Intelligent distribution system and method for medical consumables

Abstract

The invention relates to the technical field of medical consumable delivery, in particular to an intelligent medical consumable delivery system and method. The method comprises the following steps: acquiring a consumable demand information set, and generating a distribution task queue according to the consumable demand information set; acquiring a real-time equipment node state information set, and based on the real-time equipment node state information set, determining a target medicine taking node information set, an initial distribution node information set and a handover distribution node information set according to the distribution task queue; and according to the target medicine taking node information set, the initial distribution node information set and the handover distribution node information set, performing dynamic distribution task combination, determining and executing a consumable distribution plan, and determining and outputting a consumable distribution report. According to the method and the device, the balance between the medical consumable delivery efficiency and the delivery flexibility is realized, and the medical consumable demand forms in different medical scenes are effectively covered.

Description

Technical Field

[0001] The present application relates to the technical field of medical consumables distribution, and in particular to an intelligent distribution system and method for medical consumables. Background Art

[0002] Medical consumables refer to consumable equipment or items that are used once or repeatedly replaced in a short period of time during medical activities such as diagnosis, treatment, and nursing in medical institutions. They are characterized by high frequency of use, large consumption, and the need for strict management. Automated distribution technology for medical consumables can effectively reduce hospital manpower and warehousing costs, reduce manual operation errors and material loss, and is an important technology to promote refined hospital operations and realize smart medical transformation.

[0003] However, the existing automated distribution technology for medical consumables is unable to achieve rapid cross-regional transportation of medical consumables while ensuring distribution flexibility, resulting in the automated distribution process for medical consumables being unable to maintain high efficiency under complex and dynamic medical needs. Summary of the Invention

[0004] The present application provides an intelligent distribution system and method for medical consumables to solve the above-mentioned technical problems.

[0005] In a first aspect, the present application provides a method for intelligent distribution of medical consumables, the method comprising: Obtaining a consumables demand information set, and generating a delivery task queue based on the consumables demand information set; Acquire a real-time device node status information set, and determine a target drug pickup node information set, a starting delivery node information set, and a handover delivery node information set based on the real-time device node status information set and the delivery task queue; According to the target medicine collection node information set, the starting delivery node information set and the handover delivery node information set, dynamic delivery task merging is performed, the consumables delivery plan is determined and executed, and the consumables delivery report is determined and output.

[0006] Through this solution, based on the differentiated characteristics of different consumables demands in the consumables demand information set, a distribution task queue with a priority mechanism is constructed to cover the forms of medical consumables demand in different medical scenarios. On this basis, by parsing the real-time device node status information set, the corresponding status of each current distribution device is optimized and matched with each distribution task in the distribution task queue, and the target drug collection node information set, the starting distribution node information set and the handover distribution node information set are constructed respectively, so that the subsequent consumables distribution plan formulated based on the above information sets can achieve a balance between distribution efficiency and distribution flexibility. Furthermore, based on the spatial overlap relationship of the distribution paths corresponding to different distribution tasks, the distribution paths of different distribution tasks are targeted and merged, and the corresponding consumables distribution plan is formulated and executed. While improving the distribution efficiency, it can effectively reduce the distribution cost, and provide the corresponding consumables distribution report to the system maintenance personnel so that the maintenance personnel can accurately grasp the distribution status of each distribution task.

[0007] Optionally, the consumables demand information set includes a location code of a demanding department, a consumables type identifier, and a demand urgency coefficient; The consumables demand information set is obtained by an OCR medical prescription recognition unit for extracting medical consumables text keywords in cooperation with a consumables demand voice recognition unit for extracting medical consumables voice keywords; The OCR medical prescription recognition unit and the consumables demand voice recognition unit are both integrated into the hospital information management system.

[0008] Through this solution, by integrating OCR image recognition and voice interaction technology, a dual-channel mechanism is used to analyze and extract key information pointing to medical consumables in medical prescriptions and medical staff's language information, generate corresponding structured consumables demand data, and construct the corresponding consumables demand information set, thereby improving the efficiency and accuracy of the consumables demand collection process, adapting to emergency medical scenarios, and improving the processing speed of the medical consumables distribution process.

[0009] Optionally, generating a delivery task queue according to the consumables demand information set includes: Normalizing the demand urgency coefficient, assigning a delivery task label to each delivery task based on the normalization result, and using the corresponding department location code as a unique task identifier corresponding to each delivery task; The delivery task labels include scheduled delivery tasks, regular delivery tasks and emergency delivery tasks; Identify the over-limit consumables according to the consumable type identifier, decompose the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, and assign the same task unique identifier to the plurality of parallel subtasks; The emergency delivery task is inserted into the head of the delivery task queue based on the task initiation time, the regular delivery task is connected to the emergency delivery task, and the scheduled delivery task is added to the tail of the delivery task queue based on the task reservation time to construct the delivery task queue.

[0010] Through this solution, based on the normalized processing results of the demand urgency coefficient, the urgency gradient differences of different delivery tasks are clarified, and corresponding delivery task labels are assigned to each delivery task. On this basis, over-limit consumables are identified, and according to the characteristics of over-limit consumables, the corresponding delivery tasks are decomposed into several parallel sub-tasks. The emergency delivery tasks, regular delivery tasks and scheduled delivery tasks after task label division and parallel sub-task decomposition are inserted into the head, middle and tail of the delivery task queue according to their corresponding timing characteristics, thereby realizing the construction of the delivery task queue and improving the overall responsiveness of the automated delivery process of medical consumables.

[0011] Optionally, identifying the over-limit consumables according to the consumable type identifier, and decomposing the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, includes: The over-limit consumables are medical consumables that cannot be delivered by a single delivery robot in a single delivery mission; Determining the total weight and total space volume of the medical consumables required for the current delivery task based on the consumable type identifier; Comparing the total weight of the consumables and the total volume of the consumable space with the rated loading weight and rated loading capacity of the delivery robot, respectively; if the total weight of the consumables or the total volume of the consumable space exceeds the rated loading weight or the rated loading capacity, determining the medical consumables required for the current delivery task as the over-limit consumables; Determining, based on the consumable type identifier, a number of independently packaged units that can be assembled corresponding to the over-limit consumable, each of which does not exceed the rated loading weight and the rated loading volume; According to the plurality of said assemblable independent packaging units, the distribution task corresponding to the over-limit consumables is divided into a plurality of said parallel sub-tasks, and the total weight and total volume of the plurality of said assemblable independent packaging units corresponding to each of said parallel sub-tasks do not exceed the rated loading weight and the rated loading capacity.

[0012] Through this solution, based on the total weight and total space volume of the medical consumables required for the delivery task, and taking the rated loading weight and rated loading volume of the delivery robot as the benchmark, over-limit consumables are identified, and based on the number of assemblable independent packaging units that make up the over-limit consumables, the sub-task division principle is that the total weight and total volume of the number of assemblable independent packaging units corresponding to each parallel sub-task do not exceed the rated loading weight and rated loading volume. This improves the delivery stability of over-limit consumables while improving the delivery efficiency of over-limit consumables.

[0013] Optionally, the real-time device node status information set includes floor vertical pipeline topology data, robot real-time position coordinate set, and consumable wall inventory capacity status data set; The vertical pipeline topology data is three-dimensional structural data that describes the spatial layout of vertical transmission pipelines used for handing over medical consumables between floors and the real-time status of layered isolation valves.

[0014] Through this solution, real-time equipment node status information is constructed using floor vertical pipeline topology data, robot real-time position coordinate set, and consumable wall inventory capacity status data set to clearly reflect the real-time status of the entities involved in the distribution process, thereby improving the real-time performance of subsequent distribution plans. The spatial layout of vertical transmission pipelines and the real-time status of layered isolation valves are used to clarify the real-time availability of vertical transmission pipelines at different locations, providing a reliable data foundation for the formulation and implementation of subsequent distribution strategies.

[0015] Optionally, the determining of the target medicine pickup node information set, the starting delivery node information set, and the handover delivery node information set based on the real-time device node status information set and the delivery task queue includes: Through the task allocation optimization strategy, according to the department location code and consumable type identifier of each task in the delivery task queue, combined with the consumable wall inventory capacity status data set, the optimal consumable wall node corresponding to each delivery task is matched as the target drug collection node, and the target drug collection node information set is constructed; Using a shortest path optimization algorithm, based on the robot's real-time position coordinate set and the demand urgency coefficient, the shortest path from each delivery robot to the corresponding target medication pickup node is matched, the starting point corresponding to each shortest path is selected, and the starting delivery node information set is constructed; Through a dynamic efficiency analysis strategy, according to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, the corresponding optimal transfer node is selected from the intersection points of each floor of the corresponding vertical pipeline to construct the handover and distribution node information set.

[0016] Through this solution, based on the task allocation optimization strategy, the shortest path optimization algorithm and the dynamic efficiency analysis strategy, the optimal consumables wall node, the shortest path starting point and the optimal transfer node corresponding to each distribution task are matched respectively, and the target drug collection node information set, the starting distribution node information set and the handover distribution node information set are constructed. The distribution path obtained according to the above information sets is highly consistent with the real-time status of the current equipment nodes, which improves the accuracy of the distribution path analysis, prevents additional distribution delays caused by distribution path conflicts, and thus improves the distribution efficiency of medical consumables.

[0017] Optionally, the task allocation optimization strategy includes: Constructing a two-dimensional distribution benefit matrix with the consumable wall nodes as rows and the distribution tasks as columns based on the consumable wall inventory capacity status dataset and the distribution task queue; The matrix element values ​​in the two-dimensional distribution effect matrix are obtained by weighted evaluation of the estimated transportation time and inventory fulfillment rate between the consumables wall node and the corresponding demand department; Solving the two-dimensional distribution benefit matrix using the Hungarian algorithm with distribution constraints to determine the optimal consumables wall node corresponding to each distribution task; The delivery constraint condition is that the number of delivery tasks served simultaneously by a single consumable wall node does not exceed a preset consumable wall concurrency threshold.

[0018] Through this solution, a two-dimensional distribution benefit matrix is ​​constructed with consumable wall nodes as rows and distribution tasks as columns, and the matrix element values ​​are obtained by weighted evaluation of estimated transportation time and inventory fulfillment rate. On this basis, the Hungarian algorithm with distribution constraints is used to obtain the optimal consumable wall nodes corresponding to each distribution task. Through the dynamic benefit matrix optimization mechanism, the efficiency and response speed of the overall distribution process are further improved.

[0019] Optionally, the dynamic efficiency analysis strategy includes: According to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, a plurality of pre-selected target vertical pipelines connected to the floor where the demand department corresponding to the current delivery task is located are screened; Analyze the plurality of pre-selected target vertical pipelines according to the starting node corresponding to the current delivery task in the starting delivery node information set, and select the delivery entrance of the pre-selected target vertical pipeline closest to the target medicine collection node among the plurality of pre-selected target vertical pipelines as the target delivery entrance; According to the target delivery inlet, the delivery outlet corresponding to the other end of the preselected target vertical pipeline is used as the target delivery outlet; Based on the target delivery exit, analyzing the real-time position coordinate set of the robot, and selecting the delivery robot closest to the target delivery exit on the floor where the target delivery exit is located as the target delivery object; The delivery node information set is generated based on the target delivery entrance, the target delivery exit, and the target handover object corresponding to each delivery task.

[0020] Through this solution, the vertical pipeline topology data of each floor and the real-time position coordinate set of the robot are analyzed. By selecting the available delivery pipelines and the optimal handover nodes, the incidents of cross-floor consumables transfer are significantly reduced. At the same time, local pipeline congestion caused by scheduling errors is avoided, further ensuring the stability of cross-floor consumables transfer.

[0021] Optionally, the dynamic delivery task merging according to the target medicine pickup node information set, the starting delivery node information set, and the handover delivery node information set includes: Planning a delivery route corresponding to each delivery task according to the target drug pickup node information set, the starting delivery node information set, and the handover delivery node information set; The overlapping features of the delivery routes corresponding to different delivery tasks on different floors are analyzed, and the delivery tasks corresponding to the overlapping delivery routes on the same floor are partially merged to achieve the dynamic delivery task merging.

[0022] This solution utilizes a local path merging mechanism to reduce the robot's total driving distance and shorten the overall task completion time. It also reduces the number of idle runs and repeated starts and stops of the robot, thereby improving the flexibility of the delivery solution and achieving efficient resource scheduling in complex medical scenarios.

[0023] In a second aspect, the present application provides an intelligent distribution system for medical consumables, the system comprising: A task analysis module is used to obtain a consumables demand information set and generate a delivery task queue based on the consumables demand information set; A node analysis module is used to obtain a real-time device node status information set, and based on the real-time device node status information set and the delivery task queue, determine a target drug collection node information set, a starting delivery node information set, and a handover delivery node information set; The plan execution module is used to dynamically merge delivery tasks based on the target drug collection node information set, the starting delivery node information set and the handover delivery node information set, determine and execute the consumables delivery plan, and output the consumables delivery report. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0025] Figure 1 A schematic diagram of an application scenario provided in one embodiment of the present application; Figure 2 A flowchart of a method for intelligent distribution of medical consumables provided in one embodiment of the present application; Figure 3 A schematic structural diagram of an intelligent distribution system for medical consumables provided in accordance with one embodiment of the present application [Z1]. DETAILED DESCRIPTION

[0026] To make the purpose, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0027] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates an "or" relationship between the related objects.

[0028] The embodiments of the present application are described in further detail below with reference to the accompanying drawings.

[0029] Existing automated distribution technology for medical consumables is difficult to achieve rapid cross-regional transportation of medical consumables while ensuring distribution flexibility, resulting in the difficulty of the automated distribution process of medical consumables to maintain high efficiency under complex and dynamic medical needs. Based on this, the present application provides a medical consumables intelligent distribution system and method. According to the differentiated characteristics between different consumables demands in the consumables demand information set, a distribution task queue with a priority mechanism is constructed to cover the medical consumables demand forms in different medical scenarios. On this basis, by parsing the real-time device node status information set, the current corresponding status of each distribution device is optimized and matched with each distribution task in the distribution task queue, and the target drug collection node information set, the starting distribution node information set, and the handover distribution node information set are respectively constructed, so that the consumables distribution plan subsequently formulated based on the above information sets can achieve a balance between distribution efficiency and distribution flexibility. Furthermore, according to the spatial overlap relationship of the distribution paths corresponding to different distribution tasks, the distribution paths of different distribution tasks are targeted and merged, and the corresponding consumables distribution plan is formulated and executed. While improving distribution efficiency, it can effectively reduce distribution costs, and provide the corresponding consumables distribution report to system maintenance personnel so that maintenance personnel can accurately grasp the distribution status of each distribution task.

[0030] Figure 1 This is a schematic diagram of an application scenario provided by this application. In the process of distributing medical consumables, the method provided by this application is applied to achieve a balance between distribution efficiency and distribution flexibility.

[0031] Specifically, the method of the present application is applied to any server that communicates with the hospital information management system and the Internet of Things sensor. For specific implementation methods, please refer to the following embodiments.

[0032] Figure 2 This is a flowchart of a medical consumables intelligent distribution method provided in one embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario [Z2]. Figure 2 As shown, the method includes: S201: Obtain a consumables demand information set, and generate a delivery task queue based on the consumables demand information set.

[0033] The consumables demand information set may refer to a set of medical consumables demand data proposed by various departments within the hospital. The consumables demand information set may be collected through an interface provided by the hospital information management system.

[0034] The delivery task queue may be a delivery task sequence with a priority mechanism generated according to a consumables demand information set.

[0035] Specifically, existing automated distribution technologies for medical consumables usually adopt a static task allocation mechanism, that is, sequential task allocation is performed according to the time when medical consumables are required. This method is difficult to adapt to the needs of dynamic medical scenarios. This solution actively collects medical consumables needs, covers the forms of medical consumables needs in different medical scenarios, and prioritizes each medical consumables need, and constructs a distribution task queue with a priority mechanism, which serves as the data basis for subsequent distribution task allocation and execution.

[0036] S202: Acquire a real-time device node status information set, and determine a target drug pickup node information set, a starting delivery node information set, and a handover delivery node information set based on the real-time device node status information set and the delivery task queue.

[0037] The real-time device node status information set can be a status set of physical devices involved in the medical consumables distribution process, such as distribution robots, transportation pipelines, etc. The real-time device node status information set is obtained by integrating the Internet of Things sensors in each physical device node.

[0038] The target drug collection node information set can be the optimal drug collection node matched to the distribution task, which is selected based on inventory capacity and transportation efficiency.

[0039] The starting delivery node information set may be the optimal delivery departure node matched to the delivery task.

[0040] The handover delivery node information set can be the optimal pipeline transfer node matched to the delivery task, and is used for transfer coordination of cross-floor delivery tasks.

[0041] Specifically, there are two main methods for automated distribution of medical consumables in existing hospitals. One is pipeline distribution, which is fast and subject to fewer interference factors, but has high initial planning costs, low flexibility, and weak scalability. The other is robotic distribution, which is highly flexible, low-planning, and highly scalable, but its distribution efficiency is easily affected by external factors, especially when distributing consumables between multiple floors. This solution combines cross-floor pipeline distribution with same-floor robotic distribution by constructing a three-dimensional transportation network of "vertical pipelines + horizontal robots." While ensuring efficiency, it also improves the cross-floor anti-interference strength and flexibility of the medical consumables distribution process. By analyzing real-time device node status information sets (including pipeline status, robot status, and medication collection point status), the current corresponding status of each distribution device is optimally matched with each distribution task in the distribution task queue. The optimal target medication collection node, starting distribution node, and handover distribution node corresponding to each task are determined. The target medication collection node information set, starting distribution node information set, and handover distribution node information set are respectively constructed. The subsequent consumables distribution plan formulated based on these information sets can achieve a balance between distribution efficiency and distribution flexibility.

[0042] S203. Dynamically merge delivery tasks based on the target drug pickup node information set, the starting delivery node information set, and the handover delivery node information set, determine and execute the consumables delivery plan, and determine and output the consumables delivery report.

[0043] Dynamic delivery task merging can be based on the overlapping characteristics of delivery routes, merging tasks with overlapping paths on the same floor into composite tasks to reduce the number of empty driving processes of robots.

[0044] The consumables distribution plan can be a distribution scheduling instruction set including task paths, execution time and consumables allocation plan.

[0045] The consumables delivery report can be a visual report that records key indicators of the delivery process (such as task completion time and consumables loss rate).

[0046] Specifically, since the consumables distribution path involves several nodes of three major types, namely, drug collection nodes, starting distribution nodes, and handover distribution nodes, different distribution tasks may pass through the same nodes within the same distribution time period. In order to further improve the distribution efficiency, the spatial overlap relationship of the distribution paths corresponding to different distribution tasks is analyzed based on the labeled drug collection node information set, the starting distribution node information set, and the handover distribution node information set, so as to carry out targeted mergers of the distribution paths of different distribution tasks. While improving the distribution efficiency, it can effectively reduce the distribution cost. By integrating the merged distribution paths corresponding to each assigned task machine, a corresponding consumables distribution plan is generated, and through a unified consumables distribution instruction distribution control platform, according to the instructions in the consumables distribution plan, the physical equipment corresponding to the starting distribution node, drug collection node, and handover distribution node involved in the distribution process is controlled to complete the automated distribution of medical consumables. By collecting system logs and using data visualization technology, a corresponding consumables distribution report is generated, and the consumables distribution report is provided to the corresponding system maintenance personnel through human-computer interaction equipment, such as high-definition display screens.

[0047] Through this solution, based on the differentiated characteristics of different consumables demands in the consumables demand information set, a distribution task queue with a priority mechanism is constructed to cover the forms of medical consumables demand in different medical scenarios. On this basis, by parsing the real-time device node status information set, the corresponding status of each current distribution device is optimized and matched with each distribution task in the distribution task queue, and the target drug collection node information set, the starting distribution node information set and the handover distribution node information set are constructed respectively, so that the subsequent consumables distribution plan formulated based on the above information sets can achieve a balance between distribution efficiency and distribution flexibility. Furthermore, based on the spatial overlap relationship of the distribution paths corresponding to different distribution tasks, the distribution paths of different distribution tasks are targeted and merged, and the corresponding consumables distribution plan is formulated and executed. While improving the distribution efficiency, it can effectively reduce the distribution cost, and provide the corresponding consumables distribution report to the system maintenance personnel so that the maintenance personnel can accurately grasp the distribution status of each distribution task.

[0048] In some embodiments, the consumables demand information set includes a location code of the demand department, a consumables type identifier, and a demand urgency coefficient; the consumables demand information set is obtained through an OCR medical prescription recognition unit for extracting medical consumables text keywords, in conjunction with a consumables demand voice recognition unit for extracting medical consumables voice keywords; the OCR medical prescription recognition unit and the consumables demand voice recognition unit are both integrated into the hospital information management system.

[0049] The required department location code can be a unique location identifier generated by the hospital's spatial geographic information system, using the three-level coding rule of "building number-floor number-room number" (such as "B3-F12-R0815"), which is derived from the preset topological data in the hospital building BIM model to ensure that the code strictly corresponds to the physical space coordinates.

[0050] The consumable type identifier may be a unique identifier used to refer to a specific medical consumable.

[0051] The demand urgency coefficient may be a dynamic parameter that quantitatively characterizes the urgency of the demand for consumables, and the value range is (0,1).

[0052] The medical consumables text keywords may be text keywords related to medical consumables in a prescription provided by a doctor.

[0053] The OCR medical prescription recognition unit can be a functional module based on optical character recognition technology combined with a neural network to extract key information such as the name, specification, and quantity of consumables in a medical prescription.

[0054] The medical consumables voice keywords may be keywords related to medical consumables in the voice input information provided by medical staff.

[0055] The consumables demand voice recognition unit can be a voice interaction module integrated with a natural language processing engine, through which medical staff can quickly convey consumables needs.

[0056] The hospital information management system can be a heterogeneous system integration platform running within the hospital. It realizes data interoperability between the medical order system, material system and nursing system through the HL7 standard interface. The OCR and speech recognition modules are connected to the system bus through the RESTful API to ensure real-time synchronization of demand information and hospital information business flows.

[0057] Specifically, traditional consumables demand collection relies on manual entry into electronic forms, which suffers from high response delays and high error rates. In emergency medical scenarios, it is difficult for medical staff to manually enter consumables requirements through terminal operations. This solution integrates optical character recognition (OCR) image recognition and voice interaction technologies, utilizing a dual-channel mechanism to efficiently collect consumables requirements. The OCR medical prescription recognition unit monitors the HIS order queue in real time. When a new prescription image is detected, a preprocessing process is initiated: the prescription image is grayscaled, tilted, and noise filtered. The text region is then extracted using a pretrained ResNet-50 image classification network. A convolutional neural network model is then used to perform text recognition, outputting structured consumables demand data extracted from the prescription text. Simultaneously, the consumables demand voice recognition unit continuously monitors audio input from the medical staff's workstation. After capturing a preset wake-up word (such as "consumables application"), a voice activity detection algorithm is used to segment the valid instruction segments in the audio data. The Transformer model is used to convert speech to text, and the legitimacy of the consumable name is verified by combining it with the knowledge graph. The structured consumables demand data extracted from the speech is then output.

[0058] Through this solution, by integrating OCR image recognition and voice interaction technology, a dual-channel mechanism is used to analyze and extract key information pointing to medical consumables in medical prescriptions and medical staff's language information, generate corresponding structured consumables demand data, and construct the corresponding consumables demand information set, thereby improving the efficiency and accuracy of the consumables demand collection process, adapting to emergency medical scenarios, and improving the processing speed of the medical consumables distribution process.

[0059] In some embodiments, the demand urgency coefficient is normalized, and a delivery task label is assigned to each delivery task based on the normalization result, and the corresponding department location code is used as the task unique identifier corresponding to each delivery task; the delivery task labels include scheduled delivery tasks, regular delivery tasks and emergency delivery tasks; based on the consumable type identifier, over-limit consumables are identified, and based on the over-limit consumables, the corresponding delivery task is decomposed into several parallel subtasks, and the same task unique identifier is assigned to several parallel subtasks; the emergency delivery task is inserted into the head of the delivery task queue based on the task initiation time, the regular delivery task is continued after the emergency delivery task, and the scheduled delivery task is added to the tail of the delivery task queue based on the task appointment time to construct a delivery task queue.

[0060] Normalization processing can be a numerical processing process of mapping the original urgency coefficient to a standard interval.

[0061] The delivery task label can be an identification system used to classify and manage delivery tasks.

[0062] The unique task identifier may be identification information used to distinguish delivery tasks.

[0063] Scheduled delivery tasks can be daily delivery tasks submitted by medical staff, such as the daily periodic replenishment needs of the department.

[0064] Routine delivery tasks can be non-emergency consumables delivery tasks submitted by medical staff, such as the need to replenish consumables during routine medical procedures in the department.

[0065] Emergency delivery tasks can be delivery tasks submitted by medical staff that need to be processed in a short period of time, such as the need for consumables in operating rooms and emergency rooms.

[0066] Over-limit consumables may refer to the consumables demand when the scale of a single delivery exceeds the delivery capacity of a single delivery device.

[0067] The parallel subtasks may be collaborative delivery instructions generated after disassembling the over-limit consumables into multiple independent transport units.

[0068] The task initiation time may be the timestamp of submitting the data corresponding to the delivery task.

[0069] The task appointment time may be the scheduled consumables arrival time in the scheduled delivery task.

[0070] Specifically, the traditional delivery system adopts a single-dimensional priority rule (such as a single "emergency task priority"), which cannot adapt to complex clinical scenarios and does not take into account the gradient differences in task urgency, which can easily cause serious delays in non-emergency delivery tasks. The Min-Max normalization method is used to normalize the demand urgency coefficients corresponding to different delivery tasks to eliminate dimensional differences, and then assign delivery task labels (appointment delivery tasks, regular delivery tasks and emergency delivery tasks) to delivery tasks based on the division thresholds obtained by fitting historical data to accurately reflect the gradient differences in task urgency. On this basis, considering that large consumables that may be required in special medical scenarios (such as orthopedic implant sets and extracorporeal circulation equipment) are difficult to be delivered by a single delivery machine People complete the delivery in a single process, while multi-batch sequential delivery will cause significant delivery delays. Therefore, through the analysis of consumables demand, the over-limit consumables in the consumables demand are identified, and the delivery tasks corresponding to the over-limit consumables are decomposed into several parallel delivery sub-tasks. These sub-tasks are achieved by handing over the different components that make up the over-limit consumables to different delivery robots for parallel delivery within the same time period. The over-limit consumables are constrained to arrive at the same delivery destination through the same task unique identifier, thereby achieving low-latency processing of the over-limit consumables; the emergency delivery tasks, regular delivery tasks and scheduled delivery tasks after task label division and parallel sub-task decomposition are inserted into the head, middle and tail of the delivery task queue according to their corresponding timing characteristics, thereby realizing the construction of the delivery task queue.

[0071] Through this solution, based on the normalized processing results of the demand urgency coefficient, the urgency gradient differences of different delivery tasks are clarified, and corresponding delivery task labels are assigned to each delivery task. On this basis, over-limit consumables are identified, and according to the characteristics of over-limit consumables, the corresponding delivery tasks are decomposed into several parallel sub-tasks. The emergency delivery tasks, regular delivery tasks and scheduled delivery tasks after task label division and parallel sub-task decomposition are inserted into the head, middle and tail of the delivery task queue according to their corresponding timing characteristics, thereby realizing the construction of the delivery task queue and improving the overall responsiveness of the automated delivery process of medical consumables.

[0072] In some embodiments, over-limit consumables are medical consumables that a single delivery robot cannot complete the delivery of in a single delivery task; based on the consumable type identifier, the total weight of the consumables and the total space volume of the consumables required for the current delivery task are determined; the total weight of the consumables and the total space volume of the consumables are compared with the rated loading weight and rated loading volume of the delivery robot respectively. If the total weight of the consumables or the total space volume of the consumables exceeds the rated loading weight or the rated loading volume, the medical consumables required in the current delivery task are determined to be over-limit consumables; based on the consumable type identifier, a number of assemblable independent packaging units corresponding to the over-limit consumables are determined, and each assemblable independent packaging unit does not exceed the rated loading weight and the rated loading volume; based on the number of assemblable independent packaging units, the delivery task corresponding to the over-limit consumables is divided into a number of parallel sub-tasks, and the total weight and total volume of the corresponding several assemblable independent packaging units under each parallel sub-task do not exceed the rated loading weight and the rated loading volume.

[0073] The total weight of consumables may be the total mass of medical consumables required in the current delivery task.

[0074] The total space volume of consumables may be the total space volume occupied by medical consumables required in the current delivery task.

[0075] The rated loading weight may be the maximum weight load of the delivery robot, which is obtained through the delivery robot design parameter set.

[0076] The rated loading volume may be the maximum volumetric load of the delivery robot, which is obtained through the delivery robot design parameter set.

[0077] The assemblable independent packaging unit can be a standardized transport unit formed by intelligent disassembly of oversized consumables.

[0078] Specifically, according to the consumable type identifier, the medical consumable parameter information in the material database is retrieved to obtain the total weight and total space volume of the medical consumables required in the current delivery task, and the total weight and total space volume of the consumables are numerically compared with the rated loading weight and rated loading volume of the delivery robot respectively. If the total weight or total space volume of the consumables exceeds the rated loading weight or rated loading volume, the medical consumables required in the current delivery task are determined to be over-limit consumables. According to the consumable type identifier, the medical consumables technical white paper corresponding to the over-limit consumables is retrieved to determine the number of assemblable independent packaging units that constitute the over-limit consumables. The task disassembly principle is that the total weight and total volume of the corresponding several assemblable independent packaging units under each parallel subtask do not exceed the rated loading weight and rated loading volume, and the delivery task corresponding to the over-limit consumables is divided into several parallel subtasks.

[0079] Through this solution, based on the total weight and total space volume of the medical consumables required for the delivery task, and taking the rated loading weight and rated loading volume of the delivery robot as the benchmark, over-limit consumables are identified, and based on the number of assemblable independent packaging units that make up the over-limit consumables, the sub-task division principle is that the total weight and total volume of the number of assemblable independent packaging units corresponding to each parallel sub-task do not exceed the rated loading weight and rated loading volume. This improves the delivery stability of over-limit consumables while improving the delivery efficiency of over-limit consumables.

[0080] In some embodiments, the real-time device node status information set includes floor vertical pipeline topology data, a robot real-time position coordinate set, and a consumables wall inventory capacity status data set; the vertical pipeline topology data is a three-dimensional structural data that describes the spatial layout of the vertical transmission pipelines used for the transfer of medical consumables between floors and the real-time status of the layered isolation valves.

[0081] The floor vertical pipeline topology data may be a structured data set representing the three-dimensional spatial relationship of vertical transportation channels in a hospital building. The floor vertical pipeline topology data is obtained through the BIM building information model corresponding to the hospital.

[0082] The robot real-time position coordinate set may be a data set recording the real-time spatial positions of all delivery robots. The robot real-time position coordinate set is provided by a positioning module integrated within each delivery robot.

[0083] The consumables wall inventory capacity status dataset may be a data set representing the consumables storage status in the consumables wall used to store medical consumables on each floor of the hospital. The consumables wall inventory capacity status dataset is provided by an embedded IoT sensor integrated in the consumables wall.

[0084] The spatial layout may be the layout positions of each vertical transmission pipeline in the hospital.

[0085] The real-time status of the layered isolation valve may be the real-time open / close status of the isolation valve used to control the connection status between the vertical transmission pipeline and each floor.

[0086] Specifically, in this solution, the process of using delivery robots and transmission pipelines to deliver medical consumables involves three key entities, namely, the delivery robot responsible for transferring medical consumables within a single floor, the vertical transmission pipeline used to quickly transfer medical consumables between different floors, and the consumables wall used to store and manage various types of medical consumables. Real-time device node status information is constructed through the floor vertical pipeline topology data, the robot's real-time position coordinate set, and the consumables wall inventory capacity status data set, so that the real-time device node status information can clearly reflect the real-time status of the entities involved in the distribution process, so as to improve the real-time performance of subsequent distribution plans. The spatial layout of the vertical transmission pipeline and the real-time status of the layered isolation valves can be used to clarify the real-time availability of the vertical transmission pipeline at different locations, providing a reliable data basis for the formulation and implementation of subsequent distribution strategies.

[0087] Through this solution, real-time equipment node status information is constructed using floor vertical pipeline topology data, robot real-time position coordinate set, and consumable wall inventory capacity status data set to clearly reflect the real-time status of the entities involved in the distribution process, thereby improving the real-time performance of subsequent distribution plans. The spatial layout of vertical transmission pipelines and the real-time status of layered isolation valves are used to clarify the real-time availability of vertical transmission pipelines at different locations, providing a reliable data foundation for the formulation and implementation of subsequent distribution strategies.

[0088] In some embodiments, through the task allocation optimization strategy, according to the department location code and consumable type identifier of each task in the distribution task queue, combined with the consumable wall inventory capacity status data set, the optimal consumable wall node corresponding to each distribution task is matched as the target drug collection node, and the target drug collection node information set is constructed; through the shortest path optimization algorithm, based on the robot's real-time position coordinate set and the demand urgency coefficient, the shortest path from each distribution robot to the corresponding target drug collection node is matched, the path starting point corresponding to each shortest path is screened, and the starting distribution node information set is constructed; through the dynamic efficiency analysis strategy, according to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, the corresponding optimal transfer node is selected at the intersection point of each floor of the corresponding vertical pipeline, and the handover distribution node information set is constructed.

[0089] The task allocation optimization strategy can be a medical consumables distribution task allocation mechanism designed based on operations research mechanisms.

[0090] The optimal consumables wall node can be the consumables wall that meets the quantity requirements of the current delivery task and requires the shortest delivery time.

[0091] The shortest path optimization algorithm can be a graph theory method that quantifies the minimum total cost of a path between two points in a weighted graph.

[0092] The dynamic efficiency analysis strategy can be a decision model used to evaluate the transportation efficiency of the current distribution task under each vertical pipeline.

[0093] Specifically, there are usually multiple delivery tasks executed in parallel in the same period, resulting in the occupancy status of different vertical pipelines and the storage status of different consumable walls being in a highly dynamic process. Therefore, in the process of analyzing the target drug-picking node, starting delivery node and handover delivery node that determine the delivery path, it is necessary to highly combine the real-time status of each device node, and assign corresponding target drug-picking nodes, starting delivery nodes and handover delivery nodes to the current delivery task without affecting other delivery tasks; through the task allocation optimization strategy, among several consumable walls with the current target delivery consumables, the consumable wall with the shortest transportation time is screened as the target drug-picking node corresponding to the current delivery task, and the different delivery tasks are comprehensively considered. The target drug-picking node is selected and the target drug-picking node information set is constructed; the shortest path optimization algorithm, such as the A* algorithm, is used to perform descending shortest path planning based on the real-time position coordinate set of the robot and the demand urgency coefficient. That is, the higher the demand urgency coefficient, the higher the priority is for assigning the shortest path between the corresponding delivery robot and the target drug-picking node to the corresponding delivery task, and the starting point of the corresponding path is used as the corresponding starting delivery node to construct the starting delivery node information set; the dynamic efficiency analysis strategy is used to take the unoccupied vertical pipeline that is currently closest to the target drug-picking node and the delivery robot that is closest to the target transmission floor exit of the vertical pipeline as the optimal transfer nodes for the current delivery task, and to construct the handover delivery node information set.

[0094] Through this solution, based on the task allocation optimization strategy, the shortest path optimization algorithm and the dynamic efficiency analysis strategy, the optimal consumables wall node, the shortest path starting point and the optimal transfer node corresponding to each distribution task are matched respectively, and the target drug collection node information set, the starting distribution node information set and the handover distribution node information set are constructed. The distribution path obtained according to the above information sets is highly consistent with the real-time status of the current equipment nodes, which improves the accuracy of the distribution path analysis, prevents additional distribution delays caused by distribution path conflicts, and thus improves the distribution efficiency of medical consumables.

[0095] In some embodiments, based on the consumable wall inventory capacity status data set and the distribution task queue, a two-dimensional distribution benefit matrix is ​​constructed with consumable wall nodes as rows and distribution tasks as columns; the matrix element values ​​in the two-dimensional distribution effect matrix are obtained by weighted evaluation of the estimated transportation time and inventory fulfillment rate between the consumable wall node and the corresponding demand department; the two-dimensional distribution benefit matrix is ​​solved by the Hungarian algorithm with distribution constraints to determine the optimal consumable wall node corresponding to each distribution task; the distribution constraint is that the number of distribution tasks served simultaneously by a single consumable wall node does not exceed the preset consumable wall concurrency threshold.

[0096] The two-dimensional distribution benefit matrix can be a numerical matrix that comprehensively reflects the adaptation benefit of the medicine collection wall node to the distribution task.

[0097] The estimated transport time can be the estimated transport time from the drug collection wall node to the required department, which is calculated based on the floor map path distance and the average speed of the robot.

[0098] The inventory fulfillment rate can be the completeness of the target consumables in the medicine wall node.

[0099] The delivery constraint can be to limit the number of delivery tasks that a single drug collection wall node can serve simultaneously to no more than a preset drug collection wall concurrency threshold (such as processing a maximum of 3 tasks at the same time) to avoid node overload.

[0100] The Hungarian algorithm can be a combinatorial optimization algorithm based on the bipartite graph matching problem solving mechanism.

[0101] The preset concurrency threshold of the consumables wall can be the upper limit of the delivery tasks that a single drug collection wall node can handle simultaneously.

[0102] The matrix element values ​​can be quantitative values ​​used to comprehensively represent the distribution benefits.

[0103] Specifically, the traditional method uses fixed rules (such as nearest allocation) to match the drug collection wall with delivery tasks, which cannot dynamically respond to inventory changes (such as temporary shortage of consumables at a certain node) or equipment load fluctuations (such as busy consumable scheduling equipment), resulting in task backlogs or repeated scheduling; with consumable wall nodes as rows and delivery tasks as columns, a two-dimensional distribution benefit matrix is ​​constructed by combining the matrix element values ​​obtained by weighted evaluation of estimated transportation time and inventory satisfaction rate. On this basis, the Hungarian algorithm is initialized, and the matrix rows (drug collection wall nodes) and columns (delivery tasks) are used as vertices on both sides of the bipartite graph. The "maximum number of assignable tasks" is set as the concurrency threshold for each drug collection wall node to ensure that the algorithm does not exceed this limit during the matching process. Through row and column label adjustment and augmenting path search, the task-node matching scheme with the maximum total benefit is found, and the matching result, that is, the optimal drug collection wall node corresponding to each delivery task, is output.

[0104] Through this solution, a two-dimensional distribution benefit matrix is ​​constructed with consumable wall nodes as rows and distribution tasks as columns, and the matrix element values ​​are obtained by weighted evaluation of estimated transportation time and inventory fulfillment rate. On this basis, the Hungarian algorithm with distribution constraints is used to obtain the optimal consumable wall nodes corresponding to each distribution task. Through the dynamic benefit matrix optimization mechanism, the efficiency and response speed of the overall distribution process are further improved.

[0105] In some embodiments, according to the real-time status of the layered isolation valves in the floor vertical pipeline topology data, several pre-selected target vertical pipelines connected to the floor where the demand department corresponding to the current delivery task is located are screened; according to the starting node corresponding to the current delivery task in the starting delivery node information set, several pre-selected target vertical pipelines are analyzed, and the delivery entrance of the pre-selected target vertical pipeline closest to the target medicine collection node among the several pre-selected target vertical pipelines is used as the target delivery entrance; according to the target delivery entrance, the delivery outlet at the other end of the corresponding pre-selected target vertical pipeline is used as the target delivery outlet; based on the target delivery outlet, the real-time position coordinate set of the robot is analyzed, and the delivery robot closest to the target delivery outlet in the floor where the target delivery outlet is located is used as the target handover object; according to the target delivery entrance, target delivery outlet and target handover object corresponding to each delivery task, the delivery node information set is handed over.

[0106] The pre-selected target vertical pipeline may be a set of vertical pipeline candidates connected to the floor where the target department of the current delivery task is located.

[0107] The target delivery entrance may be the available vertical pipeline entrance closest to the delivery robot's starting node.

[0108] The target delivery outlet may be a floor outlet at the other end of the vertical pipeline corresponding to the target delivery inlet.

[0109] The target handover object can be an idle delivery robot that is closest to the exit on the floor where the target delivery exit is located.

[0110] Specifically, based on the unique identifier of the current delivery task, the vertical pipeline topology data of the floor is parsed, and all vertical delivery pipelines connected to the target department floor are screened. Based on the real-time status of the layered isolation valves, the real-time status of the layered isolation valves of these pipelines is checked, and only pipelines with valve status of "open" are retained to form a list of pre-selected target vertical pipelines. According to the Euclidean distance algorithm, the Euclidean distance from the robot position corresponding to the starting delivery node to the entrance of each pre-selected pipeline is quantified, and the nearest entrance is selected as the target delivery entrance. Based on the pipeline connectivity relationship in the floor vertical pipeline topology data, the corresponding exit of the pipeline on the target department floor is determined, that is, the target delivery exit. On this basis, based on the robot's real-time position coordinate set, all idle delivery robots on the target department floor are screened, the distance between each robot and the target delivery exit is quantified, and the nearest robot is selected as the target handover object. The target delivery entrance, target delivery exit, and target handover object information corresponding to the current delivery task are written into the handover delivery node information set for subsequent path planning and task merging.

[0111] Through this solution, the vertical pipeline topology data of each floor and the real-time position coordinate set of the robot are analyzed. By selecting the available delivery pipelines and the optimal handover nodes, the incidents of cross-floor consumables transfer are significantly reduced. At the same time, local pipeline congestion caused by scheduling errors is avoided, further ensuring the stability of cross-floor consumables transfer.

[0112] In some embodiments, the delivery route corresponding to each delivery task is planned based on the target drug collection node information set, the starting delivery node information set, and the handover delivery node information set; the overlapping characteristics of the delivery routes corresponding to different delivery tasks on different floors are analyzed, and the delivery tasks corresponding to the overlapping delivery routes on the same floor are partially merged to realize dynamic delivery task merging.

[0113] The overlap feature may be a characteristic indicator used to characterize the degree of overlap between different delivery routes.

[0114] Local merging can be the process of merging the delivery processes of the overlapping parts of the delivery paths corresponding to different delivery tasks.

[0115] Specifically, according to the target drug collection node information set, the starting delivery node information set and the handover delivery node information set, based on the floor map topology data, the shortest path from the starting node → drug collection node → handover node → demand department is generated for each task, and the time periods and coordinates of the same floor passed in different task paths are extracted. For tasks on the same floor whose path overlap exceeds a threshold (such as 70%), they are merged into the same batch for execution, and the same robot is assigned to the merged tasks or multiple robots are coordinated to move synchronously to ensure that the shared path segment is only occupied once, thereby realizing dynamic delivery task merging.

[0116] This solution utilizes a local path merging mechanism to reduce the robot's total driving distance and shorten the overall task completion time. It also reduces the number of idle runs and repeated starts and stops of the robot, thereby improving the flexibility of the delivery solution and achieving efficient resource scheduling in complex medical scenarios.

[0117] Figure 3 This is a schematic diagram of a medical consumables intelligent distribution system provided in one embodiment of the present application, as shown in FIG. Figure 3 As shown, a medical consumables intelligent distribution system 300 of this embodiment includes: a task analysis module 301, a node analysis module 302 and a plan execution module 303.

[0118] The task analysis module 301 is used to obtain a consumables demand information set and generate a delivery task queue based on the consumables demand information set; The node analysis module 302 is used to obtain a real-time device node status information set, and based on the real-time device node status information set and the delivery task queue, determine a target drug collection node information set, a starting delivery node information set, and a handover delivery node information set; The plan execution module 303 is used to perform dynamic delivery task merging based on the target drug collection node information set, the starting delivery node information set and the handover delivery node information set, determine and execute the consumables delivery plan, and output the consumables delivery report.

[0119] Optionally, in the task analysis module 301, the consumables demand information set includes a demand department location code, a consumables type identifier, and a demand urgency coefficient; The consumables demand information set is obtained by an OCR medical prescription recognition unit for extracting medical consumables text keywords in cooperation with a consumables demand voice recognition unit for extracting medical consumables voice keywords; The OCR medical prescription recognition unit and the consumables demand voice recognition unit are both integrated into the hospital information management system.

[0120] Optionally, the task analysis module 301 is specifically configured to: Normalizing the demand urgency coefficient, assigning a delivery task label to each delivery task based on the normalization result, and using the corresponding department location code as a unique task identifier corresponding to each delivery task; The delivery task labels include scheduled delivery tasks, regular delivery tasks and emergency delivery tasks; Identify the over-limit consumables according to the consumable type identifier, decompose the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, and assign the same task unique identifier to the plurality of parallel subtasks; The emergency delivery task is inserted into the head of the delivery task queue based on the task initiation time, the regular delivery task is connected to the emergency delivery task, and the scheduled delivery task is added to the tail of the delivery task queue based on the task reservation time to construct the delivery task queue.

[0121] Optionally, when the task analysis module 301 identifies the over-limit consumables according to the consumable type identifier and decomposes the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, it is specifically configured to: The over-limit consumables are medical consumables that cannot be delivered by a single delivery robot in a single delivery mission; Determining the total weight and total space volume of the medical consumables required for the current delivery task based on the consumable type identifier; Comparing the total weight of the consumables and the total volume of the consumable space with the rated loading weight and rated loading capacity of the delivery robot, respectively; if the total weight of the consumables or the total volume of the consumable space exceeds the rated loading weight or the rated loading capacity, determining the medical consumables required for the current delivery task as the over-limit consumables; Determining, based on the consumable type identifier, a number of independently packaged units that can be assembled corresponding to the over-limit consumable, each of which does not exceed the rated loading weight and the rated loading volume; According to the plurality of said assemblable independent packaging units, the distribution task corresponding to the over-limit consumables is divided into a plurality of said parallel sub-tasks, and the total weight and total volume of the plurality of said assemblable independent packaging units corresponding to each of said parallel sub-tasks do not exceed the rated loading weight and the rated loading capacity.

[0122] Optionally, in the node analysis module 302, the real-time device node status information set includes floor vertical pipeline topology data, robot real-time position coordinate set, and consumable wall inventory capacity status data set; The vertical pipeline topology data is three-dimensional structural data that describes the spatial layout of vertical transmission pipelines used for handing over medical consumables between floors and the real-time status of layered isolation valves.

[0123] Optionally, when the node analysis module 302 determines the target drug pickup node information set, the starting delivery node information set, and the handover delivery node information set based on the real-time device node status information set and the delivery task queue, it is specifically configured to: Through the task allocation optimization strategy, according to the department location code and consumable type identifier of each task in the delivery task queue, combined with the consumable wall inventory capacity status data set, the optimal consumable wall node corresponding to each delivery task is matched as the target drug collection node, and the target drug collection node information set is constructed; Using a shortest path optimization algorithm, based on the robot's real-time position coordinate set and the demand urgency coefficient, the shortest path from each delivery robot to the corresponding target medication pickup node is matched, the starting point corresponding to each shortest path is selected, and the starting delivery node information set is constructed; Through a dynamic efficiency analysis strategy, according to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, the corresponding optimal transfer node is selected from the intersection points of each floor of the corresponding vertical pipeline to construct the handover and distribution node information set.

[0124] Optionally, the task allocation optimization strategy in the node analysis module 302 is specifically used to: Constructing a two-dimensional distribution benefit matrix with the consumable wall nodes as rows and the distribution tasks as columns based on the consumable wall inventory capacity status dataset and the distribution task queue; The matrix element values ​​in the two-dimensional distribution effect matrix are obtained by weighted evaluation of the estimated transportation time and inventory fulfillment rate between the consumables wall node and the corresponding demand department; Solving the two-dimensional distribution benefit matrix using the Hungarian algorithm with distribution constraints to determine the optimal consumables wall node corresponding to each distribution task; The delivery constraint condition is that the number of delivery tasks served simultaneously by a single consumable wall node does not exceed a preset consumable wall concurrency threshold.

[0125] Optionally, the dynamic efficiency analysis strategy in the node analysis module 302 is specifically used to: According to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, a plurality of pre-selected target vertical pipelines connected to the floor where the demand department corresponding to the current delivery task is located are screened; Analyze the plurality of pre-selected target vertical pipelines according to the starting node corresponding to the current delivery task in the starting delivery node information set, and select the delivery entrance of the pre-selected target vertical pipeline closest to the target medicine collection node among the plurality of pre-selected target vertical pipelines as the target delivery entrance; According to the target delivery inlet, the delivery outlet corresponding to the other end of the preselected target vertical pipeline is used as the target delivery outlet; Based on the target delivery exit, analyzing the real-time position coordinate set of the robot, and selecting the delivery robot closest to the target delivery exit on the floor where the target delivery exit is located as the target delivery object; The delivery node information set is generated based on the target delivery entrance, the target delivery exit, and the target handover object corresponding to each delivery task.

[0126] Optionally, the plan execution module 303 is specifically configured to: Planning a delivery route corresponding to each delivery task according to the target drug pickup node information set, the starting delivery node information set, and the handover delivery node information set; The overlapping features of the delivery routes corresponding to different delivery tasks on different floors are analyzed, and the delivery tasks corresponding to the overlapping delivery routes on the same floor are partially merged to achieve the dynamic delivery task merging.

[0127] The system of this embodiment can be used to execute the method of any of the above embodiments. Its implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A method for intelligent distribution of medical consumables, characterized in that: include: Obtaining a consumables demand information set, and generating a delivery task queue based on the consumables demand information set; Acquire a real-time device node status information set, and determine a target drug pickup node information set, a starting delivery node information set, and a handover delivery node information set based on the real-time device node status information set and the delivery task queue; According to the target medicine collection node information set, the starting delivery node information set and the handover delivery node information set, dynamic delivery task merging is performed, the consumables delivery plan is determined and executed, and the consumables delivery report is determined and output.

2. The method according to claim 1, characterized in that The consumables demand information set includes a demand department location code, a consumables type identifier, and a demand urgency coefficient; The consumables demand information set is obtained by an OCR medical prescription recognition unit for extracting medical consumables text keywords in cooperation with a consumables demand voice recognition unit for extracting medical consumables voice keywords; The OCR medical prescription recognition unit and the consumables demand voice recognition unit are both integrated into the hospital information management system.

3. The method according to claim 2, characterized in that Generating a delivery task queue according to the consumables demand information set includes: Normalizing the demand urgency coefficient, assigning a delivery task label to each delivery task based on the normalization result, and using the corresponding department location code as a unique task identifier corresponding to each delivery task; The delivery task labels include scheduled delivery tasks, regular delivery tasks and emergency delivery tasks; Identify the over-limit consumables according to the consumable type identifier, decompose the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, and assign the same task unique identifier to the plurality of parallel subtasks; The emergency delivery task is inserted into the head of the delivery task queue based on the task initiation time, the regular delivery task is connected to the emergency delivery task, and the scheduled delivery task is added to the tail of the delivery task queue based on the task reservation time to construct the delivery task queue.

4. The method according to claim 3, characterized in that The method further includes identifying the over-limit consumables according to the consumable type identifier and decomposing the corresponding delivery task into a plurality of parallel subtasks according to the over-limit consumables, including: The over-limit consumables are medical consumables that cannot be delivered by a single delivery robot in a single delivery mission; Determining the total weight and total space volume of the medical consumables required for the current delivery task based on the consumable type identifier; Comparing the total weight of the consumables and the total volume of the consumable space with the rated loading weight and rated loading capacity of the delivery robot, respectively; if the total weight of the consumables or the total volume of the consumable space exceeds the rated loading weight or the rated loading capacity, determining the medical consumables required for the current delivery task as the over-limit consumables; Determining, based on the consumable type identifier, a number of independently packaged units that can be assembled corresponding to the over-limit consumable, each of which does not exceed the rated loading weight and the rated loading volume; According to the plurality of said assemblable independent packaging units, the distribution task corresponding to the over-limit consumables is divided into a plurality of said parallel sub-tasks, and the total weight and total volume of the plurality of said assemblable independent packaging units corresponding to each of said parallel sub-tasks do not exceed the rated loading weight and the rated loading capacity.

5. The method according to claim 4, characterized in that The real-time device node status information set includes floor vertical pipeline topology data, robot real-time position coordinate set and consumable wall inventory capacity status data set; The vertical pipeline topology data is three-dimensional structural data that describes the spatial layout of vertical transmission pipelines used for handing over medical consumables between floors and the real-time status of layered isolation valves.

6. The method according to claim 5, characterized in that The method of determining a target medicine pickup node information set, a starting delivery node information set, and a handover delivery node information set based on the real-time device node status information set and the delivery task queue includes: Through the task allocation optimization strategy, according to the department location code and consumable type identifier of each task in the delivery task queue, combined with the consumable wall inventory capacity status data set, the optimal consumable wall node corresponding to each delivery task is matched as the target drug collection node, and the target drug collection node information set is constructed; Using a shortest path optimization algorithm, based on the robot's real-time position coordinate set and the demand urgency coefficient, the shortest path from each delivery robot to the corresponding target medication pickup node is matched, the starting point corresponding to each shortest path is selected, and the starting delivery node information set is constructed; Through a dynamic efficiency analysis strategy, according to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, the corresponding optimal transfer node is selected from the intersection points of each floor of the corresponding vertical pipeline to construct the handover and distribution node information set.

7. The method according to claim 6, characterized in that The task allocation optimization strategy includes: Constructing a two-dimensional distribution benefit matrix with the consumable wall nodes as rows and the distribution tasks as columns based on the consumable wall inventory capacity status dataset and the distribution task queue; The matrix element values ​​in the two-dimensional distribution effect matrix are obtained by weighted evaluation of the estimated transportation time and inventory fulfillment rate between the consumables wall node and the corresponding demand department; Solving the two-dimensional distribution benefit matrix using the Hungarian algorithm with distribution constraints to determine the optimal consumables wall node corresponding to each distribution task; The delivery constraint condition is that the number of delivery tasks served simultaneously by a single consumable wall node does not exceed a preset consumable wall concurrency threshold.

8. The method according to claim 6, characterized in that The dynamic efficiency analysis strategy includes: According to the real-time status of the layered isolation valve in the floor vertical pipeline topology data, a plurality of pre-selected target vertical pipelines connected to the floor where the demand department corresponding to the current delivery task is located are screened; Analyze the plurality of pre-selected target vertical pipelines according to the starting node corresponding to the current delivery task in the starting delivery node information set, and select the delivery entrance of the pre-selected target vertical pipeline closest to the target medicine collection node among the plurality of pre-selected target vertical pipelines as the target delivery entrance; According to the target delivery inlet, the delivery outlet corresponding to the other end of the preselected target vertical pipeline is used as the target delivery outlet; Based on the target delivery exit, analyzing the real-time position coordinate set of the robot, and selecting the delivery robot closest to the target delivery exit on the floor where the target delivery exit is located as the target delivery object; The delivery node information set is generated based on the target delivery entrance, the target delivery exit, and the target handover object corresponding to each delivery task.

9. The method according to claim 8, characterized in that The dynamic delivery task merging according to the target medicine pickup node information set, the starting delivery node information set, and the handover delivery node information set includes: Planning a delivery route corresponding to each delivery task according to the target drug pickup node information set, the starting delivery node information set, and the handover delivery node information set; The overlapping features of the delivery routes corresponding to different delivery tasks on different floors are analyzed, and the delivery tasks corresponding to the overlapping delivery routes on the same floor are partially merged to achieve the dynamic delivery task merging.

10. An intelligent distribution system for medical consumables, characterized in that: include: A task analysis module is used to obtain a consumables demand information set and generate a delivery task queue based on the consumables demand information set; A node analysis module is used to obtain a real-time device node status information set, and based on the real-time device node status information set and the delivery task queue, determine a target drug collection node information set, a starting delivery node information set, and a handover delivery node information set; The plan execution module is used to dynamically merge delivery tasks based on the target drug collection node information set, the starting delivery node information set and the handover delivery node information set, determine and execute the consumables delivery plan, and output the consumables delivery report.

Images