A hospital logistics operation and maintenance scheduling method based on dynamic passing topology

Abstract

The application is a kind of hospital logistics operation scheduling method based on dynamic traffic topology, relating to the technical field of intelligent scheduling of hospital, comprising the following steps: obtaining scheduling trigger event and determining to-be-processed operation work order and to-be-evaluated operation personnel; forming current traffic topology according to target equipment position, real-time personnel position and current traffic state of hospital; screening reachable work order-personnel combination and determining expected arrival time based on current traffic topology; forming effective candidate matching relationship and matching evaluation in combination with personnel screening constraint, response time limit and personnel scheduling evaluation data; determining personnel allocation processing range according to reachable work order-personnel combination, formation or change of effective candidate matching relationship and matching evaluation, and determining or updating operation personnel allocation result. The application can make actual traffic state change of hospital participate in personnel scheduling calculation, and improve adaptability of scheduling result to change of on-site traffic condition.

Description

Technical Field

[0001] This invention relates to the field of hospital intelligent scheduling technology, and in particular to a hospital logistics operation and maintenance scheduling method based on dynamic access topology. Background Technology

[0002] As the construction of smart hospitals continues to advance, hospital logistics operations and maintenance are gradually shifting from manual ledgers to information-based work order management. Existing platforms typically enable functions such as repair registration, work order generation, personnel dispatch, progress recording, and equipment visualization. Some systems can also incorporate digital twin models or path planning to assist in operations and maintenance. While this approach can improve the level of information technology in operations and maintenance, personnel scheduling is still easily affected by factors such as insufficient utilization of spatial information, changes in personnel status, and resource conflicts among multiple tasks.

[0003] Especially within hospitals, elevators, access control systems, fire doors, corridors, and different functional areas collectively constitute the actual passage space for personnel. While existing intelligent operation and maintenance systems can schedule personnel or adjust routes based on real-time information such as work order status, personnel location, and personnel load, changes in physical access conditions, such as elevator outages, changes in access control status, fire door switching, construction barriers, and temporary area control, still require further solutions. Specifically, it's necessary to address how to convert these physical access status changes into the accessibility attributes and costs of nodes or edges in the hospital's access topology, and accordingly, redetermine the actual reachability relationship and estimated arrival time between personnel and target equipment. Furthermore, it's crucial to determine the specific work orders requiring reassignment based on changes in these actual reachability relationships and candidate matching relationships. If scheduling updates are only based on personnel location, status, or work order status without further impacting the actual reachability relationship and personnel allocation scope due to changes in hospital access conditions, problems may still arise such as originally assigned personnel becoming unreachable, originally estimated arrival times becoming invalid, or unaffected work orders being repeatedly rescheduled.

[0004] How to solve the above-mentioned technical problems is the problem faced by this invention. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a hospital logistics operation and maintenance scheduling method based on dynamic traffic topology, which enables changes in the actual traffic status of the hospital to be included in personnel scheduling calculations, thereby improving the adaptability of scheduling results to changes in on-site traffic conditions.

[0006] The technical solution adopted by this invention to solve its technical problem is as follows: This invention provides a hospital logistics operation and maintenance scheduling method based on dynamic access topology, comprising the following steps: S1, Obtain the scheduling trigger event, and determine the maintenance work order to be processed and the maintenance personnel to be evaluated based on the scheduling trigger event; S2, based on the pre-constructed hospital access topology, the target device location associated with the maintenance work order to be processed and the real-time location of the maintenance personnel to be evaluated are mapped to target device nodes and personnel nodes respectively, and the accessibility attributes and access resistance of nodes or edges in the hospital access topology are determined according to the current access status of the hospital to obtain the current access topology; S3. Based on the current traffic topology, perform reachability screening on each pending maintenance work order and the maintenance personnel to be evaluated to obtain reachable work order-personnel combinations with a passable path, and determine their estimated arrival time. S4. Based on the personnel screening constraints of the corresponding maintenance work order, candidate maintenance personnel are screened from the reachable work order-personnel combination to obtain an effective work order-personnel candidate matching relationship. S5. Compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding operation and maintenance work order, and determine the corresponding matching evaluation based on the comparison results and the scheduling evaluation data of the candidate operation and maintenance personnel. S6. Based on the scheduling trigger event and the formation or change of reachable work orders-personnel combinations, valid candidate matching relationships and matching evaluations, determine the personnel allocation and processing scope corresponding to this scheduling. S7. For the pending maintenance work orders within the scope of personnel allocation, determine or update the maintenance personnel allocation results based on their valid candidate matching relationships and matching evaluations to obtain the scheduling results for this time.

[0007] In the aforementioned scheduling process, the hospital's current accessibility status is not directly used as a parameter in personnel matching evaluation. Instead, it is first used to update the accessibility attributes and resistance of nodes or edges in the hospital's access topology. The updated access topology then redetermines the actual reachability relationship and estimated arrival time between work orders and personnel. This actual reachability relationship further participates in the formation or change of valid work order-person candidate matching relationships, and the formation or change of these reachable work order-person combinations, valid candidate matching relationships, and matching evaluations determine the personnel allocation and processing range. Therefore, the hospital's actual accessibility status, current access topology, actual reachability relationship, valid candidate matching relationship, and personnel allocation and processing range form a hierarchical data transmission relationship.

[0008] Further, step S1 includes the following steps: S11, Obtain scheduling trigger events, wherein the scheduling trigger events include at least one of the following: maintenance work order generation, hospital access status change, maintenance personnel status change, and maintenance work order status change; S12, Based on the affected objects and scope of the scheduling trigger event, determine the maintenance work orders affected by the scheduling trigger event, and designate them as the maintenance work orders to be processed; S13. Based on the pending maintenance work order and the impact range of the scheduling trigger event, determine the maintenance personnel participating in this scheduling evaluation from the maintenance personnel data, and designate them as the maintenance personnel to be evaluated.

[0009] Further, step S2 includes the following steps: S21, Obtain the hospital access topology pre-constructed based on the hospital digital twin spatial model, wherein the hospital access topology includes nodes representing hospital access facilities and access areas, edges connecting the nodes, and connectivity relationships between the nodes; S22, based on the unique device code associated with the maintenance work order to be processed, determine the target device node in the hospital access topology from the device space mapping data; S23, Based on the real-time location data of the maintenance personnel to be evaluated, determine the personnel node of the maintenance personnel to be evaluated in the hospital access topology; S24. Based on the current state of the hospital access facilities and access areas, determine the accessibility attributes and access resistance of the corresponding nodes or edges to obtain the current access topology.

[0010] Further, step S3 includes the following steps: S31, for each combination of maintenance work orders to be processed and maintenance personnel to be evaluated, with the corresponding personnel node as the starting point and the target device node as the ending point, a path search is performed based on the passability attributes and passability resistance of the nodes or edges in the current passability topology to obtain the corresponding path search results. S32, based on the path search results, filter the work order-personnel combinations with passable paths to obtain the reachable work order-personnel combinations, and determine the passable path and path length corresponding to each reachable work order-personnel combination; S33, determine the estimated arrival time based on the travel path corresponding to each reachable work order-personnel combination.

[0011] Further, step S4 includes the following steps: S41, obtain the personnel screening constraints corresponding to each pending maintenance work order. The personnel screening constraints include the minimum job suitability requirements, the personnel load limit, and the job qualification requirements and area access permission requirements determined according to the target equipment type and work area. S42, for each reachable work order-personnel combination, match the job information, current personnel load, work qualifications and regional access permissions of the corresponding maintenance personnel to be evaluated with the personnel screening constraints of the corresponding maintenance work order; S43, filter out reachable work order-personnel combinations that do not meet any of the personnel screening constraints, and determine the maintenance personnel to be evaluated who meet all the personnel screening constraints as candidate maintenance personnel for the corresponding maintenance work orders to be processed, thus obtaining the effective work order-personnel candidate matching relationship.

[0012] Further, step S5 includes the following steps: S51, obtain personnel scheduling evaluation data corresponding to each valid candidate matching relationship, the personnel scheduling evaluation data including job suitability, current personnel load and historical operation and maintenance evaluation; S52, compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding pending maintenance work order. When the estimated arrival time exceeds the response time limit, retain the corresponding valid candidate matching relationship and determine that it has a timeout risk. S53, based on the comparison results of the estimated arrival time and the response time limit and the personnel scheduling evaluation data, a weighted evaluation is performed on each valid candidate matching relationship to obtain the matching evaluation, wherein the matching evaluation of valid candidate matching relationships with timeout risk is reduced.

[0013] Furthermore, in step S6, the scope of personnel allocation and processing is determined according to the following rules: When only the matching evaluation changes, and the corresponding valid candidate matching relationship remains valid and the expected arrival time still meets the response time limit of the corresponding maintenance work order, the maintenance work order that has been assigned and is being executed normally will not be included in the scope of personnel allocation and processing. When the assigned personnel for a single pending maintenance work order become unreachable, the estimated arrival time exceeds the response time limit by a preset range, the order is rejected or the acceptance timeout occurs, or changes in personnel screening constraints cause the original valid candidate matching relationship to become invalid or there is no valid candidate matching relationship, the pending maintenance work order is determined to be within the scope of personnel assignment and processing. When a new high-priority maintenance work order causes competition for personnel resources, changes in hospital access status lead to changes in the effective candidate matching relationship of multiple pending maintenance work orders, or maintenance work orders are generated in batches or the status of maintenance teams changes, the multiple pending maintenance work orders affected will be determined as the scope of personnel allocation and processing.

[0014] In a specific application scenario, when a hospital access status change event indicates that an elevator E1 has changed from an accessible state to a stopped state, step S2 updates the accessible attribute of its associated vertical connection edge based on the access facility identifier and status code corresponding to elevator E1. For the maintenance work order i corresponding to the assigned maintenance personnel j1, if the original access path passes through the elevator E1, step S3 is executed again; when there is no other accessible path from personnel j1 to the target device node after elevator E1 stops operating, the original work order-personnel combination (i, j1) no longer forms a reachable work order-personnel combination, and the corresponding valid candidate matching relationship becomes invalid. For another maintenance personnel j2 to be evaluated, if there is still an accessible path to the target device node in the updated current access topology, the corresponding access path and estimated arrival time are obtained again, and steps S4 and S5 determine whether it forms a valid candidate matching relationship and the corresponding matching evaluation. Step S6 identifies the maintenance work order i that has lost accessibility of the originally assigned personnel j1 by comparing the valid candidate matching relationships before and after the change in access status, and includes the work order in the scope of this personnel allocation process; Step S7 then determines the new maintenance personnel allocation result based on the current valid candidate matching relationships and matching evaluation.

[0015] When the change in the state of elevator E1 simultaneously causes changes in the current access path, reachability, or valid candidate matching relationship of multiple incomplete maintenance work orders, step S6 only includes the multiple maintenance work orders actually affected by the state change in the scope of personnel allocation and processing, and performs multi-work order personnel allocation according to steps S71 to S73; other maintenance work orders that have no access path association with elevator E1 and whose candidate matching relationship has not changed do not participate in this reassignment.

[0016] Furthermore, when the personnel allocation processing scope includes multiple pending maintenance work orders, step S7 includes the following steps: S71, obtain the valid work order-personnel candidate matching relationship and corresponding matching evaluation for each pending maintenance work order within the personnel allocation processing range; S72, Based on the effective work order-personnel candidate matching relationship, under the allocatable load limit of candidate maintenance personnel, determine the feasible personnel allocation combination between multiple pending maintenance work orders and candidate maintenance personnel; S73, determine the sum of matching evaluations of the selected valid candidate matching relationships in each feasible personnel allocation combination, and determine the feasible personnel allocation combination with the largest sum of matching evaluations as the operation and maintenance personnel allocation result.

[0017] Furthermore, when the pending maintenance work orders determined by S1 include high-priority maintenance work orders with task preemption permissions, S1 identifies the low-priority maintenance work order currently being executed by the maintenance personnel to be evaluated as a maintenance work order to be processed. S6 determines the high-priority maintenance work orders and the low-priority maintenance work orders being executed by their valid candidate maintenance personnel as the personnel allocation and processing scope based on the valid candidate matching relationship of the high-priority maintenance work orders obtained through S2 to S5 and the current task status of the candidate maintenance personnel. S7 determines the operation and maintenance personnel allocation result based on the valid candidate matching relationship and matching evaluation of each pending operation and maintenance work order within the personnel allocation processing range. When the operation and maintenance personnel assigned to the high-priority operation and maintenance work order are executing a low-priority operation and maintenance work order, the low-priority operation and maintenance work order corresponding to the operation and maintenance personnel is paused and the current work progress is saved. When the high-priority maintenance work order is completed S1 acquires the corresponding operation and maintenance work order status change event and identifies the suspended low-priority operation and maintenance work orders as operation and maintenance work orders to be processed. After steps S2 to S5 re-obtain the valid candidate matching relationship and matching evaluation of the low-priority maintenance work order based on the current state, S6 determines the suspended low-priority maintenance work order as the scope of personnel allocation and processing; S7 updates the operation and maintenance personnel allocation results based on the current valid candidate matching relationship and matching evaluation. When the original operation and maintenance personnel are in an assignable state and are still valid candidate operation and maintenance personnel, the original personnel allocation is restored first; otherwise, the operation and maintenance personnel allocation results are re-determined.

[0018] Furthermore, when S1 determines the maintenance work order to be processed based on the scheduling trigger event, it performs a status consistency check on the IoT device status data, digital twin model status data, and maintenance work order status data associated with the maintenance work order to be determined as the maintenance work order to be processed. When the IoT device status data, digital twin model status data, and maintenance work order status data are inconsistent or status synchronization fails, and the IoT device status data is unreliable, the corresponding maintenance work order will be marked as status pending verification and will not be determined as a maintenance work order to be processed. S2 to S7 will not be executed for this maintenance work order. When the status data of the digital twin model or the status data of the maintenance work order are inconsistent with the status data of the IoT device or the synchronization is abnormal, but the status data of the IoT device is reliable, the corresponding maintenance work order is determined as a maintenance work order to be processed, and S2 to S7 are continued. When the status of the maintenance work order to be verified is successfully synchronized, the corresponding status version number is aligned and the status conflict is eliminated, the status recovery is used as a new scheduling trigger event. S1 re-determines the maintenance work order as a maintenance work order to be processed, and S2 to S7 are re-executed based on the current status, without using the valid candidate matching relationship and matching evaluation before the status abnormality.

[0019] This invention updates the traversability attributes and resistance of nodes or edges in the hospital's traversal topology based on the hospital's current traversal status, and calculates the actual reachability and estimated arrival time of personnel to target equipment based on the updated current traversal topology. This allows changes in on-site traversal such as elevator shutdowns, access control changes, passage closures, and temporary area management to directly participate in subsequent personnel scheduling calculations, which helps reduce the situation of personnel allocation based on static distances or failed paths.

[0020] This invention determines the scope of personnel allocation based on scheduling trigger events and the formation or changes in reachable work order-personnel combinations, valid candidate matching relationships, and matching evaluations. For work orders where only the evaluation changes and the original personnel still meet the execution conditions, the original allocation is maintained. When a single work order is affected, local processing is performed; when multiple work orders are affected together, multi-work order personnel allocation is performed. This reduces unnecessary overall rescheduling and scheduling disturbances to existing normally executing tasks. Furthermore, by determining the scope of actually affected maintenance work orders based on the changed nodes or edges and the resulting changes in actual reachability and valid candidate matching relationships, it helps reduce the computational processing of repeatedly performing path searches, personnel screening, and personnel allocation solutions for unaffected maintenance work orders.

[0021] This invention first forms a work order-personnel combination based on actual accessibility, then screens personnel by combining minimum job suitability requirements, personnel load limits, job qualifications, and regional access permissions, and further combines estimated arrival time, response time limit, and personnel scheduling evaluation data to form a matching evaluation. This allows personnel spatial accessibility conditions, job qualification conditions, and scheduling evaluation conditions to participate in personnel matching at different levels, which helps to improve the adaptability of effective candidate matching relationships to actual operation and maintenance task requirements.

[0022] When multiple pending maintenance work orders participate in personnel allocation simultaneously, this invention forms a feasible personnel allocation combination under the constraint of the available allocation load of candidate maintenance personnel, and determines the personnel allocation result based on the sum of the matching evaluation of the selected effective candidate matching relationships in the combination. This allows the competition for personnel resources among multiple work orders to be handled in a unified allocation process, which helps to reduce personnel resource conflicts caused by multiple work orders locally selecting the best option.

[0023] This invention performs consistency verification on IoT device status data, digital twin model status data, and maintenance work order status data. It controls whether work orders enter automatic scheduling based on the trusted status of IoT device status data. After the status is restored, it recalculates candidate matching relationships and matching evaluations based on the current status. This helps reduce the participation of abnormal or failed status data in personnel scheduling and avoids unnecessary interruptions to normal maintenance scheduling due to abnormal model or work order status synchronization. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall process of a hospital logistics operation and maintenance scheduling method based on dynamic traffic topology according to the present invention. Figure 2 This is a schematic diagram illustrating the current common topology construction and reachability calculation process of this invention; Figure 3 This is a schematic diagram of the process for screening, matching, and evaluating personnel for this invention. Figure 4 This is a flowchart illustrating the process of determining the scope of personnel allocation and assigning personnel across multiple work orders in this invention. Figure 5 This is a schematic diagram illustrating the process of preempting and restoring high-priority maintenance work orders in this invention. Figure 6 This is a schematic diagram of the state consistency verification and scheduling admission process of the present invention. Detailed Implementation

[0025] To clearly illustrate the technical features of this solution, the following detailed implementation method will be used to explain the solution.

[0026] See Figures 1 to 6 As shown, this embodiment is a hospital logistics operation and maintenance scheduling method based on dynamic access topology. The method can be executed by computing devices in the hospital logistics operation and maintenance platform, and completes a scheduling process by calling hospital digital twin space data, IoT device data, operation and maintenance work order data, and operation and maintenance personnel data. It includes the following steps: To enable subsequent steps to be directly executed by computing devices, in one specific implementation, the hospital logistics operations and maintenance platform stores the data involved in scheduling in a structured manner according to a unified identifier. The digital twin spatial data includes at least a park identifier, building identifier, floor identifier, spatial area identifier, spatial boundary, and spatial coordinates; equipment data includes at least a unique equipment code, equipment type, equipment spatial anchor point, location, and equipment status; operations and maintenance personnel data includes at least a personnel identifier, job type and skill level, current personnel load, set of work qualifications, set of area access permissions, real-time location, and current task status; operations and maintenance work order data includes at least a work order identifier, unique target equipment code, work order type, work order level, current status, and assigned personnel. The above data are linked through unique equipment codes, spatial area identifiers, and personnel identifiers.

[0027] In one specific implementation, the platform sets up a traffic topology node table, a traffic topology edge table, an equipment space mapping table, a personnel capability and permission table, a work order response rule table, and a scheduling evaluation parameter table. The traffic topology node table records node identifiers, node types, floors, areas, and spatial coordinates; the traffic topology edge table records edge identifiers, starting nodes, ending nodes, geometric lengths, traffic direction, associated access facilities, passability attributes, traffic resistance parameters, and fixed waiting times; the equipment space mapping table records the correspondence between unique equipment codes and equipment space anchor points, areas, and traffic topology nodes.

[0028] S1: Obtain the scheduling trigger event, and determine the maintenance work order to be processed and the maintenance personnel to be evaluated based on the scheduling trigger event; the scheduling trigger event is used to initiate a new personnel scheduling process. Step S1 mainly determines the work order object and personnel object that need to participate in the subsequent space calculation and personnel matching. Whether the personnel are actually reachable, whether they meet the personnel requirements of the corresponding work order, and which maintenance personnel they are ultimately assigned to will be further determined in subsequent steps.

[0029] S2, based on the pre-constructed hospital access topology, the target device location associated with the maintenance work order to be processed and the real-time location of the maintenance personnel to be evaluated are mapped to target device nodes and personnel nodes respectively, and the accessibility attributes and access resistance of nodes or edges in the hospital access topology are determined according to the current access status of the hospital to obtain the current access topology; S3. Based on the current traffic topology, perform reachability screening on each pending maintenance work order and the maintenance personnel to be evaluated to obtain reachable work order-personnel combinations with a passable path, and determine their estimated arrival time. S4. Based on the personnel screening constraints of the corresponding maintenance work order, candidate maintenance personnel are screened from the reachable work order-personnel combination to obtain an effective work order-personnel candidate matching relationship. S5. Compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding operation and maintenance work order, and determine the corresponding matching evaluation based on the comparison results and the scheduling evaluation data of the candidate operation and maintenance personnel. S6. Based on the scheduling trigger event and the formation or change of reachable work orders-personnel combinations, valid candidate matching relationships and matching evaluations, determine the personnel allocation and processing scope corresponding to this scheduling. S7. For the pending maintenance work orders within the scope of personnel allocation, determine or update the maintenance personnel allocation results based on their valid candidate matching relationships and matching evaluations to obtain the scheduling results for this time.

[0030] The above steps S1 to S7 constitute a complete operation and maintenance personnel scheduling process. Specifically, step S1 determines the work order objects and personnel objects participating in this scheduling; step S2 maps the target equipment and personnel locations to the hospital access topology and forms the current access topology based on the current actual access status; step S3 forms the actual reachability relationship and estimated arrival time based on the current access topology; step S4 further forms effective candidate matching relationships based on personnel hard constraints; step S5 forms corresponding matching evaluations; step S6 determines the scope of personnel allocation processing based on the formation or changes of the aforementioned reachable work order-personnel combinations, effective candidate matching relationships, and matching evaluations; and step S7 completes the corresponding personnel allocation. When a new scheduling trigger event occurs, the corresponding scheduling processing can be executed again according to the above steps. The following sections provide further explanation of each of the above steps.

[0031] Further, step S1 includes the following steps: S11, Obtain scheduling trigger events, wherein the scheduling trigger events include at least one of the following: maintenance work order generation, hospital access status change, maintenance personnel status change, and maintenance work order status change; S12, Based on the affected objects and scope of the scheduling trigger event, determine the maintenance work orders affected by the scheduling trigger event, and designate them as the maintenance work orders to be processed; S13. Based on the pending maintenance work order and the impact range of the scheduling trigger event, determine the maintenance personnel participating in this scheduling evaluation from the maintenance personnel data, and designate them as the maintenance personnel to be evaluated.

[0032] Specifically, maintenance work order generation events can be triggered by equipment fault alarms, inspection tasks, maintenance tasks, or manually added tasks. For equipment fault triggering scenarios, after an IoT device malfunctions or generates an abnormal alarm, the corresponding device is identified based on the unique device code associated with the alarm data, and a maintenance work order is generated. For inspection and maintenance tasks, corresponding maintenance work orders can be generated based on the inspection cycle, maintenance cycle, equipment runtime, or corresponding maintenance plan.

[0033] The unique device code can consist of device category information, spatial hierarchy information, and device serial number, and is associated with the physical device, digital twin model, and corresponding maintenance work order. The unique device code is further used in the digital twin spatial data to associate the device's three-dimensional spatial anchor point, its location, and its associated access topology nodes, for subsequent step S2 to determine the target device node.

[0034] Hospital access status changes can include elevator outages, changes in access control status, changes in fire door status, construction site fencing, passageway closures, or temporary control of work areas. Maintenance personnel status changes can include personnel completing existing tasks, changes in personnel workload, order rejection, order timeouts, or changes in team status. Maintenance work order status changes can include changes in work order level, work order completion, or task pauses—status changes that can affect current personnel scheduling relationships.

[0035] Newly generated maintenance work orders can be directly treated as pending maintenance work orders. For changes in hospital access status, maintenance personnel status, or maintenance work order status, the affected scope is determined based on the associated region, personnel, and existing work orders, and pending maintenance work orders are obtained from this scope. Subsequently, maintenance personnel who need to participate in this scheduling evaluation are identified from the maintenance personnel data. This maintenance personnel data may include job type information, current personnel load, work qualifications, area access permissions, current task status, real-time location data, and historical maintenance evaluations.

[0036] To specifically determine the impact scope of the scheduling trigger event, the platform can maintain a scheduling snapshot for maintenance work orders in an incomplete state. This snapshot records at least the target device node, assigned personnel, current valid candidate personnel identifiers, the current access path and path edge set corresponding to each valid candidate matching relationship. When the scheduling trigger event is a change in hospital access status, the changed node or edge, its floor, and access area are determined based on the status change object. For events such as reduced access capacity or closure, maintenance work orders whose current access path passes through the changed node or edge are prioritized as affected work orders. For events such as access restoration or reduced access resistance that may create new available paths, incomplete maintenance work orders whose target device node or candidate personnel node is located within the connected area of ​​the changed node or edge are included in the recalculation scope.

[0037] When the scheduling trigger event is a change in the status of maintenance personnel, maintenance work orders already assigned to that personnel or containing that personnel in the current valid candidate relationships are identified as affected work orders. When the scheduling trigger event is a change in the status of a maintenance work order, that work order and any incomplete work orders sharing candidate personnel resources with it are considered as objects to be evaluated. Step S13 then retrieves the personnel who can participate in this evaluation from the maintenance personnel records of the corresponding service area or team, and retains at least the currently assigned personnel and the available personnel that meet the service scope of the target area. Thus, S12 and S13 can form a set of work orders to be processed and a set of personnel to be evaluated that can be directly calculated based on the event objects.

[0038] Step S1 outputs the maintenance work orders to be processed and the maintenance personnel to be evaluated, which are then used as input for spatial node mapping in step S2.

[0039] Further, step S2 includes the following steps: S21, Obtain the hospital access topology pre-constructed based on the hospital digital twin spatial model, wherein the hospital access topology includes nodes representing hospital access facilities and access areas, edges connecting the nodes, and connectivity relationships between the nodes; S22, based on the unique device code associated with the maintenance work order to be processed, determine the target device node in the hospital access topology from the device space mapping data; S23, Based on the real-time location data of the maintenance personnel to be evaluated, determine the personnel node of the maintenance personnel to be evaluated in the hospital access topology; S24. Based on the current state of the hospital access facilities and access areas, determine the accessibility attributes and access resistance of the corresponding nodes or edges to obtain the current access topology.

[0040] The hospital access topology can be pre-built based on the hospital's digital twin spatial model and stored using a weighted directed graph. Nodes can correspond to corridor connections, elevators, stairs, access control systems, fire doors, area entrances, and other locations with access significance; edges represent the access connections between corresponding nodes. Step S2 does not reconstruct the complete hospital access topology; instead, it calls the pre-built topology and updates its relevant attributes based on the current state.

[0041] In a directly implementable mapping method, the hospital's digital twin spatial model consists of a spatial geometry layer, an equipment semantic layer, and an access facility layer. The spatial geometry layer records the unified coordinates of the park, buildings, floors, rooms, corridors, and area boundaries; the equipment semantic layer records the unique codes, equipment types, and spatial anchor points of equipment; the access facility layer records objects such as doors, access control systems, fire doors, staircases, elevators, and passageways, as well as the spatial areas they connect to. This spatial model can be imported from the hospital's existing BIM / IFC model, 3D GIS model, or indoor spatial model and converted to the same hospital spatial coordinate system.

[0042] When constructing the hospital's access topology, nodes are first established at corridor intersections, room or area entrances / exits, access control or fire door locations, elevator lobbies, stairwells, and inter-floor connections. Then, edges are established according to the geometric connections between adjacent accessible spaces in the digital twin spatial model. Within the same floor, horizontal edges are established between adjacent nodes based on corridor, door, or area connectivity. Vertical edges are established between corresponding nodes on different floors for elevators and staircases. For passageways with directional restrictions, only directed edges in the permitted direction are established; for bidirectional passage areas, two edges in opposite directions are established. The geometric length of the edges is calculated based on the spatial coordinates of the corresponding passage segment or polyline path.

[0043] The hospital access topology is constructed or updated when the digital twin spatial model is first released or when the spatial structure is adjusted. During daily scheduling, it is not necessary to repeatedly generate all nodes and edges. Instead, the dynamic attributes of existing nodes or edges are updated according to real-time data such as equipment status, access control status, and construction closure status. In this way, the digital twin spatial model is responsible for providing stable spatial geometric and semantic relationships, while the hospital access topology is responsible for providing nodes, edges, and dynamic access attributes for path search.

[0044] Equipment spatial mapping data can be generated during the digital twin modeling process, associating the unique equipment code with the equipment's 3D spatial anchor point, park, building, floor, room, and its corresponding access topology nodes. Step S22 obtains the above mapping data based on the unique equipment code in the maintenance work order to be processed, thereby obtaining the corresponding target equipment node.

[0045] The real-time location of the maintenance personnel to be evaluated can be obtained through the hospital's existing personnel positioning system. Based on the spatial correspondence between the personnel's real-time location and the hospital's digital twin spatial model, it is mapped to personnel nodes in the hospital's access topology. Thus, device location data and personnel positioning data from different sources are all converted into node objects within the same access topology.

[0046] In one specific implementation, when there are multiple passable candidate topology nodes in the building, floor, and accessible area corresponding to a person's real-time location, the person node v corresponding to person j... j It can be represented as:

[0047] Among them, V f,z p represents the set of passable candidate nodes corresponding to the current floor f and the accessible area z of person j; j p represents the real-time location coordinates of person j; v This represents the spatial coordinates of the candidate node v. The obtained v j This serves as the starting point for path searching in step S31. When the real-time position is located in the middle of a certain passageway, a temporary starting node is formed according to the aforementioned projection method, and this temporary starting node is used as the starting point for path searching in step S31.

[0048] Nodes or edges have drivability attributes and drivability resistance. Drivability attributes indicate whether the corresponding node or edge allows the formation of an actual drivability path under the current conditions. For example, when an elevator is out of service, the corresponding node or edge that relies on the elevator to form a floor connection will be updated to be inaccessible; when a passage is blocked by construction barriers or temporarily closed, the corresponding edge will be updated to be inaccessible; when the status of access control or fire door changes, the relevant drivability attributes will be updated accordingly.

[0049] For nodes or edges that are still passable but have different passage conditions, different passage costs can be expressed through passage resistance. For example, different passage resistances can be set for main patient passages, dedicated medical staff passages, cleaning passages, and passages affected by temporary operations but still passable. When the passage conditions of a certain road segment are reduced but not completely closed, its passage resistance can be increased; if it is completely impassable, the road segment can be excluded from the current valid paths through the passability attribute.

[0050] In one specific implementation, a access status parameter table is established for different types of access facilities, comprising "facility type - status code - passability indicator - resistance coefficient - fixed waiting time". When the status of elevator controllers, access control systems, fire doors, IoT gateways, or construction closures due to manual maintenance changes, the platform queries the parameter table based on the access facility identifier and status code, and writes the passability attribute and resistance coefficient ρ of the corresponding node or edge. e and fixed waiting time τ e A normal and well-connected corridor can allow ρ e =1、τ e =0; Road sections that are still open to traffic but affected by temporary operations, congestion, or traffic control are ordered to ρ e >1; When passage is impossible, the passability attribute is set to negative and the passage does not participate in path search. For passage facilities with waiting processes, such as elevators, τ e The estimated waiting time can be obtained from real-time reports from the equipment or from the most recent statistics. In the absence of real-time waiting data, the average waiting time pre-configured by the facility can be used.

[0051] The traffic status parameter table is pre-formed according to the type of traffic facility and its corresponding status code before scheduling. Each facility type and status code corresponds to a defined passable marker and a corresponding traffic resistance parameter. For traffic facilities with a waiting process, a fixed waiting time is also assigned. During scheduling, the corresponding parameter record is queried based on the current traffic facility identifier and its status code, and the retrieved parameters are written to the corresponding node or edge. The traffic resistance parameter is not changed in reverse based on the current candidate maintenance personnel, matching evaluation, or final personnel allocation results. When no valid parameter record corresponding to the current facility status is found, the corresponding facility status is marked as traffic parameter pending verification. Before the corresponding parameter configuration or verification is completed, automatic path calculation is not performed using the undetermined traffic resistance parameter.

[0052] In one specific implementation, for any currently traversable edge e, its estimated travel time cost c e It can be represented as:

[0053] Among them, l eRepresents the geometric length of edge e; v0 represents the pre-configured indoor reference moving speed of the platform; ρ e τ represents the resistance coefficient obtained based on the current passage state of edge e; e Let c represent the fixed waiting time corresponding to edge e. e The edge cost is used for path search in step S31 and is used to determine the expected arrival time in step S33.

[0054] After step S24 is completed, the current traffic topology reflecting the actual traffic conditions of the hospital is obtained, and it serves as the basis for path search in step S3.

[0055] Further, step S3 includes the following steps: S31, for each combination of maintenance work orders to be processed and maintenance personnel to be evaluated, with the corresponding personnel node as the starting point and the target device node as the ending point, a path search is performed based on the passability attributes and passability resistance of the nodes or edges in the current passability topology to obtain the corresponding path search results. S32, based on the path search results, filter the work order-personnel combinations with passable paths to obtain the reachable work order-personnel combinations, and determine the passable path and path length corresponding to each reachable work order-personnel combination; S33, determine the estimated arrival time based on the travel path corresponding to each reachable work order-personnel combination.

[0056] Specifically, for the same pending maintenance work order, the path search can be performed starting from the personnel node corresponding to each maintenance personnel to be evaluated and ending at the target device node associated with the work order.

[0057] In one implementation, the A-star path search algorithm can be used for path calculation. During path search, currently impassable paths are excluded based on the traversability attributes of nodes or edges, and available traversable paths are determined by combining this with the corresponding traversal resistance. In another implementation, other path search methods that can obtain the actual traversable paths between nodes based on the current topology can also be used.

[0058] When using the A-star algorithm, the current topology is denoted as G=(V,E), the personnel node is denoted as the starting point s, and the target device node is denoted as the ending point t. An open set OPEN and a closed set CLOSED are established. For any node n in OPEN, its total cost f(n) can be expressed as:

[0059] Where g(n) represents the cumulative travel time cost from the starting point s to node n along the searched edges; h(n) represents the heuristically estimated time from node n to the ending point t. To ensure that the heuristic estimate has the same dimensions as the aforementioned edge travel time cost, h(n) can be expressed as:

[0060] Where d(n,t) represents the straight-line distance between node n and endpoint t in the hospital's spatial coordinates; the meaning of v0 is the same as the definition in the aforementioned edge travel time cost. Each time, the node with the smallest f(n) is selected from OPEN for expansion, and only its adjacent and currently travelable edges are considered as candidate edges.

[0061] For the case where a node n reaches its neighboring node n' via edge e, the cumulative travel time cost can be updated as follows:

[0062] Where n' represents a node that is adjacent to node n via edge e and is currently traversable; c e The meaning is the same as the definition in the aforementioned edge travel time cost. If the same node obtains a smaller g value, its parent node is updated synchronously; when the destination t is selected as the node with the smallest current f value, the travel path from the personnel node to the target device node is obtained by backtracking according to the parent node relationship; if OPEN is empty and no destination is found, it is determined that there is no travelable path for the corresponding work order-personnel combination.

[0063] When there is no passable path between a personnel node and the target device node, the work order-person combination is not retained for subsequent personnel screening; when a passable path exists, it is identified as a reachable work order-person combination, and the corresponding actual passable path and path length are recorded. The path length is the length of the passable path formed by the personnel moving along the actual passage space of the hospital.

[0064] The estimated arrival time is further calculated based on the obtained travel path. The estimated arrival time is used as the time parameter for the reachable work order-personnel combination and is compared with the response time limit of the corresponding maintenance work order in step S5.

[0065] For any reachable work order-person combination (i,j) obtained from the search, its travel path is denoted as P. ij The corresponding path length L ij It can be represented as:

[0066] Among them, P ij L represents the travel path from person j to the target device node corresponding to work order i; ij This indicates the path length of the travel route; l eThe meaning is the same as the definition above. Based on the same travel route, its estimated arrival time T ij It can be represented as:

[0067] Among them, T ij This indicates the estimated arrival time of personnel j at the target equipment location corresponding to work order i; c e The meaning of L is the same as the definition above. ij As the path length output of step S32, the resulting T ij The estimated arrival time is output as step S33 and compared with the response time of the corresponding maintenance work order in step S5.

[0068] After the hospital's access status changes, when steps S2 and S3 are executed again, the updated current access topology is used to re-search for paths. If the original work order-person combination no longer has a passable path, the combination loses its reachability; if a path still exists, the corresponding passable path, path length, and estimated arrival time are updated.

[0069] Further, step S4 includes the following steps: S41, obtain the personnel screening constraints corresponding to each pending maintenance work order. The personnel screening constraints include the minimum job suitability requirements, the personnel load limit, and the job qualification requirements and area access permission requirements determined according to the target equipment type and work area. S42, for each reachable work order-personnel combination, match the job information, current personnel load, work qualifications and regional access permissions of the corresponding maintenance personnel to be evaluated with the personnel screening constraints of the corresponding maintenance work order; S43, filter out reachable work order-personnel combinations that do not meet any of the personnel screening constraints, and determine the maintenance personnel to be evaluated who meet all the personnel screening constraints as candidate maintenance personnel for the corresponding maintenance work orders to be processed, thus obtaining the effective work order-personnel candidate matching relationship.

[0070] Personnel screening constraints can be obtained based on the equipment attributes of the target equipment associated with the maintenance work order and the attributes of the work area. For target equipment with specific operational qualification requirements, such as medical gases and high-voltage power distribution, only personnel with the corresponding operational qualifications can meet the corresponding screening conditions; for areas with regional access requirements, further judgment is made based on the regional access permissions of the maintenance personnel.

[0071] In one specific implementation, the platform sets up a personnel screening constraint template table, using at least two of the following as search criteria: target equipment type, work area type, and maintenance work order type. Each constraint template records at least the allowed job code, minimum skill level, personnel workload limit, required set of work qualification codes, and allowed area access permission code. When generating or reading a maintenance work order to be processed, the target equipment type and its location are obtained based on the equipment's unique code, and then the corresponding constraint template is queried to form the personnel screening constraints described in S41.

[0072] When no applicable personnel screening constraint template is found based on the target equipment type, work area type, and maintenance work order type, the undefined personnel screening conditions are not used to generate a valid candidate matching relationship. Instead, the corresponding maintenance work order is marked as a personnel screening parameter pending verification. After the corresponding personnel screening constraint is configured or verified, the status of the corresponding work order is restored as a maintenance work order status change event and the process re-enters step S1.

[0073] The maintenance personnel file records the personnel type code and skill level, the number of currently incomplete work orders or equivalent load value, the set of obtained work qualification codes, and the set of regional access permissions. In a simplified implementation, the current personnel load L... j The personnel load limit L is represented by the number of work orders that the personnel currently has not completed and that are occupying personnel resources. j `max` is represented by the maximum number of work orders that the person can handle simultaneously; when a new work order requires a load unit, only `L`... j +1≤L j Only when the maximum value is reached will the load constraint be satisfied.

[0074] The personnel load limit is a hard constraint that determines whether a person can enter the candidate range. When a person's current load reaches the corresponding limit, that person can no longer form a valid candidate matching relationship with the corresponding work order.

[0075] Each screening constraint is applied to the reachable work order-personnel combination obtained from S3. If any required condition is not met, the combination is eliminated; if all personnel screening constraints are met, the corresponding personnel become candidate maintenance personnel for that work order, and a valid work order-personnel candidate matching relationship is formed.

[0076] In one specific implementation, for any reachable work order-personnel combination (i,j), the four hard constraints—job type, personnel workload, work qualifications, and regional access—are converted into judgment values ​​of 0 or 1, and the comprehensive screening result B is determined. ij It can be represented as:

[0077] Among them, b ij tradeIndicates the job type and skill level judgment value; b ij load Indicates the personnel load judgment value; b ij ual Indicates the job qualification assessment value; b ij access This represents the region access permission judgment value. Each judgment value is set to 1 if the corresponding constraint is met, and 0 if not. When B... ij When B = 1, personnel j is identified as a candidate maintenance personnel for work order i, and a valid work order-personnel candidate matching relationship is formed; when B ij When =0, the corresponding reachable work order-personnel combination is filtered out.

[0078] Among them, the upper limit of personnel workload and the current personnel workload used for matching evaluation belong to different processing levels; the minimum job suitability requirement and the subsequent job suitability degree are also used for candidate qualification judgment and candidate relationship evaluation, respectively.

[0079] Further, step S5 includes the following steps: S51, obtain personnel scheduling evaluation data corresponding to each valid candidate matching relationship, the personnel scheduling evaluation data including job suitability, current personnel load and historical operation and maintenance evaluation; S52, compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding pending maintenance work order. When the estimated arrival time exceeds the response time limit, retain the corresponding valid candidate matching relationship and determine that it has a timeout risk. S53, based on the comparison results of the estimated arrival time and the response time limit and the personnel scheduling evaluation data, a weighted evaluation is performed on each valid candidate matching relationship to obtain the matching evaluation, wherein the matching evaluation of valid candidate matching relationships with timeout risk is reduced.

[0080] Job suitability is used to indicate the relative suitability between candidates who have met the minimum job suitability requirements and specific maintenance work orders; the current personnel load is determined by the maintenance tasks currently undertaken by the personnel; historical maintenance evaluation can be determined based on maintenance data such as satisfaction, rework status, and handling status of previous work orders.

[0081] The response time limit can be obtained based on the fault level or work order level corresponding to the target equipment. For valid candidate matching relationships whose expected arrival time does not exceed the response time limit, the matching evaluation is performed based on the corresponding personnel scheduling evaluation data; for valid candidate matching relationships whose expected arrival time exceeds the response time limit, the candidate qualification of the personnel is not directly cancelled, but they are identified as having timeout risk, and the corresponding matching evaluation is reduced in step S53.

[0082] Specifically, the platform can set up a work order response rule table, which should include at least the work order level and response time limit R. i Timeout readjustment amplitude Δ i And the task preemption permission flag. For maintenance work orders generated from IoT device fault alarms, the fault level is first obtained by querying the fault level mapping table according to the device type and alarm / fault code. Then, the work order generation module writes the fault level into the final work order level according to the fault level and queries the work order response rule table to assign the corresponding R. i Write the operation and maintenance work order; for work orders created by inspection, maintenance, or manually, the work order level is determined by the corresponding task template or work order creation interface, and the R is obtained according to the same work order response rule table. i Step S52 directly reads the response time limit R already written in the pending maintenance work order. i Compare them.

[0083] When T ij ≤R i When T, the corresponding candidate relationship does not have a timeout; when T ij >R i When this occurs, it is marked as a valid candidate match with a risk of timeout. In one specific implementation, the determination condition "the expected arrival time exceeds the response time limit by a preset range" in step S6 can be expressed as:

[0084] Where, Δ i This indicates the timeout readjustment range obtained from the work order response rule table based on the work order level of work order i; T ij and R i The meaning is the same as the definition above. When the above conditions are met, the corresponding work order can be included in the personnel allocation and processing scope according to the single work order affected rules in step S6.

[0085] In one implementation, different personnel scheduling evaluation data can be normalized and then weighted for evaluation, with the sum of the normalized evaluation weights being 1. The urgency level of the work order is used to determine the scheduling rules corresponding to the work order, including response time limit, personnel screening scope, evaluation parameters, and whether task preemption authority is available, and is not considered as ordinary personnel evaluation data alongside candidate employee type suitability, current personnel load, and historical operation and maintenance evaluations.

[0086] In one specific implementation, all three evaluation dimensions in step S51 are converted to the [0,1] interval. Specifically, the job suitability A of candidate j for work order i is... ij It can be represented as:

[0087] Among them, s ijThis represents the original fit score of personnel j for work order i, obtained from the skill matrix query based on "equipment type-job type-skill level"; s max This represents the maximum allowable fit score for the skill matrix. Current personnel workload assessment F j It can be represented as:

[0088] Among them, L j and L j max The meaning is the same as the definition of personnel load constraint in step S4. Historical operation and maintenance evaluation H j It can be represented as:

[0089] Among them, S j U represents the normalized value obtained by dividing the historical work order satisfaction average of employee j by the corresponding full score; j This indicates the proportion of rework orders to completed orders within the statistical period; O j This represents the proportion of work orders completed within the specified completion timeframe out of the total number of completed work orders. If a particular employee does not have sufficient historical work orders, the average historical maintenance evaluation value for the corresponding job type or work group can be used as H. j The initial value of A. ij F j and H j As input for the weighted evaluation in step S53.

[0090] The scheduling evaluation parameter table records the job type adaptation weight w1, load weight w2, historical evaluation weight w3, and timeout penalty coefficient λ according to the job order type or job order level, and w1+w2+w3=1. After reading the parameter record of the corresponding job order in step S53, A... ij F j and H j The weights are respectively used as the normalized personnel scheduling evaluation data for weighting; if a weight for a certain work order type is not configured separately, a set of default weights preset by the platform can be used, as long as the sum of the three weights is 1.

[0091] The skill matrix and scheduling evaluation parameter table are pre-configured before scheduling. During step S51, the corresponding skill matching records are queried based on the target equipment type and the job type and skill level of the candidate maintenance personnel for the pending maintenance work order, to obtain job matching evaluation data. Under the same skill matrix version, the same combination of "target equipment type—personnel job type—skill level" corresponds to the determined matching records. w1, w2, w3, and λ read the corresponding preset parameter records according to the work order type or work order level, and do not change the corresponding evaluation parameters in reverse based on the current number of candidate personnel, the current matching evaluation result, or the final personnel allocation result during a single scheduling process.

[0092] In one specific implementation, after normalizing the personnel scheduling evaluation data participating in the matching evaluation, the basic matching evaluation Q for the effective candidate matching relationship formed between work order i and candidate j is determined. ij 0 It can be represented as:

[0093] Where, q ijk This represents the k-th normalized personnel scheduling evaluation data corresponding to the valid candidate matching relationship; w k q represents the evaluation weight corresponding to the k-th personnel scheduling evaluation data; m represents the number of personnel scheduling evaluation data participating in the matching evaluation, and the sum of all evaluation weights is 1. When using the aforementioned three evaluation dimensions, q ij1 q ij2 q ij3 Take A respectively ij F j and H j For valid candidate matches whose expected arrival time exceeds the response time limit, their relative timeout level D is... ij It can be represented as:

[0094] Among them, T ij and R i The meaning of R is the same as the definition above, and R i >0;D ij D represents the relative timeout level of the valid candidate matches. ij =0 indicates that there is no timeout, D ij A value greater than 0 indicates a timeout. Based on the basic matching evaluation and the relative timeout level, the final matching evaluation Q is... ij It can be represented as:

[0095] Where λ represents the timeout penalty coefficient greater than 0 obtained from the scheduling evaluation parameter table; Q ij0 and D ij The meaning of is the same as the definition above. D ij It increases with the timeout period, therefore Q ij It decreases as the timeout level increases; when T ij ≤R i At that time, D ij =0 and Q ij =Q ij 0 The obtained Q ij This serves as the matching evaluation input for determining the scope of personnel allocation in step S6 and for determining the personnel allocation results in step S7.

[0096] Step S5 outputs the matching evaluation corresponding to each valid candidate matching relationship, and uses the valid candidate matching relationship and the matching evaluation together in steps S6 and S7.

[0097] Furthermore, in step S6, the scope of personnel allocation and processing is determined according to the following rules: When only the matching evaluation changes, and the corresponding valid candidate matching relationship remains valid and the expected arrival time still meets the response time limit of the corresponding maintenance work order, the maintenance work order that has been assigned and is being executed normally will not be included in the scope of personnel allocation and processing. When the assigned personnel for a single pending maintenance work order become unreachable, the estimated arrival time exceeds the response time limit by a preset range, the order is rejected or the acceptance timeout occurs, or changes in personnel screening constraints cause the original valid candidate matching relationship to become invalid or there is no valid candidate matching relationship, the pending maintenance work order is determined to be within the scope of personnel assignment and processing. When a new high-priority maintenance work order causes competition for personnel resources, changes in hospital access status lead to changes in the effective candidate matching relationship of multiple pending maintenance work orders, or maintenance work orders are generated in batches or the status of maintenance teams changes, the multiple pending maintenance work orders affected will be determined as the scope of personnel allocation and processing.

[0098] Specifically, step S6 mainly determines which work orders need to re-determine personnel in this scheduling, rather than directly selecting the final personnel in this step.

[0099] When changes only lead to adjustments in personnel matching evaluations, while existing personnel remain valid candidates and can meet response deadlines, the original personnel allocation results are retained.

[0100] When a change only affects a single pending maintenance work order, such as when an assigned person loses accessibility due to the latest topology change, a person rejects a work order, or a change in personnel screening constraints causes the original candidate relationship to become invalid, only the corresponding single work order will be included in the personnel assignment processing scope.

[0101] When a new high-priority maintenance work order is added or a large-scale change in hospital access status further affects the shared personnel resources among multiple work orders, the affected multiple work orders will be included in the scope of personnel allocation and processing.

[0102] In its implementation, the platform saves a snapshot of the current valid candidate matching relationships after each S5 process, including at least the work order identifier, candidate personnel identifier, reachability flag, set of path edges, estimated arrival time, and matching evaluation. When a new scheduling trigger event occurs, S2 through S5 are re-executed for the affected objects, and the new snapshot is compared with the original snapshot according to the work order identifier and personnel identifier: if the original candidate relationship does not exist but the new candidate relationship does, it is recorded as a candidate relationship formed; if the original candidate relationship exists but the new candidate relationship does not, it is recorded as a candidate relationship invalidated; if both exist but the estimated arrival time or matching evaluation changes, it is recorded as a candidate relationship updated.

[0103] For hospital access status change events, the recalculation scope can be narrowed first by utilizing the floor and connected regions where the changed node / edge is located. Then, a set of work order identifiers that are actually affected can be formed based on the comparison results of the old and new snapshots. If this set contains only one work order, single work order reassignment is performed; if it contains multiple work orders and there are common candidate personnel resources, proceed to S71 to S73; if the existing assigned personnel are still valid candidates and T ij ≤R i If the original allocation is not maintained, then the allocation will be maintained according to the first rule of S6. Therefore, the scope of personnel allocation processing has a calculable data source.

[0104] The preset range is the timeout readjustment range Δ that matches the corresponding maintenance work order in the aforementioned work order response rule table. i Used with the expected arrival time T ij and response time limit R i Together, we determine whether the corresponding maintenance work order meets the conditions for reassigning personnel.

[0105] Furthermore, when the personnel allocation processing scope includes multiple pending maintenance work orders, step S7 includes the following steps: S71, obtain the valid work order-personnel candidate matching relationship and corresponding matching evaluation for each pending maintenance work order within the personnel allocation processing range; S72, Based on the effective work order-personnel candidate matching relationship, under the allocatable load limit of candidate maintenance personnel, determine the feasible personnel allocation combination between multiple pending maintenance work orders and candidate maintenance personnel; S73, determine the sum of matching evaluations of the selected valid candidate matching relationships in each feasible personnel allocation combination, and determine the feasible personnel allocation combination with the largest sum of matching evaluations as the operation and maintenance personnel allocation result.

[0106] When the scope of personnel assignment only includes one maintenance work order, the corresponding maintenance personnel can be determined directly based on the current valid candidate matching relationship and matching evaluation of the work order.

[0107] When multiple work orders are involved, all current valid candidate relationships for each work order are first obtained. A feasible personnel allocation combination consists of several valid work order-person candidate matching relationships, while simultaneously satisfying the allocatable load limits of the corresponding personnel. If a combination would cause candidate personnel to exceed their current allocatable load, that combination will not be determined as a feasible personnel allocation combination.

[0108] For each feasible personnel allocation combination, obtain the matching evaluation corresponding to the actual selected work order-personnel candidate matching relationship, form a total matching evaluation, compare different feasible personnel allocation combinations, and finally determine the corresponding operation and maintenance personnel allocation result.

[0109] In one implementation, a calculation method based on the weighted bipartite graph matching concept can be used to obtain the personnel allocation results. In another implementation, other optimized solution methods that can obtain the corresponding personnel allocation results under the constraints of effective candidate relationships, matching evaluation, and allocable load can also be used.

[0110] As a directly implementable multi-work-order personnel allocation method, the set of pending maintenance work orders with at least one valid candidate personnel within the personnel allocation processing scope is denoted as I. + The set of candidate maintenance personnel is denoted as J, and based on the valid work order-personnel candidate matching relationship obtained in step S71, a corresponding candidate personnel set J is formed for each work order i. i ; All satisfying j∈J i Let M be the set of valid candidate matches consisting of the work order-person pair (i,j). Establish a 0-1 decision variable x only for (i,j)∈M. ij The objective function, which aims to maximize the sum of matching evaluations for the selected valid candidate matches, can be expressed as:

[0111] Where, x ij =1 indicates that work order i is assigned to candidate j, x ij =0 indicates that the allocation will not be performed; Q ij This is the final matching evaluation obtained in step S53. For each pending maintenance work order with valid candidates, its unique assignment constraint can be expressed as:

[0112] Among them, the aforementioned unique allocation constraint applies to I. +Each work order i in the list applies. For candidate j, the set of pending maintenance work orders that can form a valid candidate match with that candidate is denoted as I. j Its assignable load constraint can be expressed as:

[0113] Where, d i L represents the personnel workload unit occupied by work order i; j and L j max The meaning is the same as the definition of the personnel load constraint mentioned above, and the load constraint applies to each person j in the candidate set J. The value constraint of the decision variable can be expressed as:

[0114] The decision variables are established only for valid candidate matching relationships; therefore, work order-personnel combinations without valid candidate matching relationships are not included in the solution. The objective function and constraints described above can be solved using a conventional 0-1 integer programming solver, branch and bound method, or an equivalent capacity-constrained weighted matching algorithm. Work orders without valid candidate personnel are not included in the solution. + The process remains pending assignment and a message indicating no valid candidates are generated. When the personnel assignment processing scope only includes one work order, a matching evaluation Q can be directly selected. ij The candidate with the highest matching score; if the highest matching scores are the same, the candidate with the estimated arrival time T is selected first. ij Smaller candidates.

[0115] In one specific implementation, for any feasible personnel allocation combination C Its combination matching evaluation Q(C) This can be represented as:

[0116] Among them, C Q represents the r-th feasible personnel allocation combination; ij The meaning is the same as the definition in step S53. Determine the Q(C) corresponding to each feasible personnel allocation combination. The feasible personnel allocation combination with the largest sum of matching evaluations is determined as the operation and maintenance personnel allocation result. This combination evaluation has the same evaluation meaning as the aforementioned 0-1 integer programming objective function.

[0117] Furthermore, when the pending maintenance work orders determined by S1 include high-priority maintenance work orders with task preemption permissions, S1 will also determine the low-priority maintenance work orders currently being executed by the maintenance personnel to be evaluated as pending maintenance work orders. S6 determines the high-priority maintenance work orders and the low-priority maintenance work orders being executed by their valid candidate maintenance personnel as the personnel allocation and processing scope based on the valid candidate matching relationship of the high-priority maintenance work orders obtained through S2 to S5 and the current task status of the candidate maintenance personnel. S7 determines the operation and maintenance personnel allocation result based on the valid candidate matching relationship and matching evaluation of each pending operation and maintenance work order within the personnel allocation processing range. When the operation and maintenance personnel assigned to the high-priority operation and maintenance work order are executing a low-priority operation and maintenance work order, the low-priority operation and maintenance work order corresponding to the operation and maintenance personnel is paused and the current work progress is saved. When the high-priority maintenance work order is completed, S1 obtains the corresponding maintenance work order status change event and determines the suspended low-priority maintenance work order as a maintenance work order to be processed. After steps S2 to S5 re-obtain the valid candidate matching relationship and matching evaluation of the low-priority maintenance work order based on the current state, S6 determines the suspended low-priority maintenance work order as the scope of personnel allocation and processing; S7 updates the operation and maintenance personnel allocation results based on the current valid candidate matching relationship and matching evaluation. When the original operation and maintenance personnel are in an assignable state and are still valid candidate operation and maintenance personnel, the original personnel allocation is restored first; otherwise, the operation and maintenance personnel allocation results are re-determined.

[0118] When a high-priority maintenance work order is generated, all low-priority tasks are not immediately suspended. Instead, the low-priority work orders that are actually competing for resources are first identified based on the current valid candidate matching relationship formed by S2 to S5. Only when the personnel who are finally assigned to the high-priority work order are executing the corresponding low-priority work order will the low-priority work order be suspended and the current work progress be saved.

[0119] After a high-priority work order is completed, a new candidate matching relationship and matching evaluation are formed based on the current traffic topology, personnel location, personnel status, and work order constraints. If the original personnel are currently in an assignable state and still meet the valid candidate conditions, the original personnel can be restored; otherwise, personnel reassignment is performed.

[0120] Whether a work order has task preemption permission can be determined by the task preemption permission flag in the aforementioned work order response rule table. When pausing a low-priority maintenance work order, the platform must at least save the current work step identifier, the set of completed steps, the uploaded work records, the current progress, and the pause timestamp for that work order, and write the work order status as paused. After a high-priority work order is completed, S2 to S5 are re-executed. If the original personnel still meet the current valid candidate conditions and are in an assignable state, the saved work step identifier is used as the recovery starting point to continue the original work order; otherwise, only the completed work records are retained, and the remaining tasks are transferred according to the new personnel allocation results.

[0121] Furthermore, when S1 determines the maintenance work order to be processed based on the scheduling trigger event, it performs a status consistency check on the IoT device status data, digital twin model status data, and maintenance work order status data associated with the maintenance work order to be determined as the maintenance work order to be processed. When the IoT device status data, digital twin model status data, and maintenance work order status data are inconsistent or status synchronization fails, and the IoT device status data is unreliable, the corresponding maintenance work order will be marked as status pending verification and will not be determined as a maintenance work order to be processed. S2 to S7 will not be executed for this maintenance work order. When the status data of the digital twin model or the status data of the maintenance work order are inconsistent with the status data of the IoT device or the synchronization is abnormal, but the status data of the IoT device is reliable, the corresponding maintenance work order is determined as a maintenance work order to be processed, and S2 to S7 are continued. When the status of the maintenance work order to be verified is successfully synchronized, the corresponding status version number is aligned and the status conflict is eliminated, the status recovery is used as a new scheduling trigger event. S1 re-determines the maintenance work order as a maintenance work order to be processed, and S2 to S7 are re-executed based on the current status, without using the valid candidate matching relationship and matching evaluation before the status abnormality.

[0122] The three types of status data can be associated using the device's unique code. When the status of a device, digital twin model, and maintenance work order changes, the corresponding status version number and timestamp can be recorded. During status synchronization, the corresponding updated data can be determined according to the version and time information. If the model status or work order status synchronization fails, a status synchronization retry can be performed.

[0123] The reliability of the IoT device status data can be determined based on the online status of the corresponding IoT device and the abnormal status of the collected data. For example, when an IoT device goes offline or the collected data shows an abnormal jump, the corresponding IoT device status data can be determined as unreliable. When the IoT device status data is unreliable, and there is a conflict among the three types of statuses or a status synchronization failure, the corresponding maintenance work order is set to a pending verification status and stops from entering subsequent automatic personnel scheduling. If the IoT device status data is reliable, and only the digital twin model status or the maintenance work order status has a synchronization anomaly, the corresponding maintenance work order can still continue to enter S2 to S7, and synchronization processing is performed on the abnormal status data.

[0124] In one specific implementation, IoT device status data, digital twin model status data, and maintenance work order status data are all stored using a structure of "device unique code + status code + status version number + timestamp". During status consistency verification, the three types of data are first associated according to the device unique code; if the status codes are consistent and the version numbers are the same, or if the status codes are consistent and the timestamps of all three types of data are within a preset synchronization time window, then the status is determined to be consistent. If the version number of the model status or work order status lags behind the trusted IoT device status, the newer version corresponding to the IoT device status is used as the synchronization source to update the model status or work order status.

[0125] The reliability of IoT device status data can be determined according to the data validity rules corresponding to the device type. These rules include at least one or more of the following: device online status, data freshness threshold, legal range of status values, or allowed state transition set. For example, if the device is online, the latest data timestamp is no more than the freshness threshold for the corresponding device type, and the status value belongs to the allowed state set for that device, the IoT device status data is considered reliable. If the device is offline, the data has not been updated for a long time, the status value exceeds the legal range, or a state transition that is not allowed by the device's state machine occurs, the data is considered unreliable. The threshold and allowed state set are pre-stored in a device type parameter table.

[0126] For work orders whose status is pending verification, the platform can re-execute consistency verification according to IoT data update events or preset retry cycles; only after the status versions are re-aligned and status conflicts are eliminated will the status be restored to form a new scheduling trigger event and S2 to S7 be re-executed.

[0127] After successful status synchronization, re-alignment of corresponding status version numbers, and elimination of status conflicts, the corresponding work orders are re-included in the new scheduling process. The current traffic topology, current personnel location, personnel load, and other current data are used to re-form candidate matching relationships, matching evaluations, and personnel allocation results.

[0128] After obtaining the scheduling result in step S7, the corresponding maintenance work order can be sent to the mobile terminal used by the corresponding maintenance personnel. The maintenance personnel will complete the repair, inspection, or maintenance work according to the work order and upload the corresponding work photos, maintenance data, and consumable records.

[0129] In one implementation, after on-site personnel submit completion information, the maintenance work order enters a pending review state. Completion verification is performed based on the uploaded on-site data and the real-time operating status of the corresponding equipment. Upon successful verification, the status of the corresponding maintenance work order and the digital twin model are updated, and the relevant maintenance data is saved to the historical maintenance record of the corresponding equipment.

[0130] Data such as satisfaction, repair status, handling time, and work order completion status generated in historical operation and maintenance records can be used to form or update the historical operation and maintenance evaluation of operation and maintenance personnel, and can be reused as personnel scheduling evaluation data in subsequent execution steps S51.

[0131] According to the specific implementation method described above, a single scheduling process can be executed directly according to the following data chain: the work order to be processed and the personnel to be evaluated are determined by the scheduling trigger event; the target equipment node and personnel node are obtained from the unique equipment code and the real-time location of the personnel; the passability attribute and ρ in the current passability topology are updated by the passability facility status parameter table. e and τ e Using the A-star algorithm and the edge travel time cost c e Get P ij and T ij ;B is obtained item by item using the personnel screening constraint template. ij And form valid candidate relations; read R i and the calculation of scheduling evaluation parameters Q ij The scope of S6 personnel allocation is determined based on snapshots of the old and new candidate relationships; finally, the maximum Q value of each work order is used. ij The personnel allocation results are obtained by selecting or using 0-1 integer programming for multiple work orders. Each of the above intermediate quantities has a clearly defined data source, calculation method, and output purpose.

[0132] The above embodiments have provided a concrete, directly executable implementation method for constructing and dynamically updating the hospital access topology, mapping digital twin spatial data to topology nodes, determining access resistance and estimated arrival time, A-star path search, obtaining and matching personnel screening constraints item by item, obtaining response time limits, configuring and calculating the weights of matching evaluation, allocating personnel across multiple work orders, and verifying state consistency. For conventional software implementations that do not alter the core data processing relationships, such as database reading and writing, interface communication, message subscription, and mobile terminal display, conventional techniques in the field can be used. Those skilled in the art can implement this using equivalent data structures or programs without changing the aforementioned input, processing, and output relationships.

Claims

1. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology, characterized in that, Includes the following steps: S1, Obtain the scheduling trigger event, and determine the maintenance work order to be processed and the maintenance personnel to be evaluated based on the scheduling trigger event; S2, based on the pre-constructed hospital access topology, the target device location associated with the maintenance work order to be processed and the real-time location of the maintenance personnel to be evaluated are mapped to target device nodes and personnel nodes respectively, and the accessibility attributes and access resistance of nodes or edges in the hospital access topology are determined according to the current access status of the hospital to obtain the current access topology; S3. Based on the current traffic topology, perform reachability screening on each pending maintenance work order and the maintenance personnel to be evaluated to obtain reachable work order-personnel combinations with a passable path, and determine their estimated arrival time. S4. Based on the personnel screening constraints of the corresponding maintenance work order, candidate maintenance personnel are screened from the reachable work order-personnel combination to obtain an effective work order-personnel candidate matching relationship. S5. Compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding operation and maintenance work order, and determine the corresponding matching evaluation based on the comparison results and the scheduling evaluation data of the candidate operation and maintenance personnel. S6. Based on the scheduling trigger event and the formation or change of reachable work orders-personnel combinations, valid candidate matching relationships and matching evaluations, determine the personnel allocation and processing scope corresponding to this scheduling. S7. For the pending maintenance work orders within the scope of personnel allocation, determine or update the maintenance personnel allocation results based on their valid candidate matching relationships and matching evaluations to obtain the scheduling results for this time.

2. The hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, S1 includes the following steps: S11, Obtain scheduling trigger events, wherein the scheduling trigger events include at least one of the following: maintenance work order generation, hospital access status change, maintenance personnel status change, and maintenance work order status change; S12, Based on the affected objects and scope of the scheduling trigger event, determine the maintenance work orders affected by the scheduling trigger event, and designate them as the maintenance work orders to be processed; S13. Based on the pending maintenance work order and the impact range of the scheduling trigger event, determine the maintenance personnel participating in this scheduling evaluation from the maintenance personnel data, and designate them as the maintenance personnel to be evaluated.

3. The hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, S2 includes the following steps: S21, Obtain the hospital access topology pre-constructed based on the hospital digital twin spatial model, wherein the hospital access topology includes nodes representing hospital access facilities and access areas, edges connecting the nodes, and connectivity relationships between the nodes; S22, based on the unique device code associated with the maintenance work order to be processed, determine the target device node in the hospital access topology from the device space mapping data; S23, Based on the real-time location data of the maintenance personnel to be evaluated, determine the personnel node of the maintenance personnel to be evaluated in the hospital access topology; S24. Based on the current state of the hospital access facilities and access areas, determine the accessibility attributes and access resistance of the corresponding nodes or edges to obtain the current access topology.

4. The hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, S3 includes the following steps: S31, for each combination of maintenance work orders to be processed and maintenance personnel to be evaluated, with the corresponding personnel node as the starting point and the target device node as the ending point, a path search is performed based on the passability attributes and passability resistance of the nodes or edges in the current passability topology to obtain the corresponding path search results. S32, based on the path search results, filter the work order-personnel combinations with passable paths to obtain the reachable work order-personnel combinations, and determine the passable path and path length corresponding to each reachable work order-personnel combination; S33, determine the estimated arrival time based on the travel path corresponding to each reachable work order-personnel combination.

5. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, S4 includes the following steps: S41, obtain the personnel screening constraints corresponding to each pending maintenance work order. The personnel screening constraints include the minimum job suitability requirements, the personnel load limit, and the job qualification requirements and area access permission requirements determined according to the target equipment type and work area. S42, for each reachable work order-personnel combination, match the job information, current personnel load, work qualifications and regional access permissions of the corresponding maintenance personnel to be evaluated with the personnel screening constraints of the corresponding maintenance work order; S43, filter out reachable work order-personnel combinations that do not meet any of the personnel screening constraints, and determine the maintenance personnel to be evaluated who meet all the personnel screening constraints as candidate maintenance personnel for the corresponding maintenance work orders to be processed, thus obtaining the effective work order-personnel candidate matching relationship.

6. The hospital logistics operation and maintenance scheduling method based on dynamic access topology as described in claim 1, characterized in that, S5 includes the following steps: S51, obtain personnel scheduling evaluation data corresponding to each valid candidate matching relationship, the personnel scheduling evaluation data including job suitability, current personnel load and historical operation and maintenance evaluation; S52, compare the estimated arrival time of each valid candidate matching relationship with the response time limit of the corresponding pending maintenance work order. When the estimated arrival time exceeds the response time limit, retain the corresponding valid candidate matching relationship and determine that it has a timeout risk. S53, based on the comparison results of the estimated arrival time and the response time limit and the personnel scheduling evaluation data, a weighted evaluation is performed on each valid candidate matching relationship to obtain the matching evaluation, wherein the matching evaluation of valid candidate matching relationships with timeout risk is reduced.

7. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, In step S6, the scope of personnel allocation and processing is determined according to the following rules: When only the matching evaluation changes, and the corresponding valid candidate matching relationship remains valid and the expected arrival time still meets the response time limit of the corresponding maintenance work order, the maintenance work order that has been assigned and is being executed normally will not be included in the scope of personnel allocation and processing. When the assigned personnel for a single pending maintenance work order become unreachable, the estimated arrival time exceeds the response time limit by a preset range, the order is rejected or the acceptance timeout occurs, or changes in personnel screening constraints cause the original valid candidate matching relationship to become invalid or there is no valid candidate matching relationship, the pending maintenance work order is determined to be within the scope of personnel assignment and processing. When a new high-priority maintenance work order causes competition for personnel resources, changes in hospital access status lead to changes in the effective candidate matching relationship of multiple pending maintenance work orders, or maintenance work orders are generated in batches or the status of maintenance teams changes, the multiple pending maintenance work orders affected will be determined as the scope of personnel allocation and processing.

8. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, When the personnel allocation processing scope includes multiple pending maintenance work orders, step S7 includes the following steps: S71, obtain the valid work order-personnel candidate matching relationship and corresponding matching evaluation for each pending maintenance work order within the personnel allocation processing range; S72, Based on the effective work order-personnel candidate matching relationship, under the allocatable load limit of candidate maintenance personnel, determine the feasible personnel allocation combination between multiple pending maintenance work orders and candidate maintenance personnel; S73, determine the sum of matching evaluations of the selected valid candidate matching relationships in each feasible personnel allocation combination, and determine the feasible personnel allocation combination with the largest sum of matching evaluations as the operation and maintenance personnel allocation result.

9. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, When the pending maintenance work orders determined by S1 include high-priority maintenance work orders with task preemption permissions, S1 identifies the low-priority maintenance work order currently being executed by the maintenance personnel to be evaluated as a maintenance work order to be processed. S6 determines the high-priority maintenance work orders and the low-priority maintenance work orders being executed by their valid candidate maintenance personnel as the personnel allocation and processing scope based on the valid candidate matching relationship of the high-priority maintenance work orders obtained through S2 to S5 and the current task status of the candidate maintenance personnel. S7 determines the operation and maintenance personnel allocation result based on the valid candidate matching relationship and matching evaluation of each pending operation and maintenance work order within the personnel allocation processing range. When the operation and maintenance personnel assigned to the high-priority operation and maintenance work order are executing a low-priority operation and maintenance work order, the low-priority operation and maintenance work order corresponding to the operation and maintenance personnel is paused and the current work progress is saved. When the high-priority maintenance work order is completed S1 acquires the corresponding operation and maintenance work order status change event and identifies the suspended low-priority operation and maintenance work orders as operation and maintenance work orders to be processed. After steps S2 to S5 re-obtain the valid candidate matching relationship and matching evaluation of the low-priority maintenance work order based on the current state, S6 determines the suspended low-priority maintenance work order as the scope of personnel allocation and processing; S7 updates the operation and maintenance personnel allocation results based on the current valid candidate matching relationship and matching evaluation. When the original operation and maintenance personnel are in an assignable state and are still valid candidate operation and maintenance personnel, the original personnel allocation is restored first; otherwise, the operation and maintenance personnel allocation results are re-determined.

10. A hospital logistics operation and maintenance scheduling method based on dynamic traffic topology as described in claim 1, characterized in that, When S1 determines the maintenance work order to be processed based on the scheduling trigger event, it performs a status consistency check on the IoT device status data, digital twin model status data, and maintenance work order status data associated with the maintenance work order to be determined as the maintenance work order to be processed. When the IoT device status data, digital twin model status data, and maintenance work order status data are inconsistent or status synchronization fails, and the IoT device status data is unreliable, the corresponding maintenance work order will be marked as status pending verification and will not be determined as a maintenance work order to be processed. S2 to S7 will not be executed for this maintenance work order. When the status data of the digital twin model or the status data of the maintenance work order are inconsistent with the status data of the IoT device or the synchronization is abnormal, but the status data of the IoT device is reliable, the corresponding maintenance work order is determined as a maintenance work order to be processed, and S2 to S7 are continued. When the status of the maintenance work order to be verified is successfully synchronized, the corresponding status version number is aligned and the status conflict is eliminated, the status recovery is used as a new scheduling trigger event. S1 re-determines the maintenance work order as a maintenance work order to be processed, and S2 to S7 are re-executed based on the current status, without using the valid candidate matching relationship and matching evaluation before the status abnormality.

Images