] The conflict-aware scheduling generation module is used to construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; to calculate the graph structure conflict based on the transfer map data to obtain path conflict data; and to generate scheduling data based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations.

[0156] Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended application documents rather than the foregoing description. Thus, it is intended that all changes falling within the meaning and scope of the equivalents of the application documents be incorporated into the invention.

[0157] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A multi-robot cloud scheduling and coordination method, characterized in that, Applied to a cloud platform, the method includes: S1. Obtain building elevator group data, and divide the building elevator service groups according to the building elevator group data to obtain building elevator service group data. S2. Obtain robot target path data; perform path resource segmentation on robot target path data and building elevator service group data to obtain split path data; S3. Obtain the current transfer floor data, and estimate the transfer waiting time based on the current transfer floor data and the split path data to obtain the transfer waiting time estimate data; S4. Construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; calculate the graph structure conflict based on the transfer map data to obtain path conflict data; generate a schedule based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations. The construction of the transfer map includes: Based on the current transfer floor data and the split path data, a path segment transfer map is constructed to obtain the path segment transfer map data; Based on the estimated transit waiting data, the transit graph data of the path segment is weighted to obtain weighted graph data. The weighted graph data is directed and scheduled to obtain the transfer graph data.

2. The method according to claim 1, characterized in that, The building elevator service group division includes: The elevator service relationship between floors is extracted from the building elevator group data to obtain the floor elevator matrix data. Elevator operation behavior data is acquired, and service overlap frequency is calculated based on the elevator operation behavior data to obtain service overlap frequency data. Based on the floor elevator matrix data and service overlap frequency data, a floor elevator connectivity graph is constructed to obtain building elevator service group data.

3. The method according to claim 2, characterized in that, The construction of the floor elevator connectivity graph includes: Based on the floor elevator matrix data and service overlap frequency data, structural coverage feature extraction and elevator collaborative scheduling feature extraction are performed to obtain structural coverage feature data and elevator collaborative scheduling feature data, respectively. Based on structural coverage feature data and elevator collaborative scheduling feature data, a coupling graph is constructed from the floor elevator matrix data to obtain the coupling graph data. Fault zone detection is performed on the coupled graph data to obtain authorized control graph data. Fault zone detection includes floor jump fault detection, floor no access record detection, and authorization restriction detection. Apply time-period service mode switching constraints to the authorized control chart data to obtain constraint chart data; Connectivity branch extraction is performed based on constraint graph data to obtain building elevator service group data.

4. The method according to claim 1, characterized in that, The path resource segment cutting includes: The robot's target path data and the building elevator service group data are divided into road segments to obtain path segmentation data; Based on the path segmentation data, service attributes and transit points are extracted to obtain service attribute data and transit point data, respectively. Based on service attribute data and transfer point data, the path partitioning data is processed to perform transfer constraint dependency processing to obtain path dependency data; Path dependency graphs are constructed based on path dependency data to obtain split path data.

5. The method according to claim 1, characterized in that, The transit waiting time estimate includes: Based on the current transfer floor data and split path data, the arrival times of multiple robots are superimposed to calculate the arrival superimposed data. Based on the arrival and overlay data, kernel density is calculated to obtain local clustering data; Based on the current transfer floor data, the throughput and the opening gate cycle are calculated to obtain the throughput data and the opening gate cycle data respectively. Based on the output volume data and the opening gate cycle data, the transfer waiting time of the local aggregation volume data is estimated to obtain the transfer waiting time estimate data.

6. The method according to claim 1, characterized in that, The graph structure conflict calculation includes: Path occupancy is mapped based on the transfer map data to obtain path occupancy data; Conflict pairs are extracted based on path occupancy data to obtain conflict pair data. Based on the conflict data, the path overlap tension is calculated to obtain the path conflict data.

7. The method according to claim 6, characterized in that, The calculation of the path overlap tension includes: Based on the conflict data, non-rigid geometric overlap and time occupancy intersection are calculated to obtain fuzzy boundary overlap data and time occupancy intersection data respectively. Based on the fuzzy boundary overlap data and the time occupation intersection data, the shared resource node competition data is processed to obtain node competition data. Based on node contention data, resource occupancy tension is propagated to obtain occupancy propagation matrix data; Based on the occupancy propagation matrix data, high-priority task scheduling gravity control is performed to obtain path conflict data.

8. The method according to claim 1, characterized in that, The scheduling generation includes: Based on the path conflict data, the core conflict groups are identified to obtain the core conflict group data. Based on the conflict core group data, path-level scheduling is sorted to obtain scheduling sorting data; Robot scheduling data is generated based on the scheduling order to enable multi-robot collaborative scheduling operations.

9. A multi-robot cloud scheduling and collaborative system, characterized in that, For executing the multi-robot cloud scheduling and collaboration method as described in claim 1, the multi-robot cloud scheduling and collaboration system comprises: The elevator service group division module is used to acquire building elevator group data and divide the building elevator service groups based on the building elevator group data to obtain building elevator service group data. The path segment structure modeling module is used to acquire robot target path data; it also performs path resource segmentation on the robot target path data and building elevator service group data to obtain split path data. The transfer waiting prediction module is used to obtain the current transfer floor data and estimate the transfer waiting based on the current transfer floor data and the split path data, thus obtaining the transfer waiting estimate data. The conflict-aware scheduling generation module is used to construct a transfer map based on the estimated transfer waiting data to obtain transfer map data; to calculate the graph structure conflict based on the transfer map data to obtain path conflict data; and to generate scheduling data based on the path conflict data to obtain robot scheduling data for multi-robot collaborative scheduling operations.

Images