A medical supply chain order scheduling method and system

Abstract

The application relates to the technical fields of warehouse logistics and data processing, and discloses a medical material supply chain order scheduling method and system. The method comprises the following steps: obtaining an order generation instruction of medical materials and extracting a target material identifier and a required quantity; mapping a preset physical inventory database according to the target material identifier to obtain a material single product set; performing elimination on the material single product set to obtain a first candidate set, and then performing reordering and intercepting to obtain a second candidate set; extracting temperature control requirements according to traceability codes in the second candidate set and performing grouping to obtain temperature control carpooling groups; calculating the temperature zone overlap degree between the vehicle temperature zone characteristics of a to-be-scheduled vehicle and the corresponding temperature control requirements of the temperature control carpooling groups, and combining the packaging volume to perform combination calculation to obtain a loading carpooling scheme; and calculating the physical grabbing strength of a mechanical arm according to the material weight of the material single products in the loading carpooling scheme to generate a warehouse picking instruction. The method can realize full-link closed-loop order scheduling.

Description

Technical Field

[0001] This invention relates to the fields of warehousing logistics and data processing technology, and in particular to a method and system for scheduling medical supply chain orders. Background Technology

[0002] Currently, the precise and timely flow of medical supplies is crucial for ensuring the quality of medical services. As modern warehousing evolves towards automation and intelligence, the scale of time-series characteristic data and status attribute data generated throughout the entire lifecycle of supplies is growing explosively. How to achieve precise scheduling at the individual item level through efficient big data management has become a focus of industry attention.

[0003] In existing technologies, conventional warehousing and transportation scheduling solutions typically rely on macro-level batch ledgers for information recording and instruction issuance. When the system receives a transfer instruction, it can usually only send macro-level management instructions to the warehouse based on broad quantity indicators. Because the scheduling system cannot capture and map the differences in physical attributes, shelf life, and specific storage constraints of each individual material item in real time during the order generation stage, the system cannot perform strict compliance quality filtering in subsequent screening stages. This lack of awareness of individual item attributes further spreads to peripheral transportation and physical execution stages, making it impossible to intelligently group and combine scattered materials with strict temperature control requirements when leaving the warehouse. This not only makes it difficult to accurately match with the actual temperature control characteristics of transport vehicles, but also makes it impossible to directly link with the underlying automated execution mechanisms to achieve precise physical grasping and path coordination.

[0004] Existing technologies suffer from a lack of end-to-end closed-loop coordination in outbound material scheduling, including single-item compliance management, multi-temperature zone cold chain consolidation and loading, and automated picking execution. Summary of the Invention

[0005] This invention provides a method and system for medical supply chain order scheduling to address the lack of end-to-end closed-loop coordination in existing technologies regarding outbound material scheduling, including single-item compliance management, multi-temperature zone cold chain grouping and loading, and automated picking execution.

[0006] Firstly, in order to solve the above-mentioned technical problems, the present invention provides a method for scheduling medical supply chain orders, comprising: Obtain the order generation instruction for medical supplies, extract the target material identifier and required quantity from the order generation instruction; map the target material identifier to a preset physical inventory database to obtain a set of individual material items containing serial number, production date, damage level, traceability code, packaging volume and material weight; Remove material items from the material item set whose production date is earlier than a preset expiration date threshold or whose damage level is greater than a preset damage threshold to obtain the first candidate set; The first candidate set is sorted according to the production date to obtain a time-series arrangement set, and the time-series arrangement set is truncated according to the demand quantity to obtain a second candidate set; Temperature control requirements are extracted from the traceability codes in the second candidate set, and the individual items in the second candidate set are grouped using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups. Obtain the vehicle temperature zone characteristics of the vehicles to be dispatched, calculate the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group, and combine the temperature control carpooling group whose temperature zone overlap meets the preset matching conditions with the packaging volume to obtain the loading carpooling scheme. Extract the serial number and weight of the individual materials in the loading and carpooling plan, calculate the physical grasping force of the robotic arm based on the weight of the materials, and generate a warehouse picking instruction based on the serial number.

[0007] In a second aspect, the present invention provides a medical supplies supply chain order scheduling system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0008] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention obtains a set of individual materials containing production dates and damage levels, compares the production dates with a preset expiration date threshold over a time span, and determines the damage level against a preset damage threshold, thereby accurately removing non-compliant individual materials before generating scheduling instructions. This single-item-level physical attribute hard filtering mechanism introduced at the forefront of the scheduling chain completely breaks the limitation of traditional scheduling systems that rely solely on macro-batch ledgers to issue broad quantity instructions, enabling the system to perceive and accurately identify physical objects with near-expiration risks and physical defects in real time. This fundamentally eliminates the hidden dangers of near-expiration or slightly damaged materials being blindly mixed into the warehouse or even left to be scrapped for a long time, and significantly improves the quality, safety, compliance, and single-item-level lifecycle management accuracy of outbound medical supplies.

[0009] (2) This invention extracts absolute temperature control requirements from the traceability codes of the candidate set, uses these as matching keys to extract target items for clustering, and further calculates the temperature zone overlap between the temperature zone characteristics of the vehicle to be dispatched and the temperature control requirements corresponding to the temperature control group. Finally, it combines the grouping that meets the intersection-union-ratio (IUU) condition with the packaging volume to generate a three-dimensional loading and carpooling scheme. This mechanism compares and optimizes the underlying temperature control constraints of multi-source scattered materials with the multi-temperature zone attributes of real physical vehicles with high confidence, avoiding random and indiscriminate carpooling of materials with huge differences in temperature control characteristics. This completely solves the problem of cold chain materials being mistakenly loaded into normal temperature vehicles or causing failure and chain breakage due to temperature zone mismatch, greatly reduces the loss rate of temperature-controlled materials in the pre-transportation stage, and significantly improves the resource loading and utilization efficiency of multi-temperature zone refrigerated vehicles.

[0010] (3) This invention extracts the actual weight characteristics of individual items in the final loading and carpooling scheme, inputs them into a pre-built dynamic calculation model to accurately calculate the physical grasping force required at the end of the robotic arm, and combines this with the automated guided vehicle path generated by the real-time warehouse topology map to uniformly encapsulate it into picking instructions. This hardware-software collaborative architecture directly transmits the flexible carpooling scheduling logic of the upper-layer data flow to the rigid physical execution link of the lower-layer warehouse hardware, enabling the automated robotic arm to adaptively calculate and apply absolutely precise target normal force when facing individual items of different weights and specifications. This achieves seamless closed-loop linkage from virtual order data flow to real physical control flow, effectively eliminating the risk of material drop or crushing damage caused by grasping force mismatch, and significantly improving the execution robustness and global physical collaborative operation efficiency of the automated warehousing system in complex environments. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the medical supplies supply chain order scheduling method provided in the first embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] Reference Figure 1 The first embodiment of the present invention provides a method for scheduling medical supply chain orders, including the following steps: S1, obtain the order generation instruction for medical supplies, extract the target material identifier and required quantity from the order generation instruction; map the target material identifier to a preset physical inventory database to obtain a set of individual material items containing serial number, production date, damage level, traceability code, packaging volume and material weight; S2, remove material items from the material item set whose production date is earlier than a preset expiration date threshold or whose damage level is greater than a preset damage threshold, to obtain the first candidate set; S3, sort the first candidate set according to the production date to obtain a time-series arrangement set, and truncate the time-series arrangement set according to the demand quantity to obtain a second candidate set; S4. Extract temperature control requirements based on the traceability codes in the second candidate set, and group the individual materials in the second candidate set using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups. S5, obtain the vehicle temperature zone characteristics of the vehicle to be dispatched, calculate the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group, and combine the temperature control carpooling group whose temperature zone overlap meets the preset matching conditions with the packaging volume to obtain the loading carpooling scheme. S6, extract the serial number and weight of the material item in the loading and carpooling plan, calculate the physical grasping force of the robotic arm based on the weight of the material, and generate a warehouse picking instruction based on the serial number.

[0014] In step S1, an order generation instruction for medical supplies is obtained, and the target material identifier and required quantity are extracted from the order generation instruction. Based on the target material identifier, a preset physical inventory database is mapped to obtain a set of individual material items containing serial number, production date, damage level, traceability code, packaging volume, and material weight.

[0015] This includes obtaining an order generation instruction for medical supplies, and extracting the target material identifier and required quantity from the order generation instruction.

[0016] In one implementation, this embodiment receives an encrypted byte stream sequence through the listening port of a distributed message queue, decrypts the encrypted byte stream sequence using an asymmetric encryption algorithm, outputs plaintext text data, and identifies the plaintext text data as an order generation instruction for medical supplies.

[0017] It should be noted that the asymmetric encryption algorithm used is RSA encryption. Specifically, the private key pre-configured in the local server's memory is extracted, and the encrypted byte stream sequence is subjected to a large integer modulo exponentiation operation using the private key to calculate and output the plaintext data.

[0018] Further, a deserialization operation is performed on the plaintext data to construct a corresponding abstract syntax tree. A depth-first search algorithm is used to traverse all nodes of the abstract syntax tree, matching preset material identifier keys with preset demand quantity keys. The string value of the node corresponding to the material identifier key is extracted and identified as the target material identifier; the integer value of the node corresponding to the demand quantity key is extracted and identified as the demand quantity.

[0019] It is worth noting that the preset material identification key name and the preset demand quantity key name are determined by reading a pre-established interface communication protocol file; this interface communication protocol file specifies the data field mapping dictionary that external business systems must follow when interacting with the current scheduling system.

[0020] For example, an encrypted hexadecimal byte stream is received through the listening port. A pre-configured 2048-bit private key is extracted, and its decryption is performed using modular exponentiation to obtain plaintext data in JSON format, which serves as the order generation instruction. The system deserializes this order generation instruction to construct an abstract syntax tree, traversing each node in depth-first order. It successfully matches the preset material identifier key name Target_Material_ID with the preset demand quantity key name Demand_Quantity. The numerical values ​​associated with these keys are extracted, ultimately yielding the target material identifier as MED-10086 and the demand quantity as 200.

[0021] Specifically, based on the target material identifier mapping to a preset physical inventory database, a set of individual material items containing serial number, production date, damage level, traceability code, packaging volume, and material weight is obtained.

[0022] In one implementation, the extracted target material identifier is configured as a hash index key, automatically triggering a memory-based retrieval mapping mechanism. A preset physical inventory database is accessed concurrently through the hash index structure. Multi-table joins and conditional matching operations are performed on the core entity data tables of the materials contained in the preset physical inventory database. The underlying static attribute row data associated with the target material identifier is retrieved, and a set of individual material items containing serial number, production date, damage level, traceability code, packaging volume, and material weight is output.

[0023] It should be noted that the pre-built physical inventory database is constructed in advance by collecting full physical features using RFID reading arrays and 3D laser scanning equipment during the material receiving process, and by combining this with daily dynamic feature updates using warehouse inspection robots. Specifically, the damage level is determined during the material receiving inspection and patrol stages by using a pre-trained defect image classification model to perform feature recognition on real-time surface images of the materials. After recognition, the information is dynamically written to and updated in the classification field of the material core entity data table, thereby avoiding the omission of new damage occurring during storage.

[0024] It is worth noting that the defect image classification model adopts the general ResNet-50 residual network architecture. Its offline training process involves collecting 5000 images of medical supply packaging labeled as intact and with varying degrees of damage as the training set, with 80% used for training and 20% for validation. The system sets numerical mapping rules for damage levels according to common packaging industry standards: a value of 1 represents intact packaging, a value of 2 represents slight dents that do not affect the interior, and a value of 3 represents severe damage; the values ​​are positively correlated with the degree of damage. Training uses the image pixel matrix as model input, employs the Adam optimizer, sets the initial learning rate to 0.001, sets the batch size to 32, and uses cross-entropy as the loss function for iterative fitting. When the classification accuracy on the validation set no longer improves after 5 consecutive training epochs, training stops and the model weights are saved, resulting in the defect image classification model.

[0025] For example, the system uses the target material identifier MED-10086 as the key-value input hash index to match and query the preset physical inventory database. After aggregation calculation through the associated tables, it retrieves and outputs a set of individual materials containing the serial number, production date of September 1, 2023, damage level of 2 as recently updated by the inspection robot, traceability code generated by a secure hash algorithm, packaging volume of 0.015 cubic meters, and material weight of 25.6 kilograms.

[0026] In step S2, items whose production date is earlier than a preset expiration date threshold or whose damage level is greater than a preset damage threshold are removed from the set of individual items, resulting in a first candidate set, including: Extract the production date and damage level of each material item from the material item set; The production date is converted into a standard timestamp, and the standard timestamp is compared with the preset expiration date threshold over a time span to obtain the comparison result; The damage level is compared with the preset damage threshold to determine the numerical value, and the determination result is obtained. Based on the comparison results and the determination results, material items whose production date exceeds a preset expiration date threshold or whose damage level exceeds a preset damage threshold are removed from the material item set to obtain a first candidate set.

[0027] In one implementation, the system pre-configures expiration date and damage thresholds, and directly reads and performs numerical comparisons at runtime. The system uses an object-relational mapping interface to extract the production date and damage level of each material item from the material item set. The production date, in string format, is converted to a standard timestamp in milliseconds, and this timestamp is compared with the preset expiration date threshold over a given time span. Simultaneously, the damage level, which is an integer value, is compared with the preset damage threshold. Based on the comparison and determination results, if the standard timestamp is less than the preset expiration date threshold (i.e., the production time is earlier than the historical critical time, resulting in an excessive in-stock time), or if the damage level is higher than the preset damage threshold, the corresponding non-compliant material item is removed from the material item set. The remaining compliant items are then reassembled in the system's cache to obtain a first candidate set.

[0028] It should be noted that the preset expiration date threshold is a historical critical time stamp obtained by subtracting the preset maximum allowable storage time from the current system running timestamp and moving back along the timeline. The maximum allowable storage time is determined by subtracting the minimum safe outbound circulation days from the general shelf life of this type of material. For conventional medical supplies, minor dents on the outer packaging are permissible as long as they do not affect the internal quality. In this embodiment, the preset damage threshold is reasonably set to a value of 2, meaning only items with a damage level of 3 (severe damage) are excluded, balancing safety with reducing unnecessary scrap costs.

[0029] For example, assuming a certain medical supply has a general shelf life of 180 days and a preset safe outbound circulation period of 10 days, the maximum allowable storage time is 170 days. Assuming the system's current running time is October 15, 2023, subtracting 170 days from the past yields the timestamp corresponding to the preset expiration date threshold, which is April 28, 2023. The system extracts the production date of a certain item as September 1, 2023, converts it to a standard timestamp, and compares it with the preset expiration date threshold. The result shows that the standard timestamp of the production date is greater than the preset expiration date threshold, meaning the item is sufficiently fresh and has not expired. Simultaneously, the extracted damage level value 2 is compared with the preset damage threshold 2. The result shows that the current level is not higher than the preset threshold. Therefore, this item is deemed compliant and retained. Ultimately, the system obtains a first candidate set containing 520 compliant items.

[0030] In step S3, the first candidate set is sorted according to the production date to obtain a time-series arrangement set. The time-series arrangement set is then truncated according to the demand quantity to obtain a second candidate set, including: Extract the production dates of the individual materials in the first candidate set, and sort the production dates in ascending order to obtain a time-series arrangement set; The truncation length is determined based on the required quantity, and a truncation operation is performed on the head position of the time sequence arrangement set according to the truncation length to obtain the second candidate set.

[0031] In one implementation, the sorting management unit retrieves the production dates associated with the individual materials in the first candidate set and loads them into a memory array. A bubble sort algorithm is used, employing a double loop structure to compare the production dates of adjacent individual materials in the memory array. If the production date timestamp of a preceding node is greater than that of a subsequent node, a memory address swap operation is triggered, moving the earlier-date material to the front. The first candidate set is then sorted in ascending order, and a time-ordered set is output.

[0032] Furthermore, the truncation control unit establishes the demand quantity extracted in step S1 as the upper limit index for truncation of the one-dimensional array. It calls the array slicing processing function, determines the truncation length according to the value corresponding to the demand quantity, and performs the truncation operation from the starting index zero of the temporal arrangement set backwards at the head position. The truncated specified number of material item objects are repackaged and written into the high-speed buffer queue to obtain the second candidate set.

[0033] For example, the sorting management unit sorts the timestamps of 520 individual items in ascending order, outputting a time-series arrangement set strictly arranged from earliest to latest. The truncation control unit determines the truncation length based on the required quantity value of 200, uses an array slicing function to truncate the individual items with indices 0 to 199, extracts 200 material objects, and encapsulates them to obtain the second candidate set.

[0034] In step S4, temperature control requirements are extracted based on the traceability codes in the second candidate set, and the individual materials in the second candidate set are grouped using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups.

[0035] Among them, extracting temperature control requirements based on the traceability codes in the second candidate set includes: Use the traceability code in the second candidate set as the key value to query the preset supplier master database and obtain the corresponding temperature control identification status code; If the temperature control status code exists, the temperature control status code is parsed to extract the corresponding lower temperature limit value and upper temperature limit value. If the temperature control status code does not exist, extract the preset upper and lower limits of the normal temperature range; The extracted upper and lower limit values ​​are combined to construct a temperature range vector, and the temperature range vector is determined as the temperature control requirement.

[0036] In one implementation, the attribute parsing engine concurrently extracts the traceability code contained in each of the material items from the second candidate set. Using the traceability code as a hash retrieval key, it concurrently accesses a preset supplier master database, retrieves matching material dictionary fields, and obtains the corresponding temperature control status code. For material items with the temperature control status code, it performs string pattern matching and parsing processing on the temperature control status code to extract the corresponding lower temperature limit and upper temperature limit values.

[0037] Specifically, first, locate the left and right square brackets in the temperature control status code, and extract the character content between them as the temperature range string. Second, using the comma in the temperature range string as a separator, divide the string into a left substring and a right substring. Remove the leading and trailing whitespace characters from the left and right substrings. If a substring starts with a minus sign, the corresponding temperature value is determined to be negative. Finally, parse the left substring into the lower temperature limit value and the right substring into the upper temperature limit value. If parsing fails or the substring is empty, the status code format is determined to be abnormal.

[0038] It is worth noting that for regular materials for which the aforementioned temperature control status code is not found, the system treats them as room temperature materials and automatically extracts the pre-set upper and lower limits of the room temperature range in the system. In this embodiment, based on the national pharmacopoeia's general standards for room temperature storage environments, the lower limit is set at 15.0 degrees Celsius and the upper limit at 30.0 degrees Celsius. The system uses the extracted upper and lower limits as independent components of a two-dimensional feature vector, concatenates them axially to construct a two-dimensional temperature range vector, and uniformly determines it as the temperature control requirement. This default parameter assignment mechanism allows room temperature materials to be seamlessly integrated into the subsequent temperature zone calculation pipeline.

[0039] Specifically, the individual items in the second candidate set are grouped using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups, including: Extract the temperature interval vector corresponding to the individual material items in the second candidate set; Clustering trials are performed on each temperature interval vector according to the preset number of candidate clusters to obtain the temporary clustering architecture corresponding to each number of candidate clusters, and the global contour coefficient of each temporary clustering architecture is calculated. The number of candidate clusters that maximizes the global contour coefficient is determined as the target number of clusters. The K-means clustering algorithm is used to perform convergent iterative calculations on each of the temperature interval vectors based on the target cluster number, thereby obtaining multiple target material clusters corresponding to the target cluster number; Based on the multiple target material clusters, the individual material items are categorized to obtain temperature-controlled carpooling groups.

[0040] In one implementation, the data clustering unit extracts the temperature interval vectors corresponding to all individual materials in the second candidate set. A preset set of K values ​​containing multiple integers is established. For each K value in this set, the algorithm execution mechanism runs the K-means clustering algorithm. After each run, the silhouette coefficient of the overall clustering result is calculated. By comparison, the K value that maximizes the overall silhouette coefficient is identified and determined as the target number of clusters. In the multidimensional geometric space, the Euclidean distance between each temperature interval vector and the initialized centroid of each cluster is iteratively calculated. According to the principle of minimizing the Euclidean distance, each temperature interval vector is divided into multiple clusters corresponding to the target number of clusters. Convergence is determined when the relative offset of the cluster centroid position coordinates between two adjacent iterations is less than 0.001. Based on the final cluster division, the individual materials are spatially classified, and a unique carpooling batch code is generated for each classification set, resulting in temperature-controlled carpooling groups.

[0041] It should be noted that the preset supplier master database is constructed in the following way: the system periodically calls the public data service platform of the National Medical Products Administration through the API interface, or imports the "Cold Chain Transportation Temperature Control Capability Filing Form" provided by the supplier through SFTP, and parses and stores the temperature control identification status code of the materials and their corresponding temperature range.

[0042] For example, the system performs traceability code queries on 200 individual items. For 45 items, the system obtains a lower temperature limit of 2.0 and an upper temperature limit of 8.0, constructing a temperature range vector [2.0, 8.0]. Simultaneously, the system finds that the remaining 155 items lack specific temperature control identifiers and automatically assigns them a normal temperature range vector [15.0, 30.0]. Finally, after multiple rounds of K-means calculations, the overall profile coefficient is maximized when K=3, thus the system accurately divides these 200 items into three temperature-controlled groups: cryogenic, refrigerated, and normal temperature.

[0043] In step S5, the vehicle temperature zone characteristics of the vehicles to be dispatched are obtained, the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group is calculated, and the temperature control carpooling group whose temperature zone overlap meets the preset matching conditions is combined with the packaging volume to obtain the loading carpooling plan.

[0044] This includes obtaining the vehicle temperature zone characteristics of the vehicles to be dispatched, and calculating the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature-controlled carpooling group, including: The structured data messages returned by the vehicles to be dispatched are parsed, and the upper and lower critical temperatures supported by the cabin of the vehicles to be dispatched are extracted as the temperature zone characteristics of the vehicles. The area crossover ratio between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group is calculated to obtain the temperature zone overlap.

[0045] In one implementation, the IoT communication gateway receives structured data packets transmitted in real time from the on-board equipment in the vehicle to be dispatched via network transmission. The data parsing unit uses an entity extraction model to parse the structured data packets, extracting the upper and lower critical temperature values ​​supported by each independent compartment of the vehicle to be dispatched, as vehicle temperature zone features. The dispatch engine calls the intersection-union algorithm to calculate the boundary overlap length between the vehicle temperature zone features of each independent compartment and the temperature control requirements corresponding to the temperature-controlled carpooling group. The boundary overlap length is divided by the union length of their one-dimensional continuous intervals to calculate the area overlap ratio corresponding to each independent compartment, and this area overlap ratio is determined as the temperature zone overlap degree.

[0046] The loading and unloading scheme is obtained by combining the temperature-controlled groupings whose temperature zone overlap meets a preset matching condition with the packaging volume, and calculating the combination. This includes: Target carpooling groups with a temperature zone overlap greater than a preset crossover ratio threshold are selected from the temperature-controlled carpooling groups. Obtain the cabin size data of the vehicle to be dispatched, and combine the packaging volume with the cabin size data to calculate the spatial location allocation of the target carpooling group to obtain the three-dimensional loading layout; Obtain the current platform occupancy status data, and assign a parking platform number to the vehicle to be dispatched based on the current platform occupancy status data; The loading and carpooling plan is obtained by combining the information of the three-dimensional loading layout, the docking platform number, and the material list corresponding to the target carpooling group.

[0047] It should be noted that for vehicles transporting goods at ambient temperature, the ambient temperature returned by their onboard sensors, such as 20.0 degrees Celsius, will fall perfectly within the ambient temperature range assigned by the system, namely 15.0 degrees Celsius to 30.0 degrees Celsius. This will naturally result in a very high degree of temperature overlap, enabling automatic matching calculations between vehicles transporting goods at ambient temperature and groups of goods transported at ambient temperature.

[0048] Furthermore, the logic filtering unit compares the temperature zone overlap of each temperature-controlled carpooling group with a preset crossover ratio (CBR) threshold. It then establishes a mapping relationship between temperature-controlled carpooling groups whose temperature zone overlap is greater than the preset CBR threshold and their corresponding independent compartments, filtering and extracting them as target carpooling groups. Considering both cold chain safety margin and vehicle loading resource utilization, the reasonable range for the preset CBR threshold is 0.85 to 0.95; in this embodiment, it is preferably set to 0.95. If the temperature zone overlap of all independent compartments of a vehicle is not greater than the preset CBR threshold, the system marks the temperature-controlled carpooling group as suspended and automatically generates an alarm work order to be pushed to the manual dispatch terminal. When the IoT gateway detects a new multi-temperature zone vehicle entering the site for registration, the system automatically wakes up the suspended temperature-controlled carpooling group and re-triggers the matching calculation.

[0049] Furthermore, the spatial planning unit obtains the actual compartment size data of the successfully matched specific independent compartments. A 3D packing algorithm based on the bottom-left heuristic rule is introduced, and combined with the packaging volume and compartment size data, the spatial location allocation of the target carpooling group within the corresponding independent compartment is calculated, outputting the 3D loading grid coordinates to obtain the 3D loading layout. Simultaneously, the resource allocation unit retrieves the current platform occupancy status data, uses a greedy optimization operation to traverse the physical platform position status matrix, selects the physical platform code with the smallest index and an idle status, and establishes it as the assigned docking platform number. The current platform occupancy status data is established through real-time monitoring by an infrared sensor array deployed at the outbound platform passage. The data encapsulation module performs data message splicing and combination for the 3D loading layout of multiple compartments, the docking platform number, and the material list corresponding to the target carpooling group to obtain the loading carpooling scheme.

[0050] For example, the area overlap ratio between the first compartment feature of a cold chain vehicle and a specific carpooling group is calculated to be 0.96, which is determined as the temperature zone overlap. The logic filtering unit filters out target carpooling groups with an overlap ratio greater than a preset intersection-union threshold of 0.95. Combining the independent compartment size data, the optimal three-dimensional loading layout is calculated using a three-dimensional packing algorithm. The vehicle is automatically assigned to docking platform 08, and the loading carpooling scheme is combined, packaged, and output.

[0051] In step S6, the serial number and weight of the material items in the loading and carpooling plan are extracted, the physical grasping force of the robotic arm is calculated based on the weight of the material, and a warehouse picking instruction is generated in combination with the serial number.

[0052] The calculation of the robotic arm's physical gripping force based on the weight of the material includes: Input the weight of the material into a pre-built dynamic calculation model; The target normal force required for the robotic arm's end effector to overcome gravity and inertia is calculated using the aforementioned dynamic calculation model. The target normal force is defined as the physical gripping force of the robotic arm.

[0053] The process of generating warehouse picking instructions based on the serial number includes: The serial number is subjected to anti-counterfeiting obfuscation calculation to obtain an anti-counterfeiting traceability verification code, and the verification bit of the anti-counterfeiting traceability verification code is calculated to obtain a picking verification bit; Obtain a real-time warehouse topology map, and perform node search and obstacle avoidance optimization calculations based on the real-time warehouse topology map to obtain the path of the automated guided vehicle; The anti-counterfeiting and traceability verification code, the picking verification digit, the automated guided vehicle path, and the physical grasping force of the robotic arm are combined to generate a warehouse picking instruction.

[0054] In one implementation, the control unit performs a depth-first traversal of the material nodes included in the loading and carpooling scheme based on a tree topology parsing algorithm to accurately extract the serial number and weight of the corresponding material item. The mechanics calculation mechanism inputs the material weight into a pre-constructed dynamics calculation model. The dynamics calculation model is implemented using rigid body mechanics formulas, the specific calculation formulas of which are shown below:

[0055] in, Indicates the target normal force; This indicates the weight of the material as reported by a high-precision sensor. This represents the constant of gravitational acceleration; in this embodiment, we take... ; The system dynamically matches a safe acceleration based on a preset maximum acceleration for the robotic arm, according to a database of material fragility levels; for example, it matches a safe acceleration for glass products. Standard cardboard box matching ; The system dynamically obtains the preset static friction coefficient between the gripper and the packaging by parsing the order's material attributes and consulting a preset friction coefficient table. For example, the coefficient is 0.5 for cardboard boxes and rubber grippers, and 0.3 for aluminum foil. The mechanical calculation mechanism calculates the target normal force required by the robotic arm's end effector by substituting these dynamic parameters into a formula and performing algebraic operations. This target normal force is initially determined as the robotic arm's physical gripping force. Furthermore, during actual execution, real-time contact feedback is obtained through a six-axis torque sensor built into the end effector. If a small amount of slippage is detected, a PID closed-loop control algorithm is used to perform real-time adaptive compensation and calibration of the gripping force.

[0056] Further, an anti-counterfeiting obfuscation calculation is performed on the serial number to obtain an anti-counterfeiting traceability verification code, and a check bit calculation is performed on the anti-counterfeiting traceability verification code to obtain a picking verification check bit. Specifically, the anti-counterfeiting obfuscation calculation uses the SHA-256 secure hash algorithm, injects a system-preset random obfuscation salt value into the beginning and end of the serial number to perform a hash operation, and outputs a fixed-length feature string as the anti-counterfeiting traceability verification code. A real-time warehouse topology map is obtained, and node search and obstacle avoidance optimization calculations are performed in conjunction with the real-time warehouse topology map to obtain the automated guided vehicle (AGV) path; the anti-counterfeiting traceability verification code, the picking verification check bit, the AGV path, and the physical grasping force of the robotic arm are combined to generate a warehouse picking instruction.

[0057] It should be noted that the node search and obstacle avoidance optimization calculation adopts the A* pathfinding algorithm, and its evaluation function consists of the actual movement cost and the estimated remaining cost. The estimated remaining cost is determined based on the Manhattan distance between the current node and the target node.

[0058] For example, the system extracts the serial number of the item to be shipped from the loading and carpooling plan as 94725, and the item's weight as 25.6 kg. The system looks up the table and finds that the item is packaged in a cardboard box and is not fragile, matching the parameters... and Substituting into the dynamic calculation formula, the target normal force is calculated to be 604.16 Newtons, that is... This force is determined as the physical gripping force of the robotic arm. The serial number is subjected to salted SHA-256 hashing and CRC checksum calculations to obtain the picking verification bit. Combined with the topology map, obstacle avoidance and optimization calculations are performed to obtain the optimal automated guided vehicle (AGV) path. Finally, these are combined and encapsulated to generate the warehouse picking instruction.

[0059] In summary, this invention introduces a single-item-level physical attribute mapping and filtering mechanism to accurately eliminate near-expiry and damaged risk materials at the scheduling source and establish strict time-series outbound queues. Furthermore, it innovatively constructs a unified temperature-controlled scheduling pipeline covering all categories of materials. Utilizing unsupervised clustering algorithms and spatial intersection-union ratio calculations, it performs high-confidence matching and three-dimensional packing combination of the underlying temperature control constraints of massive amounts of scattered materials with the physical attributes of vehicles in multiple temperature zones. Finally, it precisely integrates the upper-level virtual scheduling scheme with the underlying physical execution hardware, combining rigid body mechanics models to adaptively calculate microscopic grasping strength, and superimposing cryptographic anti-counterfeiting verification and heuristic obstacle avoidance and pathfinding to generate global collaborative instructions. This invention achieves a fundamental leap in medical supply outbound scheduling from macro-quantity assignment to a single-item-level software-hardware collaborative closed loop, completely solving the technical bottlenecks of lack of quality compliance control and cold chain temperature zone mismatch and breakage in traditional scheduling. It significantly reduces the failure and loss rate of high-value sensitive materials in the pre-transfer stage and significantly improves the robustness of underlying hardware execution and the overall logistics resource allocation efficiency in complex automated warehousing environments.

[0060] The second embodiment of the present invention provides a medical supplies supply chain order scheduling system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the method described in any of the above embodiments.

[0061] It should be noted that the medical supply chain order scheduling system provided in this embodiment of the invention is used to execute all the process steps of the medical supply chain order scheduling method in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0062] It should be noted that the system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0063] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for scheduling medical supply chain orders, characterized in that, include: Obtain the order generation instruction for medical supplies, and extract the target material identifier and required quantity from the order generation instruction; Based on the target material identifier mapping to the preset physical inventory database, a set of material items containing serial number, production date, damage level, traceability code, packaging volume and material weight is obtained; Remove material items from the material item set whose production date is earlier than a preset expiration date threshold or whose damage level is greater than a preset damage threshold to obtain the first candidate set; The first candidate set is sorted according to the production date to obtain a time-series arrangement set, and the time-series arrangement set is truncated according to the demand quantity to obtain a second candidate set; Temperature control requirements are extracted from the traceability codes in the second candidate set, and the individual items in the second candidate set are grouped using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups. Obtain the vehicle temperature zone characteristics of the vehicles to be dispatched, calculate the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group, and combine the temperature control carpooling group whose temperature zone overlap meets the preset matching conditions with the packaging volume to obtain the loading carpooling scheme. Extract the serial number and weight of the individual materials in the loading and carpooling plan, calculate the physical grasping force of the robotic arm based on the weight of the materials, and generate a warehouse picking instruction based on the serial number.

2. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The process of removing items from the set of individual items whose production date is earlier than a preset expiration date threshold or whose damage level is greater than a preset damage threshold yields a first candidate set, including: Extract the production date and damage level of each material item from the material item set; The production date is converted into a standard timestamp, and the standard timestamp is compared with the preset expiration date threshold over a time span to obtain the comparison result; The damage level is compared with the preset damage threshold to determine the numerical value, and the determination result is obtained. Based on the comparison results and the determination results, material items whose production date exceeds a preset expiration date threshold or whose damage level exceeds a preset damage threshold are removed from the material item set to obtain a first candidate set.

3. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The first candidate set is sorted according to the production date to obtain a time-series arrangement set. The time-series arrangement set is then truncated according to the demand quantity to obtain a second candidate set, including: Extract the production dates of the individual materials in the first candidate set, and sort the production dates in ascending order to obtain a time-series arrangement set; The truncation length is determined based on the required quantity, and a truncation operation is performed on the head position of the time sequence arrangement set according to the truncation length to obtain the second candidate set.

4. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The step of extracting temperature control requirements based on the traceability codes in the second candidate set includes: Use the traceability code in the second candidate set as the key value to query the preset supplier master database and obtain the corresponding temperature control identification status code; If the temperature control status code exists, the temperature control status code is parsed to extract the corresponding lower temperature limit value and upper temperature limit value. If the temperature control status code does not exist, extract the preset upper and lower limits of the normal temperature range; The extracted upper and lower limit values ​​are combined to construct a temperature range vector, and the temperature range vector is determined as the temperature control requirement.

5. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The step of grouping the individual items in the second candidate set using the temperature control requirements as the matching key to obtain temperature-controlled carpooling groups includes: Extract the temperature interval vector corresponding to the individual material items in the second candidate set; Clustering trials are performed on each temperature interval vector according to the preset number of candidate clusters to obtain the temporary clustering architecture corresponding to each number of candidate clusters, and the global contour coefficient of each temporary clustering architecture is calculated. The number of candidate clusters that maximizes the global contour coefficient is determined as the target number of clusters. The K-means clustering algorithm is used to perform convergent iterative calculations on each of the temperature interval vectors based on the target cluster number, thereby obtaining multiple target material clusters corresponding to the target cluster number; Based on the multiple target material clusters, the individual material items are categorized to obtain temperature-controlled carpooling groups.

6. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The step of obtaining the vehicle temperature zone characteristics of the vehicles to be dispatched and calculating the temperature zone overlap between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group includes: The structured data messages returned by the vehicles to be dispatched are parsed, and the upper and lower critical temperatures supported by the cabin of the vehicles to be dispatched are extracted as the temperature zone characteristics of the vehicles. The area crossover ratio between the vehicle temperature zone characteristics and the temperature control requirements corresponding to the temperature control carpooling group is calculated to obtain the temperature zone overlap.

7. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The step of combining and calculating the temperature-controlled carpooling group whose temperature zone overlap meets the preset matching conditions with the packaging volume to obtain the loading carpooling scheme includes: Target carpooling groups with a temperature zone overlap greater than a preset crossover ratio threshold are selected from the temperature-controlled carpooling groups. Obtain the cabin size data of the vehicle to be dispatched, and combine the packaging volume with the cabin size data to calculate the spatial location allocation of the target carpooling group to obtain the three-dimensional loading layout; Obtain the current platform occupancy status data, and assign a parking platform number to the vehicle to be dispatched based on the current platform occupancy status data; The loading and carpooling plan is obtained by combining the information of the three-dimensional loading layout, the docking platform number, and the material list corresponding to the target carpooling group.

8. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The calculation of the robotic arm's physical gripping force based on the weight of the material includes: Input the weight of the material into a pre-built dynamic calculation model; The target normal force required for the robotic arm's end effector to overcome gravity and inertia is calculated using the aforementioned dynamic calculation model. The target normal force is defined as the physical gripping force of the robotic arm.

9. The medical supplies supply chain order scheduling method according to claim 1, characterized in that, The step of generating a warehouse picking instruction based on the serial number includes: The serial number is subjected to anti-counterfeiting obfuscation calculation to obtain an anti-counterfeiting traceability verification code, and the verification bit of the anti-counterfeiting traceability verification code is calculated to obtain a picking verification bit; Obtain a real-time warehouse topology map, and perform node search and obstacle avoidance optimization calculations based on the real-time warehouse topology map to obtain the path of the automated guided vehicle; The anti-counterfeiting and traceability verification code, the picking verification digit, the automated guided vehicle path, and the physical grasping force of the robotic arm are combined to generate a warehouse picking instruction.

10. A medical supplies supply chain order scheduling system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 9.

Images