Fruit supply chain end-to-end collaborative management method, device and equipment and storage medium

Abstract

This application discloses a method, apparatus, equipment, and storage medium for end-to-end collaborative management and control of a fruit supply chain, relating to the field of supply chain management technology. The method includes: receiving order information sent by stores, encapsulating it as a new order event, and pushing it in real-time to the warehousing and distribution end via a data link bus to trigger the generation of a procurement task; after receiving the goods, generating a sorting list and in-warehouse picking path based on the receiving data, guiding sorting and providing feedback on the completion status, and notifying the delivery fulfillment end to execute delivery; after delivery is completed, receiving an electronic signature result containing real-time location trajectory and synchronizing it to the store to update the order status; finally, integrating full-chain business data to generate a financial settlement statement and pushing it to the financial settlement end, completing end-to-end collaborative management and control. This application, through event-driven and real-time data integration, solves the problems of business fragmentation, low efficiency, and high operating costs caused by information lag and poor coordination at various ends of the fruit supply chain, including the demand and supply sides.

Description

Technical Field

[0001] This application relates to the field of supply chain management technology, and in particular to end-to-end collaborative management and control methods, devices, equipment and storage media for fruit supply chains. Background Technology

[0002] Against the backdrop of the rapid development of fresh food consumption, fruits, as high-frequency and high-loss agricultural products, face extremely high demands on their supply chain in terms of timeliness, freshness, and responsiveness. Chain fresh food brands and large fruit distributors generally face the demand for orders from multiple stores, in small batches, and at high frequency. They urgently need an efficient collaborative mechanism that can connect store orders with supplier delivery, warehousing and sorting, delivery fulfillment, and even financial settlement, in order to achieve precise scheduling and dynamic optimization of resources across the entire chain.

[0003] Currently, the fruit supply chain generally adopts a segmented management model: stores submit order information to headquarters via WeChat or telephone, headquarters summarizes the information and then the purchasing staff negotiates prices with suppliers in tabular form; after the supplier prepares the goods, they notify the warehouse to receive them, the warehouse sorts the goods and then hands over the paper delivery note to the delivery party to arrange transportation; finally, the finance staff needs to manually check the paper documents from multiple links to reconcile accounts. The entire process relies on a lot of manual operation and unstructured communication tools, and the information systems of each link are independent of each other, forming a typical "information silo".

[0004] This traditional approach has significant problems: First, information transmission is severely delayed and prone to distortion, leading to inconsistencies between inventory and demand data relied upon by various participants. Second, business collaboration is hampered; for example, stores cannot obtain real-time supplier inventory information, and warehousing and distribution plans are disconnected, resulting in wasted resources such as "trucks waiting for goods" or "goods waiting for trucks." Third, traceability and reconciliation are difficult; once goods are damaged or quantity discrepancies occur, it is difficult to quickly pinpoint responsibility, leading to inefficient financial settlements and frequent disputes. Therefore, the business fragmentation, inefficiency, and high operating costs caused by delayed information transmission and poor collaboration at both the demand and supply sides of the fruit supply chain have become urgent problems that need to be addressed. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, equipment and storage medium for end-to-end collaborative management of the fruit supply chain, which aims to solve the technical problems of business fragmentation, low efficiency and high operating costs caused by information lag and poor coordination at the demand and supply ends of the fruit supply chain.

[0006] To achieve the above objectives, this application proposes an end-to-end collaborative management and control method for the fruit supply chain, the method comprising:

[0007] Receive order information submitted by the store's demand side, which includes demand priority and timeliness tags, and encapsulate the order information into a new order event;

[0008] The new order event is sent to the warehouse procurement and distribution terminal via a data link bus, so that the warehouse procurement and distribution terminal pushes the procurement task to the procurement officer's workbench. The procurement task is generated by the warehouse procurement and distribution terminal based on the new order event and the procurement decision model.

[0009] Receive the receiving data uploaded by the warehouse procurement and distribution terminal, and generate a picking list and in-warehouse picking route based on the receiving data;

[0010] Based on the sorting list and the picking path within the warehouse, the warehouse procurement and distribution end is instructed to sort the goods, provide feedback on the sorting completion confirmation data, and instruct the delivery fulfillment end to deliver the goods.

[0011] The system receives an electronic signature result containing real-time location data from the delivery fulfillment terminal and sends the electronic signature result to the store demand terminal so that the store demand terminal can update the order status.

[0012] A financial settlement statement is generated based on the new order event, the procurement task, the receiving data, the sorting completion confirmation data, and the electronic signature result, and then pushed to the financial settlement terminal to complete collaborative management.

[0013] In one embodiment, the step of receiving the receiving data uploaded by the warehouse procurement terminal and generating a picking list and an in-warehouse picking route based on the receiving data includes:

[0014] Receive receiving data, including quality inspection results, uploaded by the warehousing and distribution terminal;

[0015] The inventory data is updated based on the receiving data and a sorting event is generated. The sorting event is associated with the order identifier of the new order event. The sorting event includes a sorting deadline, which is determined by the timeliness tag of the new order event.

[0016] The sorting orders corresponding to the sorting events are aggregated using a wave partitioning algorithm to generate sorting orders and in-warehouse picking routes.

[0017] In one embodiment, the step of aggregating the orders to be sorted corresponding to the sorting events using a wavelet partitioning algorithm to generate sorting orders and in-warehouse picking routes includes:

[0018] Extract the sorting deadline time and product storage identifier from the sorting events, filter out the sorting orders whose sorting deadline time is in the same preset time window and whose product storage identifier is consistent, and form a set of orders to be aggregated.

[0019] The orders to be sorted in the same delivery area within the order aggregation set are divided into the same sub-order set;

[0020] Based on the preset wave capacity and product category similarity, the sub-order set is aggregated into multiple wave tasks;

[0021] Assign a sorting resource identifier to each wave task, and generate a sorting list based on the sorting resource identifier and the product storage location in the wave task;

[0022] Based on the warehouse layout data and the product storage locations in the sorting list, the shortest path from the sorting station to each shelf is calculated, and the in-warehouse picking path is generated.

[0023] In one embodiment, the warehouse layout data includes the three-dimensional coordinates of each shelf, aisle width thresholds, and no-entry zone markers;

[0024] The step of calculating the shortest path from the sorting station to each shelf based on warehouse layout data and the product storage location in the sorting list, and generating the in-warehouse picking path, includes:

[0025] Based on the three-dimensional coordinates of each shelf and the storage location of the goods in the sorting list, determine the target shelf coordinates corresponding to each goods;

[0026] A path planning matrix is ​​constructed using the coordinates of the sorting station as the starting and ending points and the coordinates of each target shelf as the waypoints.

[0027] Based on the channel width threshold, the restricted area identifier, and the path planning matrix, the optimal access order for each of the waypoints is calculated using the Dijkstra algorithm.

[0028] Based on the optimal access order, an in-warehouse picking route is generated, which includes the shelf access order, path distance, and avoidance prompts. The in-warehouse picking route is also marked with alternative routes to avoid temporary restricted areas.

[0029] In one embodiment, the procurement decision model includes a demand forecasting module, an inventory health assessment module, and a supplier matching module;

[0030] The procurement task is generated based on the predicted demand for different types of fruit, the replenishment gap, and the priority ranking of suppliers.

[0031] The supplier priority ranking is calculated by the warehousing and distribution end inputting supplier performance data and timeliness tags into the supplier matching module, and then using the NSGA-II multi-objective optimization algorithm to calculate the ranking with quality, timeliness and cost as objectives. The supplier priority ranking includes primary suppliers, alternative suppliers and replacement trigger conditions.

[0032] The shortage to be replenished is obtained by the warehousing and distribution end inputting inventory data, purchase quantity, and fruit shelf life from fruit product data into the inventory health assessment module.

[0033] The fruit category demand forecast is obtained by the warehousing and distribution end inputting historical order data, product SKUs, demand priorities and dynamic demand influencing factors into the demand forecasting module, and then using the LSTM-XGBoost ensemble algorithm to forecast demand.

[0034] The fruit product data, the inventory data, the supplier fulfillment data, and the dynamic demand influencing factors are related data corresponding to the product SKU;

[0035] The product SKU, the purchase quantity, the demand priority, and the timeliness label are demand parameters;

[0036] The associated data is retrieved based on the required parameters;

[0037] The demand parameters are extracted by the warehousing and distribution end from the new order event.

[0038] In one embodiment, the predicted demand values ​​for different fruit categories are obtained based on the basic demand forecast and the demand adjustment coefficient;

[0039] The demand adjustment coefficient is obtained by fitting the numerical weight coefficient, feature variables and historical demand values ​​in the historical demand sequence of the same type of fruit using the XGBoost algorithm.

[0040] The basic demand forecast is obtained by inputting the historical demand sequence into the LSTM network;

[0041] The characteristic variables are obtained by transforming the dynamic demand influencing factors.

[0042] The numerical weighting coefficient is obtained by mapping the demand priority;

[0043] The historical demand sequence is obtained by filtering the historical order data based on the product SKU.

[0044] In one embodiment, the steps of instructing the warehouse procurement and distribution end to sort goods according to the sorting list and the picking path in the warehouse, providing feedback on sorting completion confirmation data, and instructing the delivery fulfillment end to deliver goods include:

[0045] The sorting order and the picking route within the warehouse are sent to the warehouse procurement and distribution terminal so that the warehouse procurement and distribution terminal can sort the goods and provide sorting completion confirmation data after sorting is completed. The sorting completion confirmation data includes wave ID, actual sorted quantity and product quality verification result.

[0046] Based on the sorting completion confirmation data and the wave information of the sorting order, a delivery trigger event is generated, which includes the total weight of the goods, the delivery area, and the storage requirements. The total weight of the goods is calculated based on the actual sorting quantity.

[0047] The delivery trigger event is pushed to the delivery fulfillment terminal so that the delivery fulfillment terminal can plan the delivery route and carry out the delivery.

[0048] Furthermore, to achieve the above objectives, this application also proposes an end-to-end collaborative management and control device for a fruit supply chain, the device comprising:

[0049] The order event module is used to receive order information submitted by the store's demand side, which includes demand priority and timeliness tags, and encapsulate the order information into a new order event;

[0050] The procurement task module is used to send the new order event to the warehouse procurement terminal via the data link bus, so that the warehouse procurement terminal can push the procurement task to the procurement officer's workbench. The procurement task is generated by the warehouse procurement terminal based on the new order event and the procurement decision model.

[0051] The sorting path module is used to receive the receiving data uploaded by the warehouse procurement terminal, and generate a sorting list and in-warehouse picking path based on the receiving data.

[0052] The sorting and delivery module is used to instruct the warehouse procurement and distribution end to sort goods according to the sorting order and the picking path in the warehouse, provide feedback on sorting completion confirmation data, and instruct the delivery fulfillment end to deliver goods.

[0053] The signature synchronization module is used to receive the electronic signature result containing real-time location trajectory from the delivery fulfillment terminal, and send the electronic signature result to the store demand terminal so that the store demand terminal updates the order status;

[0054] The financial settlement module is used to generate a financial settlement statement based on the new order event, the procurement task, the receiving data, the sorting completion confirmation data, and the electronic signature result, and push the financial settlement statement to the financial settlement terminal to complete collaborative management.

[0055] In addition, to achieve the above objectives, this application also proposes an end-to-end collaborative management and control device for a fruit supply chain, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the end-to-end collaborative management and control method for a fruit supply chain as described above.

[0056] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the fruit supply chain end-to-end collaborative management and control method described above.

[0057] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the fruit supply chain end-to-end collaborative management and control method described above.

[0058] One or more technical solutions proposed in this application have at least the following technical effects:

[0059] First, the system receives order information from store demanders, including priority and timeliness tags, and encapsulates it as a new order event to structure and standardize the demand, avoiding errors from manual transcription. Then, this event is sent in real-time to the warehouse procurement and distribution end via a data link bus, triggering the automatic generation of procurement tasks based on the procurement decision model and pushing them to the procurement staff's workbench. This shifts the procurement response from passive to proactive, improving planning accuracy. Next, the system receives receiving data uploaded from the warehouse procurement and distribution end and generates picking lists and in-warehouse picking routes accordingly, ensuring that picking operations are based on data and have optimal paths, reducing unnecessary movement. Finally, the system instructs the warehouse based on the picking lists and picking routes. The procurement and distribution end performs sorting. Upon receiving confirmation data of sorting completion, it immediately notifies the delivery fulfillment end to initiate delivery, achieving seamless integration of warehousing and transportation, and shortening the overall fulfillment time. Subsequently, it receives electronic signature results from the delivery fulfillment end, including real-time location tracking, and sends them to the store demand end to update order status, ensuring authentic and reliable delivery and improving the store experience. Finally, based on five types of consistent business data across the entire chain—new order events, procurement tasks, receiving data, sorting completion confirmation data, and electronic signature results—it automatically generates financial settlement statements and pushes them to the financial settlement end, eliminating discrepancies in reconciliation of multi-source documents and achieving automatic financial closure. This application, through event-driven and real-time data integration, solves the technical problems of business fragmentation, low efficiency, and high operating costs caused by information lag and poor coordination among the demand and supply ends in the fruit supply chain. Attached Figure Description

[0060] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0061] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating an embodiment of the fruit supply chain end-to-end collaborative management method of this application.

[0063] Figure 2 This is a flowchart illustrating Embodiment 2 of the fruit supply chain end-to-end collaborative management and control method of this application;

[0064] Figure 3 A simplified flowchart illustrating the end-to-end collaborative management and control method for the fruit supply chain provided in Embodiment 2 of this application;

[0065] Figure 4 This is a schematic diagram of the module structure of the fruit supply chain end-to-end collaborative management and control device according to an embodiment of this application;

[0066] Figure 5 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the end-to-end collaborative management and control method of the fruit supply chain in this application embodiment.

[0067] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0068] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0069] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0070] It should be noted that the executing entity in this application embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, a central control platform (also known as a central data collaboration platform), etc. The following description uses a central control platform as an example to illustrate this embodiment and the subsequent embodiments.

[0071] Based on this, the embodiments of this application provide an end-to-end collaborative management and control method for the fruit supply chain, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the fruit supply chain end-to-end collaborative management and control method of this application.

[0072] In this embodiment, the end-to-end collaborative management and control method for the fruit supply chain includes steps S10 to S60:

[0073] Step S10: Receive order information submitted by the store's demand side, which includes demand priority and timeliness tags, and encapsulate the order information into a new order event.

[0074] It's important to note that "store demand side" refers to the downstream sales terminal in the supply chain, typically a chain fruit store or retail outlet, responsible for providing goods to consumers and initiating replenishment requests. Demand priority refers to the grading of the importance of different goods or the entire order in the supply response sequence, guiding the allocation of subsequent procurement, sorting, and delivery resources (e.g., high, medium, and low levels). Timeliness tags are time constraint identifiers attached to orders, specifying the maximum fulfillment time or expected delivery period from order placement to delivery to the store (e.g., same-day delivery, next-morning delivery). Order information refers to structured business data submitted by the store demand side, containing key fields such as SKU (Stock Keeping Unit, a unique identifier for specific fruit product specifications), quantity demanded, demand priority, timeliness tags, and store identification, driving subsequent operations in each link of the supply chain. A new order event refers to a standardized message object encapsulated from the order information submitted by the store, serving as the initial signal triggering the entire supply chain collaboration process, and is published to relevant business nodes through the central data collaboration platform.

[0075] Understandably, the central control platform first receives order information submitted by stores in real time through standardized interfaces. Then, the platform performs format validation and structured processing on the received raw order data to ensure it conforms to the platform's unified data specifications, and packages it into a "new order event" message object with a unique identifier, timestamp, and event type. Finally, the central control platform publishes this "new order event" to a data link bus built on message middleware (such as Kafka or RabbitMQ), allowing downstream business nodes such as warehouse procurement and distribution, and suppliers to subscribe and respond in real time. This is done to overcome the delays caused by traditional manual transmission or batch synchronization, enabling all participants in the entire chain to perceive the store's real needs within milliseconds, thereby initiating subsequent automated collaborative processes.

[0076] Step S20: The new order event is sent to the warehouse procurement terminal via the data link bus, so that the warehouse procurement terminal pushes the procurement task to the procurement officer's workbench. The procurement task is generated by the warehouse procurement terminal based on the new order event and the procurement decision model.

[0077] It should be noted that the data link bus refers to an asynchronous communication channel built on message middleware, supporting a publish-subscribe model, used for real-time and reliable transmission of structured event messages between various business ends of the supply chain. The warehouse procurement and distribution end refers to an integrated business unit responsible for inventory management, order sorting, and procurement coordination. Its functions cover the entire process from receiving store requirements to generating procurement suggestions, organizing receiving and sorting operations. A procurement task refers to a structured instruction generated by the warehouse procurement and distribution end, containing elements such as the SKU of the goods to be procured, the suggested procurement quantity, recommended suppliers, and expected delivery time, used to guide procurement staff in executing specific procurement operations. The procurement staff workbench refers to a digital operating interface provided to procurement personnel, integrating functions such as task lists, approval portals, supplier communication tools, and historical record queries, used to receive, review, and execute procurement tasks pushed by the platform. The procurement decision model refers to an algorithmic model running on the warehouse procurement and distribution end, which integrates multiple dynamic factors such as current available inventory, allocated inventory, safety stock threshold, demand forecasts for the next few days, and supplier procurement lead times, and generates procurement tasks through preset logic.

[0078] As an example, the procurement decision model includes a demand forecasting module, an inventory health assessment module, and a supplier matching module. The procurement task is generated based on the demand forecast values ​​for different fruit categories, the replenishment gap, and supplier priority ranking. The supplier priority ranking is calculated by the warehousing and distribution end inputting supplier fulfillment data and timeliness tags into the supplier matching module, using the NSGA-II multi-objective optimization algorithm with quality, timeliness, and cost as objectives. The supplier priority ranking includes primary suppliers, alternative suppliers, and replacement trigger conditions. The replenishment gap is determined by the warehousing and distribution end inputting inventory data, procurement quantity, and the fruit shelf life from the fruit product data into the inventory health assessment module. The demand forecast values ​​for different fruit categories are obtained from the Kangdu assessment module. These forecasts are generated by the warehousing and distribution end inputting historical order data, product SKUs, demand priorities, and dynamic demand influencing factors into the demand forecasting module, and then using the LSTM-XGBoost ensemble algorithm for demand forecasting. The fruit product data, inventory data, supplier fulfillment data, and dynamic demand influencing factors are related data corresponding to the product SKUs. The product SKU, purchase quantity, demand priority, and timeliness label are demand parameters. The related data are retrieved based on the demand parameters. The demand parameters are extracted by the warehousing and distribution end from the new order events.

[0079] The demand forecasting module is an algorithmic component used to quantitatively predict the demand for different types of fruit over a certain period based on historical order data and dynamic external factors. The inventory health assessment module is a logical unit used to determine whether inventory is sufficient and calculate replenishment gaps by comprehensively considering current inventory status, procurement demand, and fruit preservation characteristics. The supplier matching module is a decision-making component that comprehensively scores and ranks available suppliers based on multi-dimensional performance indicators and constraints such as timeliness tags. Category-specific fruit demand forecasts refer to the future demand values ​​output for each type of fruit (e.g., apples, bananas), serving as an important input for procurement planning. The replenishment gap refers to the additional quantity of goods or services that need to be procured to meet demand, calculated by the warehousing and procurement department after considering current available inventory, allocated inventory, safety stock, and procurement quantity. Supplier priority ranking is a list of primary and secondary suppliers formed by comprehensively ranking all available suppliers based on three objectives: quality, timeliness, and cost, using an optimization algorithm. Supplier performance data refers to structured data recording the actual performance of each supplier over a period of time, including dimensions such as the pass rate of quality spot checks in the past 30 days, logistics transit time, and pricing. Timeliness tag refers to the maximum allowable fulfillment time or expected delivery window specified in the order from order placement to delivery to the store, used to constrain subsequent procurement and delivery arrangements. NSGA-II (Non-dominated Sorting Genetic Algorithm II) is a non-dominated sorting genetic algorithm used to find Pareto optimal solutions among multiple conflicting objectives, thereby generating a reasonable supplier ranking. Primary supplier, alternative supplier, and replacement trigger condition refer to: the primary supplier being the default preferred partner; alternative suppliers being replacement options when the primary supplier is unavailable; and the replacement trigger condition being a preset switching rule, such as automatically activating an alternative supplier when the primary supplier cannot meet the timeliness tag requirements. Inventory data refers to the set of information describing the current inventory status of fruit products in the warehouse, including current available inventory, allocated inventory, and preset safety stock dynamically adjusted according to shelf life. Purchase quantity refers to the specific order quantity of a certain SKU of fruit required by the store in a new order event, serving as one of the core input parameters for generating the procurement task. Fruit product data refers to the basic attribute information associated with a specific fruit SKU, including shelf life, real-time quality grade, and historical spoilage rate for that category. Historical order data, product SKUs, demand priority, and dynamic demand influencing factors are a set of input variables used to drive the demand forecasting model: historical order data reflects past sales patterns; product SKUs identify specific fruit categories; demand priority reflects the urgency of orders; and dynamic demand influencing factors include external disturbance variables such as holiday factors, weather factors, and promotional plans.The LSTM-XGBoost ensemble algorithm is a hybrid machine learning method that uses Long Short-Term Memory (LSTM) networks to capture trends and periodic features in time series data, and then feeds the output of LSTM into the XGBoost model to further integrate static and dynamic factors for regression prediction. Related data refers to a collection of various auxiliary information associated with a specific product SKU, including fruit product data, inventory data, supplier fulfillment data, and dynamic demand influencing factors, used to support multi-dimensional analysis in the procurement decision-making model. Demand parameters are key fields extracted from new order events that directly reflect the store's current demand, including product SKU, purchase quantity, demand priority, and timeliness tags; they serve as the initial basis for retrieving related data and initiating the procurement decision-making process.

[0080] First, the warehousing and distribution end reads the product SKU, purchase quantity, demand priority, and timeliness tag from new order events. Using these fields as query conditions, it pulls corresponding fruit product data, inventory data, supplier fulfillment data, and dynamic demand influencing factors from the data warehouse. Then, the warehousing and distribution end feeds historical order data, product SKU, demand priority, and dynamic demand influencing factors into a pre-trained LSTM-XGBoost ensemble model. The LSTM layer processes the daily sales sequence of the past 90 days (for example only) and outputs a time-series feature vector. The XGBoost layer receives this vector and integrates structured factors such as holidays, weather, and promotions. Finally, it outputs the daily granular demand forecast value of the fruit category to which the SKU belongs for a preset future time period (e.g., the next 7 days), and summarizes the forecast values ​​for each fruit category. Next, the warehouse procurement and distribution end reads the current available inventory and allocated inventory from the inventory data, calculates the net inventory value = current available inventory - allocated inventory, and then determines whether to adjust the safety stock based on the shelf life in the fruit product data (if the shelf life is <3 days, then the safety stock = original safety stock × 0.8). Subsequently, it determines whether the net inventory value is ≥ the purchase quantity + the adjusted safety stock. If it does not meet the requirement, it calculates the replenishment gap = (purchase quantity + adjusted safety stock) - net inventory value. Meanwhile, the warehousing and distribution end inputs the pass rate of all available suppliers in the past 30 days, average transit time, latest price, and the maximum allowable transit time specified by the timeliness label into the deployed NSGA-II multi-objective optimization algorithm. The algorithm uses quality (pass rate × 0.4), timeliness (1 / max(1, actual transit time - allowable time by timeliness label) × 0.3, with a timeliness score of 0 if the time is exceeded), and cost (1 / price × 0.3) as objective functions to perform multiple rounds of non-dominated sorting and congestion calculation. Finally, it outputs a list of suppliers ranked by comprehensive merit, and marks the first-ranked supplier as the primary supplier and the second-ranked supplier as the alternative supplier. The replacement trigger condition is set as "the primary supplier's estimated arrival time > the timeliness label's required time". Finally, the warehousing and distribution end will use the demand forecast values ​​for different types of fruits to verify the long-term supply trend, take the unfilled gap as the core quantity basis for this purchase, and combine it with the primary supplier information in the supplier priority ranking to assemble and generate a purchase task record containing product SKU, purchase quantity (using the unfilled gap value), primary supplier ID, alternative supplier ID and automatic switching rules, and write it into the purchase task queue and push it to the purchaser's workbench.

[0081] As an example, the predicted demand value for each fruit category is obtained based on the basic demand prediction value and the demand adjustment coefficient; the demand adjustment coefficient is obtained by fitting the numerical weight coefficient, feature variables, and historical demand values ​​in the historical demand sequence of the same fruit category using the XGBoost algorithm; the basic demand prediction value is obtained by inputting the historical demand sequence into an LSTM network; the feature variables are obtained by transforming the dynamic demand influence factor; the numerical weight coefficient is obtained by mapping the demand priority; and the historical demand sequence is obtained by filtering the historical order data based on the product SKU.

[0082] The basic demand forecast refers to the preliminary time series prediction result output by an LSTM network after inputting the historical demand sequence corresponding to a specific product SKU into the network, without considering external disturbances. The demand adjustment coefficient is a multiplicative factor used to correct the basic demand forecast after regression fitting of numerical weight coefficients and feature variables using the XGBoost algorithm, reflecting the influence of priority and external environment on actual demand. The numerical weight coefficient is a value converted from the demand priority in the order according to a preset rule: high priority is mapped to 1.2, medium priority to 1.0, and low priority to 0.8, used to quantify the amplification or reduction effect of different priorities on demand. Feature variables are structured input data that can be used by machine learning models after standardization or one-hot encoding of dynamic demand influencing factors (such as holidays, weather, and promotional plans). The historical demand sequence refers to the daily (or periodic) actual order quantity sequence selected from historical order data based on product SKUs and arranged in chronological order. The historical demand value is each specific value in the historical demand sequence, representing the actual demand for that product SKU at a certain point in the past.

[0083] First, the warehousing and distribution end performs a structured query in the historical order database based on the product SKU in the new order event. This filters out all order records belonging to the same fruit category as the SKU (e.g., "Red Fuji apples" are categorized under "apples"), and aggregates them by date to calculate the daily total order volume, forming a continuous time series—the historical demand series. Second, the warehousing and distribution end converts the demand priority of this order into numerical weight coefficients according to a preset mapping rule. Simultaneously, it preprocesses dynamic demand influencing factors: holiday factors are converted to 0 / 1 Boolean values, weather factors are encoded as discrete values ​​based on temperature and precipitation levels, and promotional plans are converted into numerical features indicating participation and discount levels. This generates a set of feature variables aligned with the prediction period. Then, the warehousing and distribution end inputs the historical demand series into a deployed LSTM network model. This model uses data from the past 90 days as an input window and learns the seasonality, trends, and fluctuation patterns of demand through multi-layered recurrent neurons, outputting basic demand forecasts for each day of the next 7 days. Next, the warehousing and distribution end calls the XGBoost regression model. This model was trained offline using historical data, and its inputs include numerical weight coefficients and feature variables for the corresponding historical time periods. The output is the actual historical demand value. During the online inference phase, the numerical weight coefficients and feature variables corresponding to the current forecast period are input into the model to obtain the daily demand adjustment coefficients for the next 7 days. Finally, the warehousing and distribution end multiplies the daily basic demand forecast value output by the LSTM with the corresponding daily demand adjustment coefficient output by XGBoost element-wise to generate the final demand forecast value for each fruit category. This is done to dynamically incorporate the impact of order urgency and external environmental disturbances while preserving the inherent regularity of the time series, making the forecast results closer to the actual demand in real business scenarios.

[0084] Step S30: Receive the receiving data uploaded by the warehouse procurement terminal, and generate a picking list and in-warehouse picking route based on the receiving data.

[0085] It should be noted that receiving data refers to structured information uploaded to the central control platform by the warehousing and distribution end after completing the supplier's acceptance of the goods. This information includes fields such as product SKU, actual received quantity, receiving time, batch number, and quality inspection results. A picking list is a list of work instructions generated by the central control platform after matching the receiving data with orders to be fulfilled. It clearly lists the product SKUs to be picked, the corresponding stores, the picking quantity, and the order association identifier, guiding the picking staff to perform the picking operations. In-warehouse picking routes refer to the path optimization algorithms (such as mileage saving or A / B picking) that the central control platform uses based on the warehouse layout, storage location distribution, and the location of each SKU in the picking list. The algorithm calculates the shortest or most efficient picking route to reduce unnecessary movement of pickers and improve operational efficiency.

[0086] Understandably, firstly, the central control platform receives receiving data uploaded by the warehouse procurement and distribution end in real time via API interface after the receiving operation is completed, and performs integrity verification and format standardization on the data. Secondly, the platform performs multi-dimensional matching of the verified receiving data with all store orders in the "purchased, awaiting sorting" state—using the product SKU as the key, combined with timeliness tags and demand priority, to filter out order lines that can be covered by this receiving, and aggregates the matching results by store ID and product SKU to generate a structured sorting list containing store identifier, product SKU, quantity to be sorted, associated order number, and batch information. Then, the platform retrieves the latest warehouse location layout map and real-time inventory snapshot, locates the current physical location of each product SKU in the sorting list (e.g., A-section-03-02), and runs a pre-deployed path optimization algorithm (using A... The algorithm (considering aisle width, customer flow density, and shelf height limitations) calculates the shortest, non-backtracking route from the sorting starting point, through all target storage locations, and finally back to the packing area—the in-warehouse picking path. Finally, the platform packages the generated picking list and the corresponding in-warehouse picking path into a single picking task package, which is then pushed to the designated sorter's mobile terminal via a message queue. This ensures that the sorting operation accurately matches received goods with store orders while minimizing unnecessary movement through the optimal path, significantly improving sorting efficiency and order fulfillment timeliness.

[0087] Step S40: Based on the sorting list and the picking path in the warehouse, instruct the warehousing and distribution end to sort the goods, provide feedback on the sorting completion confirmation data, and instruct the delivery fulfillment end to deliver the goods.

[0088] It should be noted that sorting completion confirmation data refers to structured feedback information submitted by the warehousing and distribution end through mobile terminals or the sorting system after the sorting operation is completed. This includes the sorting order number, the actual SKU and quantity of sorted goods, the sorting completion time, the operator ID, and anomaly markers (such as out of stock, damage, etc.), used to confirm that the sorting task has been completed. The delivery fulfillment end refers to the logistics execution unit responsible for transporting the sorted goods from the warehouse to various stores. This includes the dispatching system, drivers, delivery vehicles, and supporting mobile applications, and is responsible for fulfillment processes such as route planning, loading, in-transit tracking, and store receipt confirmation.

[0089] As an example, the steps of instructing the warehousing and distribution end to sort goods according to the sorting list and the in-warehouse picking route, feeding back sorting completion confirmation data, and instructing the delivery fulfillment end to deliver goods include: sending the sorting list and the in-warehouse picking route to the warehousing and distribution end so that the warehousing and distribution end can sort goods and feed back sorting completion confirmation data after sorting is completed, the sorting completion confirmation data including wave ID, actual sorted quantity, and product quality verification results; generating a delivery trigger event including total weight of goods, delivery area, and storage requirements based on the sorting completion confirmation data and the wave information of the sorting list, the total weight of goods being calculated based on the actual sorted quantity; and pushing the delivery trigger event to the delivery fulfillment end so that the delivery fulfillment end can plan the delivery route and deliver goods.

[0090] Wave ID refers to a unique batch identifier assigned by the central control platform when generating a sorting list. It is used to aggregate multiple orders from the same time period, delivery area, or with similar product characteristics into a single sorting and delivery unit, facilitating batch processing and tracking. Actual sorted quantity refers to the specific quantity of each SKU in the sorting list that is actually counted, packed, and shipped out during the sorting operation at the warehouse procurement and distribution end. This quantity may differ from the planned sorted quantity due to insufficient inventory or losses. Product quality verification result refers to the quality status recorded by the sorter after a secondary inspection of the fruit's appearance, ripeness, and damage during the sorting process. It is usually expressed as a grade (e.g., A / B / C) or anomaly labels (e.g., "minor dent" or "overripe"). Total weight of goods refers to the total weight of all goods in that wave, calculated by summing the actual sorted quantity from the sorting completion confirmation data and the preset unit weight of each SKU (e.g., 5 kg per box of apples). Delivery area refers to the geographical region information of a store (e.g., "a certain city and a certain district"), used to guide the delivery fulfillment end in clustering orders by region, optimizing vehicle scheduling, and route planning. Storage requirements refer to the special conditions required for fruit products during transportation, such as temperature control, humidity, and stacking methods (e.g., "refrigerated at 0–4℃" and "do not invert"), derived from the fruit product data corresponding to the product SKU, used to ensure quality stability during transportation. Delivery trigger events are structured message objects generated by the central control platform to initiate the delivery process, containing wave ID, total weight of goods, delivery area, storage requirements, and a list of associated stores, serving as input for the delivery fulfillment end to execute route planning and loading scheduling.

[0091] First, the central control platform sends the generated sorting list and the in-warehouse picking route to the mobile terminal of the warehouse procurement and distribution end in real time via internal API. The sorting staff at the warehouse procurement and distribution end check the product SKUs one by one in the warehouse and complete the physical picking according to the location sequence indicated by the in-warehouse picking route. After the picking is completed, the operation is confirmed on the terminal. The warehouse procurement and distribution end collects and uploads the sorting completion confirmation data, which includes the wave ID corresponding to the task, the actual sorting quantity of each SKU (by scanning or manual entry), and the product quality review results of the sorting staff after on-site inspection of the fruit's appearance, ripeness, etc. Secondly, after receiving the sorting completion confirmation data, the central control platform immediately associates it with the sorting order corresponding to the wave ID, extracts the delivery area information, and queries the unit weight of each SKU from the product master data. It then multiplies the actual sorted quantity of each SKU by its unit weight and sums the results to calculate the total weight of the goods in that wave. Simultaneously, it reads the storage requirements shared by all orders in that wave and packages the wave ID, total weight of goods, delivery area, and storage requirements into a delivery trigger event. Finally, the central control platform pushes this delivery trigger event to the dispatch system at the delivery fulfillment end via a message middleware. Upon receiving the event, the delivery fulfillment end automatically triggers the delivery task creation process: it aggregates all stores covered by this wave based on the delivery area, selects refrigerated or ambient temperature vehicles of appropriate tonnage based on the total weight of goods, configures the vehicle temperature control parameters according to the storage requirements, and calls the route planning engine (such as based on the Dijkstra algorithm or Clarke-Wright saving algorithm) to generate the optimal delivery route that avoids congestion and meets the timeliness requirements. The task is then dispatched to the driver's app, and loading and delivery are initiated. This is done to automatically trigger precise and efficient delivery scheduling once sorting is completed, avoiding delays caused by manual handling and ensuring that the fruit is delivered to the store quickly in optimal condition.

[0092] Step S50: Receive the electronic signature result containing real-time location trajectory from the delivery fulfillment terminal, and send the electronic signature result to the store demand terminal so that the store demand terminal updates the order status.

[0093] It should be noted that real-time location tracking refers to the sequence of vehicle or delivery personnel geographic coordinates continuously collected and uploaded by the delivery fulfillment end through the GPS module of the driver's mobile terminal during the delivery process, including information such as timestamps, latitude and longitude, and positioning accuracy. Electronic signature results refer to the structured signature voucher collected by the delivery fulfillment end through a mobile terminal after the goods are handed over at the store, including the signature time, the name or employee number of the signatory, a photo of the signature, a handwritten electronic signature, and the simultaneously attached real-time location tracking. Order status refers to the business stage identifier of the order recorded in the central control platform, such as "order placed," "purchased," "sorted," "in delivery," and "signed for," used to reflect the order's execution progress throughout the entire supply chain process.

[0094] Understandably, firstly, the central control platform monitors and receives electronic signature results uploaded by the delivery fulfillment terminal via a pre-defined API interface after the driver completes the store delivery operation. This result is generated synchronously by the terminal when the signature action is triggered. Secondly, the platform automatically verifies the received electronic signature results: firstly, it checks whether the required fields are complete; secondly, it compares whether the endpoint coordinates of the real-time location trajectory fall within the store's pre-defined geofence (e.g., a circular area with a radius of 200 meters centered on the store's latitude and longitude). If both are successful, the signature is deemed valid. Finally, the platform pushes the complete electronic signature result to the corresponding store's order management system via an internal message queue. Upon receiving the result, the store immediately locates the order record and changes its order status field from "in transit" to "signed." Simultaneously, the electronic signature, photo, and location trajectory are stored as tamper-proof fulfillment credentials on the order details page. This is done to ensure the authenticity and traceability of the signature behavior, automatically drive the closed-loop order lifecycle, avoid delays or errors in manual confirmation, and provide authoritative evidence for subsequent automatic financial reconciliation.

[0095] Step S60: Generate a financial settlement statement based on the new order event, the procurement task, the receiving data, the sorting completion confirmation data, and the electronic signature result, and push the financial settlement statement to the financial settlement terminal to complete collaborative management.

[0096] It should be noted that a financial settlement statement is a structured document automatically generated by the central control platform based on end-to-end business events. It is used for reconciliation and payment, and includes fields such as product SKU, purchase quantity, actual received quantity, sorting loss, receipt confirmation status, supplier information, unit price, total amount, and timestamps for each stage, ensuring consistency between business flow and cash flow. The financial settlement terminal refers to the internal financial system or personnel's terminal responsible for accounts payable / receivable management, invoice verification, cost accounting, and payment execution. It receives, reviews, and processes financial settlement statements to complete the final accounting and fund settlement.

[0097] Understandably, firstly, the central control platform uses the unique order identifier (such as order ID or wave ID) to query and retrieve five key records from the internal data bus: new order event (extracting product SKU, purchase quantity, and store demand ID), purchase task (extracting supplier ID, agreed unit price, and purchase time), receiving data (extracting actual received quantity, receiving time, and batch number), sorting completion confirmation data (extracting actual sorted quantity and product quality verification results), and electronic signature result (extracting signature status, signature time, and signature recipient information). Secondly, the platform aligns the above data line by line by product SKU, using the "signed" status in the electronic signature result as a prerequisite for settlement validity—if the signature is successful, the actual received quantity in the receiving data is used as the settlement quantity; if there is sorting loss or quality degradation, the reason for the difference is noted in the remarks field, and the settlement quantity is multiplied by the agreed unit price in the purchase task to calculate the amount payable for each product. Finally, a complete financial settlement statement is generated, including supplier name, store name, order cycle, timestamps of each stage, settlement details, and total amount. Finally, the platform pushes the financial settlement statement to the designated interface of the financial settlement system via an encrypted API. The financial settlement system automatically receives it and stores it in the pending bill pool, triggering subsequent review and payment processes. This is done to eliminate the cumbersome operations of manually collecting paper documents and comparing data across systems, ensuring that financial settlement is based on consistent and tamper-proof business facts throughout the entire process, and achieving an automatic, accurate, and traceable closed loop of integrated business and finance.

[0098] In a pilot project of a large fruit chain enterprise, the following quantitative benefits were achieved after applying this embodiment:

[0099]

[0100] This embodiment provides an end-to-end collaborative management method for the fruit supply chain. First, it receives order information from store demanders, including priority and timeliness tags, and encapsulates it as a new order event, achieving structured and standardized demand to avoid errors from manual transcription. Then, it sends this event to the warehouse procurement and distribution end in real time via a data link bus, triggering the end to automatically generate procurement tasks based on a procurement decision model and push them to the procurement staff's workbench, shifting procurement response from passive to proactive and improving planning accuracy. Next, it receives receiving data uploaded from the warehouse procurement and distribution end and generates sorting lists and in-warehouse picking routes accordingly, ensuring that sorting operations are based on data and have optimal routes, reducing unnecessary movement. Then, based on… The sorting list and picking route instruct the warehousing and distribution end to perform sorting. Upon receiving confirmation data of sorting completion, the delivery fulfillment end is immediately notified to initiate delivery, achieving seamless integration of warehousing and transportation, and shortening overall fulfillment time. Subsequently, the system receives electronic signature results from the delivery fulfillment end, including real-time location tracking, and sends them to the store demand end to update order status, ensuring authentic and reliable delivery and improving the store experience. Finally, based on five types of consistent business data across the entire chain—new order events, procurement tasks, receiving data, sorting completion confirmation data, and electronic signature results—a financial settlement statement is automatically generated and pushed to the financial settlement end, eliminating discrepancies in reconciliation of multi-source documents and achieving automatic financial closure. This embodiment, through event-driven and real-time data integration, solves the technical problems of business fragmentation, low efficiency, and high operating costs caused by information lag and poor coordination among the demand and supply ends in the fruit supply chain.

[0101] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the fruit supply chain end-to-end collaborative management and control method of this application. Step S30 of the fruit supply chain end-to-end collaborative management and control method includes steps S31 to S33:

[0102] Step S31: Receive the receiving data, which includes quality inspection results, uploaded by the warehousing and distribution terminal.

[0103] It should be noted that the quality inspection results refer to the quality assessment information recorded after sampling or full inspection of the received fruits in the warehouse receiving process. This includes the judgment results of indicators such as appearance (such as rot, color difference), maturity, firmness, and sugar content. They are usually presented in the form of qualified / unqualified, grade (such as A / B / C grade) or specific numerical values, and are used to determine whether the batch of goods meets the procurement standards and decide whether to put it into the warehouse.

[0104] Step S32: Update the inventory data according to the receiving data and generate a sorting event. The sorting event is associated with the order identifier of the new order event. The sorting event includes a sorting deadline, which is determined by the timeliness tag of the new order event.

[0105] It should be noted that a "pending sorting event" refers to a structured message object generated by the central control platform after receiving goods, used to trigger sorting operations. This object includes information such as the SKU of the goods to be sorted, quantity, associated order identifier, and sorting deadline, serving as the basis for initiating the sorting process at the warehouse procurement and distribution end. An order identifier is a unique number or ID that identifies a store order in a new order event, used to track and associate all subsequent operations and data of that order throughout the entire supply chain process. The sorting deadline refers to the latest time to complete sorting calculated by the central control platform based on the time-sensitive tags (such as "same-day delivery" or "next-morning delivery") attached to the new order event, ensuring that subsequent deliveries can be fulfilled on time.

[0106] Understandably, firstly, after receiving the receiving data uploaded by the warehouse procurement and distribution end, the central control platform parses the product SKU, actual received quantity, batch number, and receiving time, and writes this information into the inventory database, updating the current available inventory field for that SKU. Simultaneously, it associates the newly added inventory with the corresponding batch to support subsequent scheduling based on first-in-first-out (FIFO) or shelf-life priority. Secondly, based on the order source implicit in the receiving data, the platform traces back and matches the corresponding new order events, extracting a unique order identifier and timeliness tag. Then, based on preset timeliness rules (e.g., for "same-day delivery" orders requiring delivery before 18:00, a default 2-hour delivery time is reserved, and sorting must be completed no later than 16:00), the platform calculates the specific sorting deadline. Finally, the platform creates a new pending picking event record, filling in fields such as order identifier, product SKU, quantity to be picked (the smaller of the purchase quantity and the actual received quantity in the new order event), picking deadline, and warehouse ID according to a standard format, writing it to the event queue, and publishing it through the message bus to ensure that the subsequent picking scheduling system can automatically trigger picking tasks based on this event. This is done to activate the picking process immediately after receiving the goods and to enforce timely fulfillment by binding the picking deadline, avoiding failure to meet the store's timeliness label requirements due to operational delays.

[0107] Step S33: Aggregate the orders to be sorted corresponding to the sorting events using a wave partitioning algorithm to generate sorting orders and in-warehouse picking paths.

[0108] It should be noted that wave partitioning algorithm refers to a scheduling algorithm used to cluster and merge multiple orders awaiting sorting according to preset rules (such as similar delivery areas, consistent time-delivery labels, similar product categories, or overlapping time windows). The purpose is to reduce repetitive picking trips and improve warehouse operation efficiency. Orders awaiting sorting refer to store orders that have been received but have not yet been sorted, which are associated with the pending sorting event. Each order contains information such as product SKU, quantity to be sorted, order identifier, sorting deadline, and delivery destination, serving as the basic unit for wave partitioning and sorting task generation.

[0109] As an example, the step of aggregating the orders to be sorted corresponding to the sorting events using a wave partitioning algorithm to generate sorting orders and in-warehouse picking paths includes: extracting the sorting deadline time and product storage identifier from the sorting events; filtering out the orders to be sorted whose sorting deadline time is in the same preset time window and whose product storage identifier is consistent, forming a set of orders to be aggregated; dividing the orders to be sorted in the same delivery area in the set of orders to be aggregated into the same sub-order set; aggregating the sub-order set into multiple wave tasks according to the preset wave capacity and product category similarity; assigning sorting resource identifiers to each wave task, and generating sorting orders according to the sorting resource identifiers and the product storage locations in the wave tasks; calculating the shortest path from the sorting station to each shelf according to the warehouse layout data and the product storage locations in the sorting orders, and generating in-warehouse picking paths.

[0110] Product storage identifiers are codes used to distinguish the storage area or temperature zone type of fruits in the warehouse, such as "Refrigerated Zone-A" and "Ambient Temperature Zone-B," ensuring that products with consistent storage conditions are grouped and processed. A preset time window refers to a time interval threshold (e.g., within 2 hours) set by the central control platform to filter orders with similar sorting deadlines, ensuring that the same batch of operations can be completed under uniform timeliness requirements. The set of orders to be aggregated refers to the collection of all orders that meet the same preset time window and have the same product storage identifier after initial screening, serving as the basic input for subsequent clustering. The same sub-order set refers to a subset of orders in the set of orders to be aggregated, further divided by delivery area, with adjacent or identical destinations, facilitating subsequent delivery route optimization. The preset wave capacity refers to the maximum number of orders that a single wave task can contain (e.g., 40 orders), used to control the scale of sorting tasks and avoid overload or delays. Product category similarity refers to the degree of overlap in fruit categories among different orders. It is typically calculated based on the intersection ratio of the major categories to which the SKUs belong (e.g., apples, bananas). Higher similarity tends to merge orders into the same wave, reducing cross-region picking. A wave task is a sorting operation unit with a unique wave ID, aggregated from one or more orders to be sorted. It serves as the basic unit for generating sorting orders and scheduling resources. Sorting resource identification refers to the physical and manpower configuration information assigned to a wave task, including the sorting station number (designating the workstation) and the sorting personnel ID (designating the executor). Product storage location refers to the specific coordinates of each product SKU in the warehouse, such as "Area A - Row 03 - Floor 02," used for path planning and picking navigation. Warehouse layout data refers to digital map information describing the internal structure of the warehouse, including shelf coordinates, aisle widths, turning restrictions, and restricted areas, supporting accurate path modeling. A sorting station is a fixed workstation within the warehouse where sorters receive tasks, retrieve containers, and complete order-based picking or sorting operations; it is the starting and ending point of the sorting process.

[0111] First, the central control platform traverses all pending sorting events, reads the sorting deadline and product storage identifier for each event, and performs a filtering operation: only those pending sorting orders whose sorting deadlines differ by no more than a preset time window and whose product storage identifiers are exactly the same are retained, and these orders are merged into a set of orders to be aggregated, so as to ensure that orders processed in the same batch are compatible in terms of timeliness requirements and storage temperature zones. Secondly, the order set to be aggregated is grouped according to the delivery area field, and orders with the same delivery area are grouped into the same sub-order set so that the delivery route can be efficiently clustered in subsequent deliveries. Then, for each sub-order set, the similarity of product categories between any two orders is calculated (by comparing the intersection ratio of the fruit categories to which their respective SKUs belong). Then, combined with the preset wave capacity, a greedy clustering strategy is used to gradually merge orders with high similarity and whose total number does not exceed the limit, forming multiple wave tasks. Each task generates a unique wave ID. Next, for each wave task, an idle sorting station number and a sorting personnel ID are allocated from the available resource pool as sorting resource identifiers. The corresponding product storage location is queried according to all the product SKUs included in the wave task, and a structured sorting list is generated, which clearly lists the wave ID, each product SKU, the quantity to be sorted, and the associated order identifier. Finally, the central control platform loads the warehouse layout data, using the assigned sorting station as the starting point and all the product storage locations in the sorting list as target points. It then uses a path planning algorithm to calculate a continuous route that covers all target points, has the shortest total walking distance, and avoids restricted areas, generating the final in-warehouse picking route. This is done to maximize sorting and delivery efficiency while minimizing redundant paths and resource conflicts, while ensuring timely fulfillment and consistent storage conditions.

[0112] As an example, the warehouse layout data includes the three-dimensional coordinates of each shelf, aisle width thresholds, and restricted area markers. The step of calculating the shortest path from the sorting station to each shelf based on the warehouse layout data and the product storage location in the picking list, and generating the in-warehouse picking path, includes: determining the target shelf coordinates for each product based on the three-dimensional coordinates of each shelf and the product storage location in the picking list; constructing a path planning matrix with the sorting station coordinates as the start and end points and each target shelf coordinate as a waypoint; calculating the optimal access order for each waypoint using the Dijkstra algorithm based on the aisle width threshold, the restricted area markers, and the path planning matrix; and generating an in-warehouse picking path that includes the shelf access order, path distance, and avoidance prompts based on the optimal access order. The in-warehouse picking path also marks alternative aisles to avoid temporary restricted areas.

[0113] The aisle width threshold refers to the minimum passage width limit preset by the central control platform (e.g., 1.2 meters), used to determine whether a certain aisle allows sorting equipment or personnel to pass through, avoiding path planning into excessively narrow areas. Restricted area markers are codes or labels marking prohibited areas within the warehouse, such as "Equipment Maintenance Area" or "Forklift Operation Area," ensuring paths avoid safety hazards or areas of operational conflict. Target shelf coordinates refer to the specific three-dimensional coordinates of the shelves retrieved from the shelf database based on the storage location corresponding to each product SKU in the picking list, serving as necessary path points for path planning. The path planning matrix is ​​a table of distances and connectivity between nodes constructed based on warehouse layout data, using the sorting station and all target shelf coordinates as nodes, to support shortest path algorithm calculations. Dijkstra's algorithm is a classic graph search algorithm used to solve for single-source shortest paths in weighted directed graphs; here, it is used to find the optimal walking route while satisfying aisle and restricted area constraints. Optimal access order refers to the order in which target shelves are visited to minimize the total walking distance or time under all constraints, and is the core output of path optimization. Shelf access sequence, path distance, and avoidance prompts refer to the specific details included in the picking path within the warehouse: the shelf access sequence lists the shelf numbers to be visited in order; the path distance indicates the actual walking distance between adjacent nodes; avoidance prompts include safety reminders such as "Slippery area ahead, please proceed slowly." Alternative routes refer to alternative routes pre-calculated and embedded in the path by the central control platform when a section of the main path is temporarily marked as prohibited (e.g., due to sudden equipment failure), ensuring uninterrupted picking operations.

[0114] First, the central control platform iterates through each SKU in the picking list, accurately retrieves the unique code of the corresponding shelf from the warehouse master data table based on its associated product storage location field, and extracts the pre-stored 3D coordinates (X, Y, Z) of that shelf as the target shelf coordinates for that product, ensuring that each product to be picked has a clear spatial location. Second, the central control platform sets the currently assigned 3D coordinates of the picking station as the fixed start point and final return point of the path, and treats all target shelf coordinates as necessary waypoints. Based on the warehouse layout data, it calculates the straight-line distance between any two nodes (including the picking station and each shelf, and the distance between shelves), and combines this with the connectivity of the passageway to determine whether it is passable, constructing an N×N path planning matrix, where N is the total number of nodes, and the matrix elements represent the distance of the feasible path between two points or are marked as unreachable. Then, the central control platform performs constraint filtering on the path planning matrix: based on the aisle width threshold, it removes aisle connections with widths smaller than the equipment or personnel passage requirements; simultaneously, based on prohibited area markers, it marks shelves or road sections located in areas such as equipment maintenance areas and forklift operation areas as impassable, forming a restricted feasibility graph. On this basis, it calls the Dijkstra algorithm, starting from the sorting station, to sequentially expand the shortest path tree, comprehensively considering minimizing the total walking distance to all passing points, to find an optimal access order that satisfies the constraints. Finally, the central control platform generates a complete in-warehouse picking route according to this optimal access order. The route content includes the ordered shelf access sequence, the precise path distance for each movement (in meters), automatically inserted avoidance prompts in narrow aisles or slippery areas, and dynamically activates pre-cached detour schemes when a main path passes through a temporarily added prohibited area, marking an alternative route to avoid that area. This is done to achieve high-precision, high-efficiency, and high-safety picking guidance in complex warehouse environments, avoiding unnecessary backtracking, passage conflicts, or operation interruptions, and significantly improving picking efficiency and the operator experience.

[0115] This embodiment first receives receiving data, including quality inspection results, uploaded from the warehouse procurement and distribution end to ensure that the incoming fruits meet quality requirements, and updates the inventory data accordingly. Then, based on the order identifier and timeliness tag in the new order event associated with this receiving data, a sorting event is generated and automatically set with a sorting deadline determined by the timeliness tag, ensuring that subsequent operations start on time. Next, a wave partitioning algorithm is used to aggregate the sorting orders corresponding to the sorting events, comprehensively considering the sorting deadline window, product storage identifier, delivery area, wave capacity, and category similarity to form a structured sorting list and optimized in-warehouse picking path. This embodiment achieves controllable receiving quality, manageable sorting timeliness, and efficient warehouse operations, effectively supporting the precise fulfillment and collaborative operation of the fruit supply chain.

[0116] For example, to help understand the implementation process of the end-to-end collaborative management and control method for the fruit supply chain obtained by combining this embodiment with the above embodiment one, please refer to... Figure 3 , Figure 3 A simplified flowchart illustrating an end-to-end collaborative management method for the fruit supply chain is provided, specifically:

[0117] This diagram illustrates the full-process architecture of a collaborative management and control method for the fruit supply chain. Centered on a central management platform, it facilitates information exchange and process linkage between various business units via a data link bus. The top of the diagram represents the demand side, including the "store order placement" stage, where stores initiate orders. These orders enter the central management platform via order events and are then transmitted downstream via the "data link bus (message queue / API)." The central management platform acts as the core hub, connecting the warehousing and procurement end, the delivery fulfillment end, the supplier end, and the financial settlement end. At the warehousing and procurement end, the system receives "synchronized orders" from the data link bus, triggering the "order receipt" process. Subsequently, the "procurement decision algorithm automatically generates procurement tasks," which are then pushed to the procurement staff for execution via "procurement events." Upon receiving the "synchronized procurement" instruction, the supplier confirms the order and completes shipment. After the goods arrive, the warehousing and procurement end executes "warehouse receipt," uploads the receipt information, triggers a sorting completion event, and initiates the intelligent sorting process. On the delivery fulfillment side, after receiving a synchronization task, the system enters the process through "Receive Delivery Task," uses route optimization algorithms to plan the delivery route, and the driver executes the delivery. Electronic signatures are then completed at the store, generating a "Signature Completed Event" which is fed back to the central control platform. Finally, the central control platform synchronizes all the data to the financial settlement end, achieving automatic reconciliation. The entire process is linked through event-driven mechanisms and a data link bus, forming an end-to-end closed-loop collaborative system from store order placement to financial settlement.

[0118] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the end-to-end collaborative management method of the fruit supply chain in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0119] This application also provides an end-to-end collaborative management and control device for the fruit supply chain. Please refer to [link / reference]. Figure 4 The fruit supply chain end-to-end collaborative management and control device includes:

[0120] The order event module 10 is used to receive order information containing demand priority and timeliness tags submitted by the store demand side, and encapsulate the order information into a new order event;

[0121] The procurement task module 20 is used to send the new order event to the warehouse procurement terminal via the data link bus, so that the warehouse procurement terminal can push the procurement task to the procurement officer's workbench. The procurement task is generated by the warehouse procurement terminal based on the new order event and the procurement decision model.

[0122] The sorting path module 30 is used to receive the receiving data uploaded by the warehouse procurement terminal and generate a sorting list and in-warehouse picking path based on the receiving data.

[0123] The sorting and delivery module 40 is used to instruct the warehouse procurement and distribution end to sort goods according to the sorting order and the picking path in the warehouse, provide feedback on sorting completion confirmation data, and instruct the delivery fulfillment end to deliver goods.

[0124] The signature synchronization module 50 is used to receive the electronic signature result containing real-time location trajectory from the delivery fulfillment terminal, and send the electronic signature result to the store demand terminal so that the store demand terminal updates the order status;

[0125] The financial settlement module 60 is used to generate a financial settlement statement based on the new order event, the procurement task, the receiving data, the sorting completion confirmation data, and the electronic signature result, and push the financial settlement statement to the financial settlement terminal to complete collaborative management.

[0126] The fruit supply chain end-to-end collaborative management and control device provided in this application, employing the fruit supply chain end-to-end collaborative management and control method in the above embodiments, can solve the technical problems of business fragmentation, low efficiency, and high operating costs caused by information transmission delays and poor coordination at various ends of the fruit supply chain, such as the demand side and the supply side. Compared with the prior art, the beneficial effects of the fruit supply chain end-to-end collaborative management and control device provided in this application are the same as those of the fruit supply chain end-to-end collaborative management and control method provided in the above embodiments, and other technical features in the fruit supply chain end-to-end collaborative management and control device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0127] This application provides an end-to-end collaborative management and control device for a fruit supply chain. The end-to-end collaborative management and control device for a fruit supply chain includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the end-to-end collaborative management and control method for a fruit supply chain in the first embodiment described above.

[0128] The following is for reference. Figure 5The diagram illustrates a structural schematic suitable for implementing the end-to-end collaborative management and control device for the fruit supply chain in the embodiments of this application. The end-to-end collaborative management and control device for the fruit supply chain in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The fruit supply chain end-to-end collaborative management equipment shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments in this application.

[0129] like Figure 5 As shown, the end-to-end collaborative management and control equipment for the fruit supply chain may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to programs stored in ROM (Read Only Memory) 1002 or programs loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the end-to-end collaborative management and control equipment for the fruit supply chain. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the fruit supply chain end-to-end collaborative management equipment to exchange data with other devices wirelessly or via wired communication. Although the figure shows a fruit supply chain end-to-end collaborative management equipment with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0130] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0131] The fruit supply chain end-to-end collaborative management and control device provided in this application, employing the fruit supply chain end-to-end collaborative management and control method described in the above embodiments, can solve the technical problems of business fragmentation, low efficiency, and high operating costs caused by information lag and poor coordination at various ends of the fruit supply chain, such as the demand side and the supply side. Compared with the prior art, the beneficial effects of the fruit supply chain end-to-end collaborative management and control device provided in this application are the same as those of the fruit supply chain end-to-end collaborative management and control method provided in the above embodiments, and other technical features of the fruit supply chain end-to-end collaborative management and control device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0132] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0133] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0134] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the end-to-end collaborative management and control method for the fruit supply chain in the above embodiments.

[0135] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory or Flash Memory), optical fibers, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0136] The aforementioned computer-readable storage medium may be included in the end-to-end collaborative management and control equipment for the fruit supply chain; or it may exist independently and not be assembled into the end-to-end collaborative management and control equipment for the fruit supply chain.

[0137] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the fruit supply chain end-to-end collaborative management and control device, the fruit supply chain end-to-end collaborative management and control device: receives order information submitted by the store demand side, containing demand priority and timeliness tags, and encapsulates the order information into a new order event; sends the new order event to the warehousing and distribution end via a data link bus, so that the warehousing and distribution end pushes the procurement task to the purchasing agent's workbench, the procurement task being generated by the warehousing and distribution end based on the new order event and the procurement decision model; receives receiving data uploaded by the warehousing and distribution end, and according to the... The system generates a picking list and in-warehouse picking path from the receiving data; based on the picking list and in-warehouse picking path, it instructs the warehouse procurement and distribution end to sort the goods, provides feedback on the sorting completion confirmation data, and instructs the delivery fulfillment end to deliver the goods; it receives the electronic signature result containing real-time location trajectory from the delivery fulfillment end and sends the electronic signature result to the store demand end so that the store demand end can update the order status; based on the new order event, the procurement task, the receiving data, the sorting completion confirmation data, and the electronic signature result, it generates a financial settlement statement and pushes the financial settlement statement to the financial settlement end to complete collaborative management.

[0138] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0139] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0140] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0141] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., computer programs) for executing the aforementioned end-to-end collaborative management and control method for the fruit supply chain. This solves the technical problems of business fragmentation, low efficiency, and high operating costs caused by information delays and poor coordination at the demand and supply ends of the fruit supply chain. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the end-to-end collaborative management and control method for the fruit supply chain provided in the above embodiments, and will not be repeated here.

[0142] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described end-to-end collaborative management and control method for the fruit supply chain.

[0143] The computer program product provided in this application can solve the technical problems of business fragmentation, low efficiency, and high operating costs caused by information lag and poor coordination at the demand and supply sides of the fruit supply chain. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the end-to-end collaborative management and control method for the fruit supply chain provided in the above embodiments, and will not be repeated here.

[0144] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for end-to-end collaborative management of a fruit supply chain, characterized in that, The method comprises: receiving order information submitted by a store demand end, the order information containing demand priority and time limit label, and encapsulating the order information as a new order event; sending the new order event to a warehouse procurement end through a data link bus, so that the warehouse procurement end pushes a procurement task to a procurement staff workstation, the procurement task being generated by the warehouse procurement end according to the new order event and a procurement decision model; receiving receipt data uploaded by the warehouse procurement end, and generating a sorting list and an in-warehouse sorting path according to the receipt data; instructing the warehouse procurement end to sort goods according to the sorting list and the in-warehouse sorting path, feeding back sorting completion confirmation data, and instructing a distribution performance end to distribute the goods; receiving electronic receipt results containing real-time position trajectories fed back by the distribution performance end, and sending the electronic receipt results to the store demand end, so that the store demand end updates the order state; generating a financial settlement sheet according to the new order event, the procurement task, the receipt data, the sorting completion confirmation data and the electronic receipt results, and pushing the financial settlement sheet to a financial settlement end to complete collaborative management; the procurement decision model comprises a demand prediction module, an inventory health degree evaluation module and a supplier matching module; the procurement task is generated according to a sub-category fruit demand prediction value, a to-be-replenished gap and a supplier priority ranking; the supplier priority ranking is calculated by inputting supplier performance data and the time limit label into the supplier matching module by the warehouse procurement end through an NSGA-II multi-objective optimization algorithm with quality, time limit and cost as targets, and the supplier priority ranking contains a primary selected supplier, an alternative supplier and a replacement trigger condition; the to-be-replenished gap is obtained by inputting inventory data, a procurement quantity and fruit preservation period in fruit commodity data into the inventory health degree evaluation module by the warehouse procurement end; the sub-category fruit demand prediction value is obtained by inputting historical order data, a commodity SKU, the demand priority and a dynamic demand influence factor into the demand prediction module by the warehouse procurement end through an LSTM-XGBoost integrated algorithm for demand prediction; the fruit commodity data, the inventory data, the supplier performance data and the dynamic demand influence factor belong to associated data corresponding to the commodity SKU; the commodity SKU, the procurement quantity, the demand priority and the time limit label belong to demand parameters; the associated data is obtained by calling according to the demand parameters; the demand parameters are extracted from the new order event by the warehouse procurement end.

2. The method of claim 1, wherein, The step of receiving the receipt data uploaded by the warehouse procurement end, and generating the sorting list and the in-warehouse sorting path according to the receipt data comprises: receiving the receipt data uploaded by the warehouse procurement end, the receipt data containing quality detection results; updating inventory data according to the receipt data and generating a to-be-picked event associated with an order identifier of the new order event, the to-be-picked event containing a picking deadline determined by the time limit label of the new order event; aggregating to-be-picked orders corresponding to the to-be-picked event through a wave division algorithm to generate a picking list and an in-warehouse picking path.

3. The method of claim 2, wherein, The step of aggregating to-be-picked orders corresponding to the to-be-picked event through the wave division algorithm to generate a picking list and an in-warehouse picking path comprises: extracting the picking deadline and the commodity storage identifier in the to-be-picked event, and screening to-be-picked orders with the same preset time window and consistent commodity storage identifier to form a to-be-aggregated order set; dividing the to-be-picked orders in the same delivery area in the to-be-aggregated order set into a same sub-order set; aggregating the sub-order set into multiple wave tasks according to a preset wave capacity and commodity category similarity; allocating a picking resource identifier to each wave task, and generating a picking list according to the picking resource identifier and the commodity storage location in the wave task; calculating the shortest path from the picking station to each shelf according to the warehouse layout data and the commodity storage location in the picking list to generate an in-warehouse picking path.

4. The method of claim 3, wherein, The warehouse layout data comprises three-dimensional coordinates of each shelf, a channel width threshold, and a forbidden area identifier; The step of calculating the shortest path from the picking station to each shelf according to the warehouse layout data and the commodity storage location in the picking list to generate an in-warehouse picking path comprises: determining the target shelf coordinates corresponding to each commodity according to the three-dimensional coordinates of each shelf and the commodity storage location in the picking list; constructing a path planning matrix with the picking station coordinates as the starting point and the ending point, and each target shelf coordinate as the passing point; calculating the optimal access order of each passing point through the Dijkstra algorithm according to the channel width threshold, the forbidden area identifier, and the path planning matrix; generating an in-warehouse picking path containing shelf access order, path distance, and avoidance prompt, the in-warehouse picking path also marked with an alternative channel to avoid temporary forbidden areas.

5. The method of claim 1, wherein, The sub-category fruit demand prediction value is obtained according to a basic demand prediction value and a demand adjustment coefficient; The demand adjustment coefficient is obtained by fitting the numerical weight coefficient, the characteristic variable, and the historical demand value in the historical demand sequence of the same category fruit through the XGBoost algorithm; The basic demand prediction value is obtained by inputting the historical demand sequence into an LSTM network; The characteristic variable is converted according to the dynamic demand influence factor; The numerical weight coefficient is mapped from the demand priority; The historical demand sequence is screened from the historical order data according to the commodity SKU.

6. The method of any one of claims 1 to 5, wherein, The step of instructing the warehouse picking and delivery end to pick goods, feeding back picking completion confirmation data, and instructing the delivery performance end to deliver goods according to the picking list and the in-warehouse picking path comprises: The sorting order and the in-warehouse picking path are sent to the warehouse fulfillment end to make the warehouse fulfillment end sort the goods and feed back sorting completion confirmation data after sorting is completed, the sorting completion confirmation data including wave ID, actual sorting quantity and commodity quality review result; Based on the sorting completion confirmation data and wave information of the sorting order, a delivery trigger event including total weight of goods, delivery area and storage requirement is generated, the total weight of goods being calculated according to the actual sorting quantity; The delivery trigger event is pushed to the delivery performance end to make the delivery performance end plan a delivery route and deliver the goods.

7. A device for end-to-end collaborative management of a fruit supply chain, characterized in that, The device comprises: An order event module is configured to receive order information submitted by the store demand end and containing demand priority and time limit label, and encapsulate the order information as a new order event; A procurement task module is configured to send the new order event to the warehouse fulfillment end through a data link bus, so that the warehouse fulfillment end pushes a procurement task to a procurement clerk workstation, the procurement task being generated by the warehouse fulfillment end according to the new order event and a procurement decision model; the procurement decision model comprises a demand prediction module, an inventory health degree evaluation module and a supplier matching module; the procurement task is generated according to a sub-category fruit demand prediction value, a to-be-replenished gap and a supplier priority ranking; the supplier priority ranking is calculated by the NSGA-II multi-objective optimization algorithm with quality, time limit and cost as targets by inputting supplier performance data and the time limit label into the supplier matching module by the warehouse fulfillment end, and the supplier priority ranking contains a primary supplier, an alternative supplier and a replacement trigger condition; the to-be-replenished gap is obtained by inputting inventory data, procurement quantity and fruit shelf life in fruit commodity data into the inventory health degree evaluation module by the warehouse fulfillment end; the sub-category fruit demand prediction value is obtained by inputting historical order data, commodity SKU, the demand priority and a dynamic demand influence factor into the demand prediction module by the warehouse fulfillment end and performing demand prediction by the LSTM-XGBoost integrated algorithm; the fruit commodity data, the inventory data, the supplier performance data and the dynamic demand influence factor belong to associated data corresponding to the commodity SKU; the commodity SKU, the procurement quantity, the demand priority and the time limit label belong to demand parameters; the associated data is obtained by calling according to the demand parameters; the demand parameters are extracted from the new order event by the warehouse fulfillment end; A sorting path module is configured to receive receipt data uploaded by the warehouse fulfillment end and generate a sorting order and an in-warehouse picking path according to the receipt data; A sorting and delivery module is configured to instruct the warehouse fulfillment end to sort the goods according to the sorting order and the in-warehouse picking path, feed back sorting completion confirmation data, and instruct a delivery performance end to deliver the goods. The signing synchronization module is configured to receive the electronic signing result fed back by the delivery performance end and containing real-time position track, and send the electronic signing result to the store demand end, so that the store demand end updates the order state; The financial settlement module is configured to generate a financial settlement sheet according to the new order event, the purchase task, the receipt data, the sorting completion confirmation data and the electronic signing result, and push the financial settlement sheet to a financial settlement end, so as to complete the collaborative management and control.

8. A fruit supply chain end-to-end collaborative management device, characterized by, The device comprises a memory, a processor and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the fruit supply chain end-to-end collaborative management method according to any one of claims 1 to 6.

9. A storage medium, characterized by The storage medium is a computer readable storage medium, and the storage medium stores a computer program. When the computer program is executed by the processor, the steps of the fruit supply chain end-to-end collaborative management method according to any one of claims 1 to 6 are implemented.

Images