Intelligent logistics scheduling method and device and computer equipment

Abstract

This application relates to an intelligent logistics scheduling method, apparatus, and computer equipment. The method includes: in response to a resource scheduling trigger event, acquiring order data and / or capacity data; acquiring a disturbance configuration template, and adjusting the order data and / or capacity data accordingly based on the disturbance configuration template to obtain disturbance combination data; predicting a reference logistics backlog period and corresponding reference logistics backlog data based on the disturbance combination data; within the reference logistics backlog period, if a target reference logistics backlog data is determined based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period, then acquiring reference scheduling resources and a reference scheduling method corresponding to the target reference logistics backlog data for resource scheduling. This method can significantly improve the response speed of resource scheduling by predicting scheduling schemes in advance, thereby improving the stability of the system during peak hours.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to an intelligent logistics scheduling method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] With the development of data processing technology, it can be applied in most scenarios. For example, in the field of smart logistics, data processing technology can be used for intelligent scheduling of logistics.

[0003] Logistics systems may face periodic high concurrency challenges, necessitating solutions for these extreme scenarios. However, current traditional technologies often employ static rules or historical average loads for resource allocation, which can lead to either over- or under-allocation of resources. Summary of the Invention

[0004] Based on this, it is necessary to provide an intelligent logistics scheduling method, device, computer equipment, computer-readable storage medium, and computer program product that can provide resource allocation that matches the actual logistics situation and significantly improve the response speed of resource scheduling by predicting scheduling schemes in advance, thereby improving the stability of the system during peak periods.

[0005] Firstly, this application provides an intelligent logistics scheduling method, including:

[0006] In response to resource scheduling trigger events, obtain order data and transportation capacity data;

[0007] Obtain the disturbance configuration template, and adjust the order data and capacity data accordingly based on the disturbance configuration template to obtain the disturbance combination data;

[0008] Based on the perturbation combination data, the reference logistics backlog period and the reference logistics backlog data corresponding to the reference logistics backlog period are predicted; the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data are obtained and stored in the shared storage area;

[0009] If a target reference logistics backlog data is determined based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog data during the reference logistics backlog period, then the reference scheduling resources and reference scheduling methods corresponding to the target reference logistics backlog data are obtained from the shared storage area. Based on the reference scheduling resources and reference scheduling methods and the reference logistics backlog data, the target scheduling resources and target scheduling methods corresponding to the current logistics backlog data are determined, and the corresponding target scheduling methods are sent to the target scheduling resources for resource scheduling.

[0010] Secondly, this application also provides an intelligent logistics scheduling device, comprising:

[0011] The communication module is used to obtain order data and capacity data in response to resource scheduling trigger events;

[0012] The simulation module is used to obtain a disturbance configuration template, adjust the order data and capacity data accordingly according to the disturbance configuration template, and obtain disturbance combination data. Based on the disturbance combination data, the module predicts the logistics backlog period and the corresponding logistics backlog data. Based on the logistics backlog data, the module obtains the corresponding scheduling resources and scheduling methods.

[0013] The scheduling module is located on the same physical host as the simulation module. The simulation module and the scheduling module interact with each other through the shared memory area in the physical host. During the logistics backlog period, if the difference between the current logistics backlog data and the current logistics backlog data is within a preset range, the scheduling resource is invoked and scheduled according to the scheduling method.

[0014] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps described in the above method embodiments.

[0015] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps described in the above method embodiments.

[0016] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps described in the above method embodiments.

[0017] The aforementioned intelligent logistics scheduling method, device, computer equipment, computer-readable storage medium, and computer program product, in response to resource scheduling trigger events, adjust order data and capacity data according to disturbance configuration templates to obtain disturbance combination data. Based on the disturbance combination data, predictions are made to obtain reference logistics backlog periods and corresponding reference logistics backlog data. This allows for the prediction of potential logistics congestion periods and scales through disturbance simulation, thereby shortening the decision-making time when actual backlogs occur. Secondly, by adjusting order and capacity data according to disturbance configuration templates, multiple possible backlog scenarios are generated. This reduces scheduling failures caused by unforeseen disturbances during peak periods, improves the completeness of scheduling strategies, and ultimately enhances system stability and order fulfillment capabilities during peak hours. Finally, by comparing the current logistics backlog data with the reference logistics backlog data within the predicted reference logistics backlog period, the target reference logistics backlog data is determined. Then, based on the reference scheduling resources and reference scheduling methods corresponding to the target reference logistics backlog data, the target scheduling resources and target scheduling methods are determined. In this way, resource allocation with a high degree of matching with the current logistics status can be determined, reducing the occurrence of resource over-scheduling or resource under-scheduling. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an application environment diagram of the intelligent logistics scheduling method in one embodiment;

[0020] Figure 2 This is a flowchart illustrating an intelligent logistics scheduling method in one embodiment;

[0021] Figure 3 This is a schematic diagram of a heatmap of logistics nodes in one embodiment;

[0022] Figure 4 This is a structural block diagram of an intelligent logistics scheduling device in one embodiment;

[0023] Figure 5 This is an internal structural diagram of a computer device in one embodiment;

[0024] Figure 6 This is a diagram of the internal structure of a computer device in another embodiment. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0026] The intelligent logistics scheduling method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, the pressure scheduling integrated machine can obtain relevant data about logistics nodes (e.g., the number of packages arriving at the current node, the number of packages to be processed at the current node, and the number of packages processed at the current node) through the execution device controller. The execution device controller includes AGVs (Automated Guided Vehicles), parcel lockers, execution mobile terminals, and third parties (e.g., carrier API gateways). In the intelligent logistics scheduling system provided in this embodiment, a performance monitoring probe can also be additionally set up. This performance monitoring probe is used to monitor each execution device to collect the load information of each logistics node in real time. The load information of each logistics node collected by the performance monitoring probe can then be applied to the scheduling strategy generated by the pressure scheduling integrated machine. Furthermore, the pressure scheduling integrated machine also communicates with upstream systems (e.g., order systems, GPS terminals, human resources systems, etc.) to obtain order data and human resource data and perform resource scheduling operations. The pressure scheduling integrated machine can be a hybrid architecture of CPU (Central Processing Unit) / GPU (Graphics Processing Unit) / FPGA (Field-Programmable Gate Array). In this way, an intelligent logistics scheduling system integrating "perception → simulation → decision-making → execution → feedback" is formed.

[0027] The stress scheduling appliance comprises a communication module, a simulation module, and a scheduling module. The communication module can consist of an API (Application Programming Interface) gateway server and a data caching unit. The simulation module can consist of programmable logic devices and a fault injection unit. The scheduling module can consist of a GPU server and a redundancy policy library. Furthermore, the stress scheduling appliance can acquire order data and / or capacity data at various times through a monitoring module (such as the aforementioned performance monitoring probe). The monitoring module can consist of monitoring probes and an automatic switch controller. The stress scheduling appliance may also include a resource pool manager, composed of a resource controller and an external interface module, for communicating with scheduling resources. The simulation module and the scheduling module are housed in the same chassis and share memory via a high-speed PCIe (Peripheral Component Interconnect Express) bus to avoid network transmission latency and achieve microsecond-level data exchange.

[0028] In this embodiment, the "simulation module" is not a software program running on a general-purpose computer, but rather an embedded hardware module with a defined physical structure. For example, the simulation module is an industrial-grade edge computing box or a PCIe accelerator card. The simulation module includes a casing, a core chip, a storage unit, and physical interfaces. The casing is an IP65-rated aluminum alloy heat sink; the core chip's internal PCB board has an FPGA programmable logic device soldered on as the core computing unit for embedding stress test logic; the storage unit includes onboard DDR4 memory chips and Flash memory chips for storing test templates and cached data; the physical interfaces include an RJ45 Ethernet port (for connecting to a scheduling host), a DB9 serial port (for debugging), and a power interface (12VDC input).

[0029] The aforementioned components are physically connected via PCB traces and internal buses (such as PCIe and AXI buses). Specific pins of the FPGA chip are connected to the PHY chip of the Ethernet port via circuit traces, forming a defined data path. This hardware connection determines the physical path of data processing, rather than the logical path of software calls.

[0030] The simulation module and the elastic scheduling engine (corresponding to the scheduling module mentioned above) are connected via physical network cables or board gold fingers, and have a clear physical boundary.

[0031] In one exemplary embodiment, such as Figure 2 As shown, an intelligent logistics scheduling method is provided, which can be applied to... Figure 1Taking the integrated pressure dispatching machine as an example, the explanation includes the following steps 202 to 208. Wherein:

[0032] Step 202: In response to the resource scheduling trigger event, obtain order data and / or capacity data.

[0033] Among them, resource scheduling trigger events refer to various events that initiate the logistics resource scheduling process. These events can be periodic (such as at a fixed time every day), event-based (such as detecting that the logistics backlog threshold has been exceeded, or the sudden launch of a major promotion), or triggered by manual instructions.

[0034] Order data is a collection of data reflecting logistics tasks awaiting delivery or processing, including but not limited to order volume, order distribution, estimated delivery time, goods type, customer location, etc. For example, order data could be the total order data for the previous day, or the total order data for a period of time after a resource scheduling trigger event occurs, such as the total order data for one hour after the trigger event.

[0035] Transportation capacity data is a collection of data reflecting currently available logistics resources, including but not limited to the number of vehicles, vehicle location data, and employee work schedules.

[0036] For example, a resource scheduling trigger event is triggered daily. In response to the resource scheduling trigger event, the above-mentioned communication module can be used to obtain order data within a specified time window from the order management system and capacity data from the transportation management system or logistics node management platform through interface calls, database queries or message queues. The captured order data and capacity data are then standardized into a unified format and stored in a cache (e.g., Redis cache) for subsequent steps to read and use.

[0037] Optionally, resource scheduling events can also be triggered based on specific events. For example, a resource scheduling event can be triggered upon receiving a promotional email. The stress scheduling appliance can monitor the inbox of the target email address through a monitoring module. When an email containing target keywords (such as "lockdown," "power outage," or "promotion") is detected in the inbox, the natural language processing module is triggered. This module performs deep analysis of the email content, extracting the corresponding event type, affected area, and affected time. Based on the extracted event type, affected area, and affected time, the corresponding order data and capacity data are then obtained.

[0038] Step 204: Obtain the disturbance configuration template, and adjust the order data and / or capacity data accordingly according to the disturbance configuration template to obtain the disturbance combination data.

[0039] The disturbance configuration template is a predefined data structure or rule file used to describe various abnormal or changing scenarios that may affect the logistics system. Each template contains a set of adjustment parameters, such as the percentage increase in order volume, the percentage decrease in transport capacity, changes in order concentration in a specific region, and traffic delay time. Different disturbance configuration templates can be used for different resource scheduling trigger events. That is, the corresponding disturbance configuration template is determined based on the trigger type of the resource scheduling trigger event. For example, the percentage increase in order volume in the disturbance configuration template corresponding to a daily scheduled resource scheduling trigger event may be lower than the percentage increase in order volume in the disturbance configuration template corresponding to a resource scheduling trigger event triggered by a major promotional email. Similarly, the percentage decrease in transport capacity in the disturbance configuration template corresponding to a resource scheduling trigger event triggered by a major promotional email may be lower than the percentage decrease in transport capacity in the disturbance configuration template corresponding to a resource scheduling trigger event triggered by a lockdown email or power outage email. Therefore, a matching disturbance configuration template can be selected based on the trigger type of the resource scheduling trigger event, so that the reference logistics backlog data predicted based on the disturbance configuration template is closer to the actual backlog situation.

[0040] Among them, adjustment refers to performing mathematical transformations, filtering, expansion and contraction, or adding noise on the original order data or capacity data according to the parameters in the disturbance configuration template, in order to simulate the data state after the disturbance occurs.

[0041] Among them, disturbance combined data refers to a complete set of data obtained by merging the adjustment results of order data and capacity data using disturbance configuration templates, representing the order and capacity status under a specific disturbance scenario.

[0042] For example, an intelligent logistics scheduling system can maintain a disturbance template library, where each disturbance configuration template can be defined by a set of key-value pairs or structured rules. For instance, a disturbance configuration template corresponding to a "major promotion" event could be defined as: total order volume increases by 200%, the proportion of orders from the East China region increases by 40%, and available capacity time is reduced by 10%. Based on the selected disturbance configuration template, operations such as adjusting order magnitude and regional weights are performed on the order data, and operations such as reducing capacity and compressing service time windows are performed on the capacity data. The adjusted order data and capacity data together constitute the disturbance combination data corresponding to that disturbance configuration template.

[0043] Optionally, multiple disturbance configuration templates matching the resource scheduling trigger event can be obtained, and the order data and capacity data can be adjusted accordingly according to each disturbance configuration template to obtain multiple disturbance combination data. It can be understood that the adjustment trend of the multiple matched disturbance configuration templates is consistent (i.e., simulating logistics congestion and backlog by reducing scheduling resources and increasing pending logistics volume), the difference lies in the degree of adjustment.

[0044] In some embodiments, a disturbance configuration template is obtained, and the order data and capacity data are adjusted accordingly based on the disturbance configuration template to obtain disturbance combination data. Simulations using this obtained disturbance combination data can simulate potentially complex situations in actual logistics by injecting composite disturbances.

[0045] Step 206: Based on the disturbance combination data, predict the reference logistics backlog period and the reference logistics backlog data corresponding to the reference logistics backlog period; obtain the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data and store them in the shared storage area.

[0046] The reference logistics backlog period refers to the predicted time interval for logistics backlog under a specific combination of disturbance data. For example, "14:00 on November 11, 2025 to 18:00 on November 11, 2025".

[0047] Reference logistics backlog data refers to quantitative indicator data corresponding to the reference logistics backlog period, such as a list of backlog sites, the expected number of backlog orders for each backlog site, the average order delay time, bottleneck heat map, order backlog growth rate, etc.

[0048] Reference scheduling resources refer to the specific logistics resources that are used to make decisions regarding application backlog situations. Examples include scheduling trucks, delivery personnel, sorting personnel, and sorting machines. Reference scheduling resources can also include third-party platforms, execution paths (such as delivery paths and sorting paths), and drone landing station controllers.

[0049] A reference scheduling method is a specific execution plan for scheduling resources, such as "how many sorting personnel to allocate to backlog sites" or "how many additional delivery personnel to perform deliveries." A reference scheduling method can also be a communication method with reference scheduling resources, such as communicating with a third-party platform via an industrial interface gateway, or communicating with the AGV main control PLC (Programmable Logic Controller) via a CAN (Controller Area Network) bus to execute path changes (e.g., allocating industrial robots), or controlling the drone take-off and landing station controller via a 4G module to initiate drone flight.

[0050] A shared storage area is a data storage area that can be accessed by multiple modules, services, or nodes. It can be a distributed cache, database, or object storage, used to persistently save reference plans.

[0051] For example, the disturbance combination data is input into the simulation module to simulate the entire process of order dispatch, vehicle travel, and delivery completion over a future period. The simulation module outputs the number of backlogged orders at each time point, identifies the time window when the backlog continues to rise or reaches a preset backlog level (e.g., the number of backlogged orders exceeds a preset backlog threshold) as a reference logistics backlog period, and records the maximum backlog amount, average order delay time, and other indicators corresponding to this reference logistics backlog period as the corresponding reference logistics backlog data. Subsequently, for this reference logistics backlog period, operations research optimization algorithms (e.g., integer programming, heuristic search) are used to solve for the resource scheduling scheme with the minimum cost or the shortest time to resolve the backlog, obtaining the corresponding scheduling resources and scheduling method. Finally, the scheduling resources, the scheduling method, the corresponding reference logistics backlog period, and the reference logistics backlog data are associated and stored in a shared storage area.

[0052] In other words, this step is used to conduct stress tests based on the combined disturbance data. By simulating multi-dimensional extreme scenarios (such as order surges, road closures, and personnel shortages), it generates early warnings and corresponding scheduling resource plans. This allows the scheduling resource plans to be automatically activated when the actual business load approaches the early warning threshold in real-world situations, thus completing the transformation from "passive response" to "active defense" and significantly improving system stability and order fulfillment capabilities during peak periods.

[0053] Alternatively, a neural network model can be trained in advance. By inputting the obtained order data and transportation capacity data into the trained neural network model, the neural network model outputs the predicted reference logistics backlog periods and the corresponding reference logistics backlog data.

[0054] In some embodiments, a bottleneck heatmap is obtained by analyzing the number of backlogged orders at each logistics node at each time point output by the simulation module; then, a reference logistics backlog period is identified based on this bottleneck heatmap. For example, a reference... Figure 3 Logistics node numbers are used to identify logistics nodes. For example, number 1 is used to identify East China Warehouse A, number 2 is used to identify highway entrance B, number 3 is used to identify distribution station C, and so on. Figure 3 Different shades of gray are used to represent different probabilities of logistics backlog. Figure 3 The peak position in the bottleneck heatmap corresponds to the highest probability of logistics backlog. The corresponding time will be included in the reference logistics backlog period. For example, based on this bottleneck heatmap, each time when the probability of logistics backlog is greater than a preset probability condition can be included in the reference logistics backlog period. Optionally, there can be multiple reference logistics backlog periods, and the minimum time difference between different reference logistics backlog periods should be greater than or equal to a preset time difference.

[0055] Step 208: If, within the reference logistics backlog period, a target reference logistics backlog data is determined based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period, then the reference scheduling resources and reference scheduling methods corresponding to the target reference logistics backlog data are obtained from the shared storage area. Based on the reference scheduling resources and reference scheduling methods and the reference logistics backlog data, the target scheduling resources and target scheduling methods corresponding to the current logistics backlog data are determined, and the corresponding target scheduling methods are sent to the target scheduling resources for resource scheduling.

[0056] Among them, the current logistics backlog data refers to the logistics backlog status data monitored in real time during actual operation, such as the number of undelivered or unsorted orders at each logistics node at the current moment, the queue length at each station, and the growth rate of backlog orders.

[0057] The gap refers to a measure of the similarity or difference between current logistics backlog data and reference logistics backlog data. For example, the gap includes the difference between the current logistics backlog volume and the reference logistics backlog volume, or the difference between the average order delay time in the current logistics backlog data and the average order delay time in the reference logistics backlog data. The gap can also include the difference between the current logistics backlog status score for a specific site and the reference logistics backlog status score, or the similarity between the current logistics backlog data and the reference logistics backlog data for a specific site (which can be obtained by calculating the differences between each indicator data in the current logistics backlog data and each reference indicator data in the reference logistics backlog data).

[0058] Target scheduling resources refer to the actual set of resources that need to be utilized after adapting and adjusting the reference scheduling resources based on the current logistics backlog data. For example, if the reference logistics backlog data and scheduling resources are "allocate 5 sorters when there are 200 backlogged orders," then if the current logistics backlog data shows 160 backlogged orders, it can be adjusted to add 4 sorters. This allows for the allocation of appropriate scheduling resources and reduces the occurrence of resource over-scheduling.

[0059] The target scheduling method refers to the specific scheduling instructions actually executed after adapting and adjusting the reference scheduling method based on the current logistics backlog data.

[0060] For example, at each moment during the reference logistics backlog period, the node input and output of each logistics node are collected, and the current logistics backlog data for each logistics node is determined based on their respective node input and output. The following explanation focuses on a specific logistics node: The reference logistics backlog data for the current logistics node during the reference logistics backlog period is obtained. After collecting the current logistics backlog data, it is determined whether the current logistics backlog data is greater than a preset backlog threshold. If the current logistics backlog data is greater than the preset backlog threshold, the difference between the current logistics backlog data and each reference logistics backlog data is calculated, and the reference logistics backlog data corresponding to the smallest difference is determined as the target reference logistics backlog data. Subsequently, the reference scheduling resources and reference scheduling methods corresponding to the target reference logistics backlog data are obtained from the shared storage area, and the target scheduling resources and target scheduling methods matching the current logistics backlog data are determined based on the reference scheduling resources and reference scheduling methods. Finally, the target scheduling method is sent to the determined target scheduling resources to schedule the target scheduling resources according to the target scheduling method. For example, a task notification can be sent to five selected sorters to indicate that they have arrived at a logistics station to sort orders, or a task notification can be sent to five deliverymen to indicate that they have arrived at a logistics station to deliver orders.

[0061] In the aforementioned intelligent logistics scheduling method, in response to resource scheduling trigger events, order data and capacity data are adjusted according to a disturbance configuration template to obtain disturbance combination data. Based on this disturbance combination data, predictions are made to obtain reference logistics backlog periods and corresponding reference logistics backlog data. This allows for the prediction of potential logistics congestion periods and scales through disturbance simulation, thereby shortening the decision-making time when actual backlogs occur. Secondly, by adjusting order and capacity data according to the disturbance configuration template, multiple possible backlog scenarios are generated. This reduces scheduling failures caused by unforeseen disturbances during peak periods, improves the completeness of the scheduling strategy, and enhances the system's stability and order fulfillment capabilities during peak periods. Finally, by comparing the current logistics backlog data with the reference logistics backlog data within the predicted reference logistics backlog period, a target reference logistics backlog data is determined. Then, based on the reference scheduling resources and methods corresponding to the target reference logistics backlog data, target scheduling resources and methods are determined. This ensures that resource allocation with a high degree of matching with the current logistics status is identified, reducing the occurrence of resource over-scheduling or under-scheduling.

[0062] In one exemplary embodiment, such as Figure 3As shown, the steps for predicting the reference logistics backlog period and the corresponding reference logistics backlog data based on the perturbation combination data include: determining the logistics transportation period based on resource scheduling trigger events; simulating the perturbation combination data to predict the simulation index data of each logistics node at each moment within the logistics transportation period, and determining the simulated logistics backlog information of each logistics node at each moment based on the corresponding simulation index data; the simulated logistics backlog information is used to characterize whether the logistics node is in a logistics backlog state at a given moment; determining the reference logistics backlog period based on the simulated logistics backlog information of each logistics node at each moment; and determining the corresponding reference logistics backlog data based on the simulation index data corresponding to the reference logistics backlog period.

[0063] The logistics transportation period refers to the time period that needs to be predicted (or simulated). In other words, the simulation module predicts the relevant data for the corresponding logistics transportation period based on the obtained order and capacity data. For example, if the resource scheduling trigger event is a daily scheduled event, the logistics transportation period could be the day following the trigger time. If it's a weekly scheduled event, the logistics transportation period could be the week following the trigger event. As another example, if the resource scheduling trigger event is an email trigger event, the logistics transportation period could be the period extracted based on natural language parsing of the email content. For instance, it might be necessary to predict the bottleneck heatmap for the three days from January 1st to January 3rd.

[0064] Simulation index data refers to specific numerical indicators calculated during the simulation process to measure the operational status of logistics, such as the parcel processing speed of each logistics node, vehicle utilization rate, and the length of the queue of pending orders.

[0065] Simulated logistics backlog information is a state conclusion derived from simulation index data. It is used to mark whether a logistics node is in a "backlog" state at a specific moment (e.g., if the queue length is greater than the threshold, it is a backlog).

[0066] Specifically, the simulation process described above is executed using a simulation module. The simulation module consists of a programmable logic device (FPGA or ASIC) and a fault injection unit (software-defined load generator).

[0067] After identifying a resource scheduling trigger event, the logistics transportation period is determined based on the trigger type of the resource scheduling trigger event in order to predict the logistics operation status of each logistics node during the logistics transportation period, thereby identifying the possible backlog situation of each logistics node.

[0068] An interface module exists on the programmable logic device (PLD). When a resource scheduling event is triggered, the PLD receives a simulation signal. Internally, a "global simulation clock counter" is established. The scale between the global simulation clock counter and real time is set. For example, 1 millisecond of PLD operation represents 1 minute of real-world operation. The PLD sets the termination value of the global simulation clock counter based on previously derived logistics transportation time periods. Then, the global simulation clock counter is started. When order data and capacity data flow into the PLD, they need to pass through a fault injection unit. The fault injection unit processes the order data and capacity data based on the perturbation combination data. For example, if "orders increase by 30%" is configured, the fault injection unit cyclically replicates the corresponding number of data packets. If "capacity decreases by 20%" is configured, the fault injection unit can intercept 20% of the "vehicle permission signals" at the hardware circuit level. Inside the PLD, dozens or even hundreds of node processing modules (corresponding to various logistics nodes) operate in parallel. Each node processing module has a hardware counter. When an input port detects a package arrival signal, the input counter increments by 1. When an output port detects a package output signal, the output counter increments by 1. The current backlog amount equals the previous backlog amount plus the input counter value minus the output counter value. This is used to predict the simulation index data of each logistics node at each moment within the logistics transportation period, and to determine the simulated logistics backlog information of each logistics node at each moment based on the corresponding simulation index data. Next, a recording module is stored in the programmable logic device to determine and record the reference logistics backlog period and the corresponding reference logistics backlog data. The recording module detects the backlog signal of each logistics node. When the backlog signal changes from 0 to 1, a hardware-triggered latch records the current global simulation time, designated as the backlog start time. When the backlog signal changes from 1 to 0, the current global simulation time is recorded again, designated as the backlog end time. Thus, the reference logistics backlog period is determined based on the backlog start and end times. When the backlog signal is detected to change to 1, a "maximum value tracker" can also be activated to continuously compare and update the current maximum backlog amount. When the logistics node exits the backlog state or the simulation ends, the programmable logic device packages the relevant data of each logistics node (backlog period, backlog data, etc.) for generating heat maps or scheduling schemes.

[0069] In other embodiments, since logistics backlogs may be a dynamic process (e.g., from initial development to outbreak and then to decline), different scheduling schemes can be used for different stages to allocate resources more rationally. Specifically, the simple comparator in the node processing module is upgraded to a more complex logic circuit. The output status code of the node processing module is designed according to preset backlog grading thresholds. For example, if there are 3 backlog grading thresholds, the output status code consists of two bits. If there are 7 backlog grading thresholds, the output status code can consist of three bits. Taking 3 backlog grading thresholds as an example, mild backlog: backlog amount greater than 1000 and less than 3000; moderate backlog: backlog amount greater than 3000 and less than 5000; severe backlog: backlog amount greater than 5000. When the backlog amount jumps between different intervals, the programmable logic device records the duration of each interval and the corresponding maximum backlog amount. Specifically, if the backlog is mild, the output status code is "00"; if the backlog is moderate, the output status code is "01"; and if the backlog is severe, the output status code is "10". This refines the reference logistics backlog period, making the simulated reference logistics backlog period more consistent with reality and more detailed and accurate.

[0070] In this embodiment, by performing fine-grained simulation and index analysis on the combined disturbance data during the logistics transportation period, potential and vague risks can be transformed into specific and quantifiable time windows and backlog values, thereby achieving an accurate "profile" of future backlog situations. This accurate prediction enables the scheduling system to identify high-risk periods and nodes before the actual backlog occurs, providing solid data support for subsequent advance scheduling resource reserves and the development of targeted solutions, effectively solving the inefficiency problem caused by response delays in traditional scheduling.

[0071] In some embodiments, based on perturbation combination data, a reference logistics backlog period and the reference logistics backlog data corresponding to the reference logistics backlog period are predicted, including: determining the logistics transportation period based on resource scheduling trigger events; performing a preset number of simulations on the perturbation combination data, predicting the simulation index data of each logistics node at each moment within the logistics transportation period in each simulation, determining the simulation logistics backlog information of each logistics node at each moment based on the corresponding simulation index data, so as to obtain the simulation results of each simulation; obtaining a logistics node heatmap corresponding to the logistics transportation period based on each simulation result; the logistics node heatmap is used to characterize the correlation between the logistics backlog probabilities of each logistics node at each moment within the logistics transportation period; querying the target logistics backlog probability that meets the preset probability conditions in the logistics node heatmap, and determining the reference logistics backlog period related to the target logistics backlog probability; and determining the corresponding reference logistics backlog data based on the statistical data corresponding to the simulation index data corresponding to the reference logistics backlog period.

[0072] The preset number of simulations refers to performing multiple simulations to improve the accuracy of the prediction. The simulation result for each simulation refers to the output result after a single simulation run, which includes the specific state of each logistics node at each time point under that simulation scenario (such as whether there is backlog).

[0073] A logistics node heatmap is a visual or data-driven probability distribution map generated based on the statistical results of multiple simulations. It does not show the definitive outcome of a single simulation, but rather the "probability" of backlog at each logistics node at each time point.

[0074] Logistics backlog probability refers to the frequency with which a certain node experiences backlog at a certain moment in multiple simulations (e.g., backlog occurs 80 times out of 100 simulations, with a probability of 80%).

[0075] The preset probability condition is a threshold standard for screening high-risk periods (e.g., backlog probability greater than 90%). Only those periods that meet this condition are identified as "reference logistics backlog periods" that require attention.

[0076] The target logistics backlog probability is a probability value that meets the preset probability conditions and is found in the heat map. It is used to locate the specific backlog time period.

[0077] Specifically, multiple simulations are performed according to the above embodiments to obtain simulation results. Subsequently, the simulation module can send the simulation results to the upper-layer software, which then statistically analyzes the frequency of logistics backlogs in different time periods to generate a logistics node heatmap (see reference). Figure 3 Next, query the target logistics backlog probability greater than the preset probability threshold in the logistics node heatmap, determine the time period corresponding to these target logistics backlog probabilities as the reference logistics backlog time period, and determine the statistical data (or aggregated data) of the simulation index data of each simulation corresponding to this time period as the reference logistics backlog data.

[0078] In this embodiment, by simulating the combined disturbance data a preset number of times and generating a heatmap representing the backlog probability, this method can effectively eliminate the uncertainty caused by random interference or model bias in a single prediction, making implicit risks explicit into specific probability values. This allows the system to accurately pinpoint high-risk periods with extremely high probability of occurrence, filter out invalid or occasional false alarms, and thus provide a statistically based and highly reliable decision-making basis for subsequent scheduling strategies, greatly improving the accuracy of resource scheduling.

[0079] In some embodiments, the reference logistics backlog data includes the total backlog of parcels. Obtaining reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data and storing them in a shared storage area includes: obtaining the target evacuation duration; obtaining the current total resources of the logistics nodes corresponding to the reference logistics backlog period; calculating the total number of evacuable parcels within the target evacuation duration based on the current total resources; obtaining the number of reference scheduling resources based on the positive difference between the total backlog of parcels and the total number of evacuable parcels, and the number of evacuable parcels corresponding to a single scheduling resource; and obtaining the target policy rule corresponding to the reference scheduling resource from a preset policy rule base to obtain the reference scheduling method corresponding to the reference scheduling resource.

[0080] The target evacuation time, which can be preset by the system or dynamically determined, is the expected length of time to completely process the backlog of packages and is a key indicator for measuring the urgency of scheduling. For example, the target evacuation time can be determined based on the total number of backlogged packages. The larger the total number of backlogged packages, the longer the corresponding target evacuation time may be. In practical applications, it may also be determined based on the total number of backlogged packages and the order information corresponding to these backlogged packages. For example, the average delivery time corresponding to the backlogged packages can be obtained based on the order information of the backlogged packages, and then the target evacuation time can be determined based on the average delivery time. Different target evacuation times can be pre-configured for different delivery times, thus obtaining the corresponding target evacuation time based on the average delivery time.

[0081] The current total resources refer to the total transportation capacity resources (such as the number of existing vehicles, sorting staff, etc.) that a logistics node currently has that can be used to handle the backlog during the reference logistics backlog period.

[0082] The total number of evacuable packages refers to the maximum number of packages that the existing resources can theoretically handle within a limited target evacuation time.

[0083] The positive difference is used to quantify the magnitude of the gap between supply and demand.

[0084] The preset policy rule base stores a database of operation specifications corresponding to different orders of magnitude and different types of scheduling resources, which is used to transform the number of resources into specific execution plans.

[0085] Specifically, first, the configuration interface is called to obtain the "target evacuation duration" (e.g., set to 2 hours), and the "current total resources" of the logistics node during the current reference logistics backlog period are obtained (e.g., 10 existing sorters, with a sorting capacity equivalent to 2000 pieces / hour). The total number of parcels that can be evacuated within the specified target evacuation duration is calculated based on the product of the current total resources and the target evacuation duration (e.g., 2000 pieces / hour × 2 hours = 4000 pieces). Assuming the total backlog of parcels is 5000 pieces and the total number of evacuable parcels is 4000 pieces, the positive difference between the two is calculated to be 1000. Then, combined with the "number of evacuable parcels corresponding to a single scheduling resource" (e.g., one sorter can handle 100 pieces per hour), a division operation is used to determine the required resource quantity as 1000 ÷ 100 = 10. Finally, it is determined that 10 sorters need to be scheduled. The scheduling resources corresponding to sorters can be determined based on the number of sorters. For example, if one sorting machine is configured for 5 sorters, then 2 more sorting machines need to be scheduled to that logistics node. Next, the target policy rule corresponding to the reference scheduling resource can be obtained from a preset policy rule base to obtain a reference scheduling method corresponding to the reference scheduling resource. For example, the communication method for communicating with the reference scheduling resource. The preset policy rule base stores the scheduling methods corresponding to various scheduling resources. In some embodiments, the corresponding scheduling method can also be determined based on the type of scheduling resource and the current logistics backlog data (e.g., the identifier of the logistics node in a backlog state, the backlog amount, etc.).

[0086] In some other embodiments, the total backlog of packages may be less than the total number of packages that can be evacuated. In this case, the positive difference between the two is zero. Based on the obtained positive difference, the number of resources to be scheduled is determined to be zero, that is, no resource scheduling is performed.

[0087] In other words, if the total backlog of packages is greater than the total number of packages that can be evacuated, the number of reference scheduling resources obtained is greater than zero; if the total backlog of packages is less than or equal to the total number of packages that can be evacuated, the number of reference scheduling resources obtained is zero, that is, no related resource scheduling is performed.

[0088] In this embodiment, a time-constrained resource calculation model is established by setting a target evacuation duration and quantifying the processing potential of existing resources, achieving a precise mapping from the abstract "backlog situation" to the specific "resource quantity". The use of positive interpolation calculation ensures the non-negativity of resource demand, and combined with the automatic generation of scheduling methods from a strategy rule base, effectively avoids the blindness of manual estimation, thereby improving the accuracy and timeliness of logistics scheduling in responding to sudden backlogs.

[0089] In some embodiments, order data includes order volume, and capacity data includes data related to transportation routes and manpower. The order data and capacity data are adjusted according to a perturbation configuration template to obtain perturbation combination data, including: obtaining a first multiple interval corresponding to the order data and a second multiple interval corresponding to the capacity data from the perturbation configuration template; the first multiple in the first multiple interval is greater than one, and the second multiple in the second multiple interval is less than one; randomly selecting a first multiple from the first multiple interval, and expanding the order volume based on the first multiple to obtain a corresponding target order volume; randomly selecting a second multiple from the second multiple interval, and reducing the number of available transportation routes and / or the number of available manpower based on the second multiple to obtain corresponding target route data and / or target manpower data; and obtaining perturbation combination data based on the target order volume, target route data, and / or target manpower data.

[0090] Specifically, the perturbation configuration template is first parsed to extract key parameter ranges corresponding to order data and transportation capacity data. For example, the order quantity adjustment parameter is read as the first multiple range [1.5, 2.0], the order weight adjustment parameter as the third multiple range, and the available transportation route adjustment parameter or the available transportation manpower and available sorting manpower adjustment parameter as the second multiple range [0.6, 0.8]. These two ranges represent the "upper limit of demand-side pressure" and the "lower limit of supply-side pressure," respectively. The perturbation configuration template sets the adjustment parameters corresponding to the data types that need to be adjusted. This allows the data to be perturbed. Subsequently, a random number generator is used to randomly select a value from the first multiple range as the first multiple (e.g., randomly selecting 1.8). Then, the order quantity in the original order data is multiplied by 1.8. Assuming the original order quantity is 1000, the target order quantity is 1800. Similarly, a value is randomly selected from the second multiple range as the second multiple (e.g., randomly selecting 0.7). The number of transport routes and the number of transport personnel are each multiplied by 0.7. Assuming there are originally 100 routes and 50 drivers, the adjusted data yields 70 target routes and 35 target personnel. These 1800 orders, 70 routes, and 35 personnel are then packaged together to form a complete set of "perturbation combination data." This data describes a high-pressure scenario of "orders doubling but transport capacity decreasing," and is fed into subsequent modules for simulation and prediction.

[0091] In some other embodiments, the disturbance configuration template may only adjust order data or only capacity data. Specifically, the disturbance configuration template may also only adjust available transportation routes or only available transportation manpower or available sorting manpower. Different degrees of adjustment may also be applied to different data types; that is, the corresponding multiple ranges may be different.

[0092] In this embodiment, by setting a first multiple greater than one and a second multiple less than one, and introducing a random selection mechanism, this method can flexibly construct extreme supply and demand imbalance scenarios where "demand surge" and "capacity reduction" occur simultaneously. Randomness ensures the diversity of each simulation scenario, avoids overfitting to a single fixed scenario, and thus can comprehensively cover various combinations of sudden risks that may occur in real logistics, effectively verifying the risk resistance capability of the logistics system under extreme pressure and the robustness of the scheduling strategy.

[0093] In some embodiments, determining the target scheduling resources and target scheduling method corresponding to the current logistics backlog data based on scheduling resources and scheduling methods includes: calculating the ratio between the current logistics backlog amount in the current logistics backlog data and the reference logistics backlog amount in the reference logistics backlog data; and determining the target scheduling resources based on the ratio and scheduling resources.

[0094] For example, the ratio of the current logistics backlog to the reference logistics backlog can be calculated, and the required number of target scheduling resources can be determined based on this ratio. It can be understood that target scheduling resources include, for example, sorters, delivery personnel, sorting machines, etc.

[0095] In some other embodiments, the required amount of target scheduling resources can also be determined based on the ratio between the current average order delay time and the reference average order delay time.

[0096] In this embodiment, by calculating the ratio between the current actual data and the predicted data, and dynamically adjusting the scheduling resources accordingly, the problem of discrepancies between the prediction model and the actual execution can be effectively solved. This avoids resource waste when the actual situation is better than the prediction, and also prevents the risk of insufficient resource investment when the actual situation is worse than the prediction, thereby achieving refined scheduling decisions and optimized resource utilization.

[0097] In some embodiments, determining the target reference logistics backlog data based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period includes: calculating the difference between the current logistics backlog data and each reference logistics backlog data corresponding to the reference logistics backlog period; selecting the smallest difference from the differences; if the smallest difference meets a preset difference condition, then the reference logistics backlog data corresponding to the smallest difference is determined as the target reference logistics backlog data. If the smallest difference does not meet the preset difference condition, the target reference logistics backlog data cannot be determined, and subsequent steps are not performed. That is, the current logistics backlog data has not yet reached the point where resource scheduling is required.

[0098] The current logistics backlog data can include the backlog amount and the rate of backlog growth.

[0099] Specifically, after obtaining the current logistics backlog data, since one reference logistics backlog period may correspond to multiple reference logistics backlog data, and the scheduling resources collected for different backlog states will also be different, the difference between the current logistics backlog data and each reference logistics backlog data can be calculated. If the current logistics backlog data (or reference logistics backlog data) consists of multiple types of data, the similarity between the current logistics backlog data and each reference logistics backlog data can also be calculated. That is, the target reference logistics backlog data is determined by comprehensively considering multiple types of data (such as the backlog amount and backlog growth rate mentioned above). It can be understood that if the current logistics backlog data does not reach the target level of backlog state (e.g., only a few packages are backlogged), no additional resources need to be scheduled. Therefore, in order to determine accurate resources and reduce resource waste, it is also necessary to determine whether the difference between the current logistics backlog data and the target reference logistics backlog data reaches a preset difference condition. If the difference does not meet the preset difference condition or the similarity does not reach the preset similarity threshold, the relevant resource scheduling will not be performed temporarily. If the minimum difference reaches the preset difference condition or the similarity reaches the preset similarity threshold, the subsequent steps are executed.

[0100] In this embodiment, by calculating the difference between actual data and multiple sets of predicted data and selecting the minimum value, precise alignment between simulation prediction and real-time decision-making is achieved. By introducing a preset gap condition as a "safety valve," the risk of blindly applying the solution when the actual situation deviates significantly from the prediction is effectively avoided. This ensures the speed of scheduling decision response while significantly improving the accuracy of decision-making and robustness to abnormal scenarios.

[0101] In one exemplary embodiment, the intelligent logistics scheduling system consists of five functional modules, with data processing strictly following the sequential logic of "routine data collection → timed simulation → risk identification → contingency plan generation → real-time activation". The five functional modules are: real-time data access module (corresponding to the communication module), stress test simulator (corresponding to the simulation module), elastic scheduling engine (corresponding to the scheduling module), resource pool manager, and performance monitoring and activation module (corresponding to the monitoring module). The real-time data access module comprises an API gateway server (e.g., Dell PowerEdge R650) and a data caching unit (e.g., a Redis cluster). The input data for the real-time data access module includes ERP / TMS order flows (corresponding to order data), vehicle location data (corresponding to a portion of the capacity data), and employee shift schedules (corresponding to a portion of the capacity data). The real-time data access module captures upstream system data (e.g., order system, human resources system, etc.) in real time and stores the captured upstream system data in the Redis cache for subsequent retrieval. The output data of the real-time data access module includes standardized event streams (i.e., standardized formatted order data and capacity data). The stress test simulation module consists of programmable logic devices and a fault injection unit (software-defined load generator). The input data for the stress test simulation module includes basic business data and user-configured stress templates (corresponding to disturbance configuration templates). The stress test simulation module starts the simulation task periodically (e.g., at 2 AM every day) and simulates the load by injecting composite disturbances (order amplification: multiplied by 5; route blocking: randomly closing 20% ​​of main roads; manpower reduction: removing drivers based on historical absenteeism rate × 1.2). The frequency of bottleneck occurrence is statistically analyzed by running 100 Monte Carlo simulations. The output data of the stress test simulator includes bottleneck heatmaps, emergency resource requirement lists, and key scenario identification codes. The elastic scheduling engine consists of GPU servers and a redundancy policy library (stored on solid-state drives). The input data for the elastic scheduling engine includes the bottleneck heatmap and emergency resource requirement list output by the simulator. The elastic scheduling engine executes the following steps: 1. Parse the heatmap (corresponding to the logistics node heatmap) to locate the highest probability node and time window (corresponding to the reference logistics backlog period); 2. Find the emergency resource demand list corresponding to the node (corresponding to the scheduling resource); 3. Match the policy ID (corresponding to the scheduling method) in the redundancy policy library; 4. Extract the instruction sequence and generate an executable instruction package (used to communicate with the scheduling resource). It can be understood that the output data of the elastic scheduling engine includes standardized instructions (including target device address, protocol type, parameter values, etc.). The resource pool manager consists of a physical resource controller (industrial control computer) and an external interface module. The input data of the resource pool manager is the instruction package issued by the elastic scheduling engine.The resource pool manager can, for example, send "reserve riders" instructions to the crowdsourcing platform interface, or send "lock trucks" notifications to partner fleets, or maintain a resource status table locally (idle / locked / in use). The resource pool manager's output data includes resource lock confirmation receipts. The performance monitoring and activation module consists of monitoring probes (lightweight agents deployed at key nodes) and automatic switch controllers (dedicated comparison circuits). The input data for the performance monitoring and activation module includes actual order throughput, vehicle response latency, and warehouse operation rate. In other words, the performance monitoring and activation module is used to collect current logistics backlog data. For example, the performance monitoring and activation module can collect performance indicators every 30 seconds and determine whether dual threshold conditions are met: "time is within the high-risk period predicted in the simulation," "current logistics backlog is not lower than the simulation preset lower limit (corresponding to the lower limit of the backlog in the reference logistics backlog data)," and "growth rate is not lower than 80% of the simulation predicted value (corresponding to the growth rate in the reference logistics backlog data)." If these conditions are met, a "start contingency plan" signal is sent to the elastic scheduling engine (i.e., additional scheduling resources are scheduled according to the scheduling method). In addition, the performance monitoring and activation module can also monitor the physical indicators of the equipment layer (AGV battery voltage, motor temperature, positioning accuracy) and the physical indicators of the environment layer (warehouse temperature and humidity, local light intensity).

[0102] The aforementioned redundancy strategy library can be referenced from the following table:

[0103]

[0104] In other words, if the order backlog at the East China warehouse sorting point is greater than 500 orders based on the current logistics backlog data, then the scheduling strategy (i.e., scheduling method) with ID STRAT_WH_001 will be executed. Specifically, this scheduling strategy involves sending the instruction SET_ROUTE_OVERRIDE=RTE_BK_02 (indicating the addition of a new sorting line) to the AGV scheduling system via the industrial API gateway. If the current logistics backlog data satisfies multiple trigger conditions in the redundancy strategy library, then the scheduling strategy identifier, execution instruction, and physical carrier matching each satisfied trigger condition will be obtained. Subsequently, the elastic scheduling engine in the pressure scheduling integrated machine can send the corresponding execution instruction (i.e., the aforementioned standardized instruction) to the corresponding terminal through the corresponding physical carrier.

[0105] The redundancy strategy library is a fixed, executable emergency operation package, not an algorithmic model, and must demonstrate physical feasibility. That is, by querying the redundancy strategy library, one can obtain the scheduling methods for scheduling resources corresponding to the current logistics backlog data.

[0106] The standard resource list stored in the shared storage area (i.e., the list corresponding to the reference logistics backlog period and reference logistics backlog data mentioned above) can be found in the following table:

[0107]

[0108] In other words, the reference logistics backlog data includes the location of bottlenecks, the predicted backlog volume, and the backlog growth rate. The corresponding scheduling resources include the required resource types and quantities, as well as alternative routes. Then, by querying the redundancy policy library, corresponding scheduling instructions are sent to the terminals associated with the scheduling resources to generate the scheduling of these resources. For example, a scheduling instruction is sent to the sorting operator's mobile phone to notify them to report for duty. A scheduling instruction is also sent to the AGV scheduling system to execute alternative routes.

[0109] The output of the stress test simulator (corresponding to the simulation module) can be found in the following table:

[0110]

[0111] Among them, the three-dimensional array corresponds to the heat map of the logistics nodes mentioned above, the JSON array corresponds to the scheduling resources, and the string corresponds to the resource scheduling trigger event. A unique scene identifier is generated for the resource scheduling trigger event to facilitate subsequent auditing and traceability or model iteration.

[0112] The following example illustrates this concept: During the peak pre-sale period of the "Double Eleven" e-commerce promotion, e-commerce platforms anticipate order volume on the first day to reach 10 times the daily average. The system execution process can be referenced below:

[0113] 1. When time T = 2:00 AM the previous day:

[0114] The stress test simulator is started, injecting a combined scenario of "10 orders + 3 sections of main road under construction + 25% courier leave rate".

[0115] Output bottleneck heatmap: shows that the East China sorting center will experience backlogs from 9:00 to 11:30 in the morning (98% probability).

[0116] Output the emergency resource requirements list: "12 temporary sorters, 3 electric forklifts, and 5 backup charging stations are needed."

[0117] 2. When time T = 3:05 AM the previous day:

[0118] The elastic scheduling engine parses the heatmap and matches the strategy STRAT_WH_001 corresponding to the bottleneck period (9:00–11:30);

[0119] The corresponding execution command sent to the resource manager is: RESERVE_RIDERS=12, DURATION=4H, SET_ROUTE_OVERRIDE=RTE_BK_02.

[0120] 3. When time T = 8:45 AM on the same day:

[0121] The performance monitoring and activation module detected that the current time 8:45 does not belong to the reference logistics backlog period [9:00, 11:30], so monitoring will continue.

[0122] 4. When time T = 9:02 AM on the same day:

[0123] Real-time backlog = 520 orders (≥500 threshold), growth rate is +92 orders / hour (≥0.8×110).

[0124] The automatic switch controller outputs a high level, triggering the loading of the pre-set plan.

[0125] 5. When time T = 9:03 AM on the same day:

[0126] The AGV scheduling system received the instruction to divert 30% of the packages to the backup route RTE_BK_02.

[0127] The crowdsourced rider app received a standby notification, and 12 riders confirmed their response.

[0128] Final result: No serious backlog occurred in the East China warehouse, and the average delivery time for orders was only extended by 1.2 hours (compared to 4.5 hours in previous years).

[0129] In other words, in some embodiments, the system only outputs an activation signal to drive the elastic scheduling engine to load and execute the contingency plan (i.e., the corresponding scheduling resources and scheduling method) when the real-time monitored business indicators simultaneously meet three hard conditions: "the current time is within the high-risk period predicted by the simulation (i.e., the reference logistics backlog period)," "the current backlog amount is not lower than the preset lower limit of the simulation (i.e., the lower limit of a portion of the data in the reference logistics backlog period)," and "the current growth rate is not lower than a preset percentage (e.g., 80%) of the simulation prediction value." This logic is implemented by a dedicated comparison circuit hardware within the performance monitoring module, not a software rule engine, with a response latency of <100 microseconds. Refer to the following table for details:

[0130]

[0131] To achieve the above functions, the following components are deployed in the industrial control computer of the "Performance Monitoring and Activation Module": an ADC analog-to-digital converter (composed of a chip), a preset threshold storage area (composed of a ROM memory), a dedicated comparator circuit (composed of internal logic units), and an output driver (composed of a TTL level buffer chip). The functions of these components can be found in the following table:

[0132]

[0133] The above measures can improve system stability, advance resource preparation, increase response speed, save costs, enhance cross-system collaboration, and reduce the burden of operation and maintenance.

[0134] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0135] Based on the same inventive concept, this application also provides an intelligent logistics scheduling device for implementing the intelligent logistics scheduling method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more intelligent logistics scheduling device embodiments provided below can be found in the limitations of the intelligent logistics scheduling method described above, and will not be repeated here.

[0136] In one exemplary embodiment, such as Figure 4 As shown, an intelligent logistics scheduling device 400 is provided, including: a communication module 401, a simulation module 402, and a scheduling module 403, wherein:

[0137] The communication module 401 is used to obtain order data and transportation capacity data in response to resource scheduling trigger events.

[0138] The simulation module 402 is used to obtain a disturbance configuration template, adjust the order data and the transportation capacity data according to the disturbance configuration template to obtain disturbance combination data; based on the disturbance combination data, predict the logistics backlog period and the logistics backlog data corresponding to the logistics backlog period; and obtain the corresponding scheduling resources and scheduling methods based on the logistics backlog data.

[0139] The scheduling module 403 is located in the same physical host as the simulation module. The simulation module and the scheduling module interact with each other through a shared memory area in the physical host. If the difference between the current logistics backlog data and the logistics backlog data is within a preset range during the logistics backlog period, the scheduling resource is invoked and the scheduling resource is scheduled according to the scheduling method.

[0140] In some embodiments, regarding the prediction of a reference logistics backlog period and the reference logistics backlog data corresponding to the reference logistics backlog period based on the disturbance combination data, the simulation module 402 is specifically configured to: determine the logistics transportation period based on the resource scheduling trigger event; simulate the disturbance combination data to predict the simulation index data of each logistics node at each moment within the logistics transportation period, and determine the simulated logistics backlog information of each logistics node at each moment based on the corresponding simulation index data; the simulated logistics backlog information is used to characterize whether the logistics node is in a logistics backlog state at that moment; determine the reference logistics backlog period based on the simulated logistics backlog information of each logistics node at each moment; and determine the corresponding reference logistics backlog data based on the simulation index data corresponding to the reference logistics backlog period.

[0141] In some embodiments, when predicting a reference logistics backlog period and reference logistics backlog data corresponding to the reference logistics backlog period based on the disturbance combination data, the simulation module 402 is specifically used for: determining the logistics transportation period based on the resource scheduling trigger event; performing a preset number of simulations on the disturbance combination data, predicting the simulation index data of each logistics node at each moment within the logistics transportation period in each simulation, determining the simulation logistics backlog information of each logistics node at each moment based on the corresponding simulation index data, so as to obtain the simulation result of each simulation; obtaining a logistics node heatmap corresponding to the logistics transportation period based on each simulation result; the logistics node heatmap is used to characterize the correlation between the logistics backlog probabilities of each logistics node at each moment within the logistics transportation period; querying the target logistics backlog probability that meets the preset probability conditions in the logistics node heatmap, and determining the reference logistics backlog period related to the target logistics backlog probability; determining the corresponding reference logistics backlog data based on the statistical data corresponding to the simulation index data corresponding to the reference logistics backlog period.

[0142] In some embodiments, regarding obtaining the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data and storing them in the shared storage area, the simulation module 402 is specifically used to: obtain the target evacuation duration corresponding to the total backlog of parcels; obtain the current total resource amount of the logistics node corresponding to the reference logistics backlog period; calculate the total number of evacuable parcels within the target evacuation duration based on the current total resource amount; obtain the number of reference scheduling resources based on the difference between the total backlog of parcels and the total number of evacuable parcels, and the number of evacuable parcels corresponding to a single scheduling resource; and obtain the target strategy rule corresponding to the reference scheduling resource from the preset strategy rule base to obtain the reference scheduling method corresponding to the reference scheduling resource.

[0143] In some embodiments, regarding the adjustment of the order data and the capacity data according to the disturbance configuration template to obtain disturbance combination data, the simulation module 402 is specifically used to: obtain a first multiple interval corresponding to the order data and a second multiple interval corresponding to the capacity data from the disturbance configuration template; the first multiple in the first multiple interval is greater than one, and the second multiple in the second multiple interval is less than one; randomly select a first multiple from the first multiple interval, expand the order volume based on the first multiple, and obtain a corresponding target order volume; randomly select a second multiple from the second multiple interval, reduce the number of available transportation roads and the number of available transportation personnel based on the second multiple, and obtain corresponding target road data and target personnel data; obtain the disturbance combination data based on the target order volume, the target road data, and the target personnel data.

[0144] In some embodiments, in determining the target scheduling resource and target scheduling method corresponding to the current logistics backlog data based on the scheduling resource and the scheduling method, the scheduling module 403 is specifically configured to: calculate the ratio between the current logistics backlog amount in the current logistics backlog data and the reference logistics backlog amount in the reference logistics backlog data; and determine the target scheduling resource based on the ratio and the scheduling resource.

[0145] In some embodiments, in determining the target reference logistics backlog data based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period, the scheduling module 403 is specifically configured to: calculate the difference between the current logistics backlog data and each reference logistics backlog data corresponding to the reference logistics backlog period; select the smallest difference from each difference; and if the smallest difference meets a preset difference condition, determine the reference logistics backlog data corresponding to the smallest difference as the target reference logistics backlog data.

[0146] Each module in the aforementioned intelligent logistics scheduling device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0147] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data such as preset policy rule bases and disturbance configuration templates. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements an intelligent logistics scheduling method.

[0148] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements an intelligent logistics scheduling method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0149] Those skilled in the art will understand that Figure 5 or Figure 6The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0150] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0151] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0152] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0153] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0154] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0155] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0156] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A smart logistics scheduling method, characterized in that, The method includes: In response to a resource scheduling trigger event, order data and transportation capacity data are acquired; the order data includes the order quantity, and the transportation capacity data includes data related to available transportation roads and available transportation manpower. Obtain a disturbance configuration template; from the disturbance configuration template, obtain a first multiple interval corresponding to the order data and a second multiple interval corresponding to the transportation capacity data; the first multiple in the first multiple interval is greater than one, and the second multiple in the second multiple interval is less than one; randomly select a first multiple from the first multiple interval, and expand the order volume based on the first multiple to obtain the corresponding target order volume; randomly select a second multiple from the second multiple interval, and reduce the number of available transportation roads and the number of available transportation personnel based on the second multiple to obtain the corresponding target road data and target personnel data; obtain disturbance combination data based on the target order volume, the target road data, and the target personnel data; Based on the disturbance combination data, a reference logistics backlog period and corresponding reference logistics backlog data are predicted, including: determining a logistics transportation period based on the resource scheduling trigger event; simulating the disturbance combination data to predict the simulation index data of each logistics node at each moment within the logistics transportation period, and determining the simulation logistics backlog information of each logistics node at each moment based on the corresponding simulation index data; the simulation logistics backlog information is used to characterize whether the logistics node is in a logistics backlog state at that moment; determining the reference logistics backlog period based on the simulation logistics backlog information of each logistics node at each moment; determining the corresponding reference logistics backlog data based on the simulation index data corresponding to the reference logistics backlog period; or determining the logistics transportation period based on the resource scheduling trigger event. The disturbance combination data is simulated a preset number of times. In each simulation, the simulation index data of each logistics node at each moment within the logistics transportation period is predicted. Based on the corresponding simulation index data, the simulation logistics backlog information of each logistics node at each moment is determined to obtain the simulation result of each simulation. Based on each simulation result, a logistics node heatmap corresponding to the logistics transportation period is obtained. The logistics node heatmap is used to characterize the correlation between the logistics backlog probabilities of each logistics node at each moment within the logistics transportation period. The target logistics backlog probability that meets the preset probability conditions in the logistics node heatmap is queried, and the reference logistics backlog period related to the target logistics backlog probability is determined. The corresponding reference logistics backlog data is determined based on the statistical data corresponding to the simulation index data corresponding to the reference logistics backlog period. Obtain the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data and store them in the shared storage area; During the reference logistics backlog period, if a target reference logistics backlog data is determined based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period, then the reference scheduling resource and the reference scheduling method corresponding to the target reference logistics backlog data are obtained from the shared storage area. Based on the reference scheduling resource and the reference scheduling method, and the reference logistics backlog data, the target scheduling resource and the target scheduling method corresponding to the current logistics backlog data are determined, and the corresponding target scheduling method is sent to the target scheduling resource for resource scheduling.

2. The method according to claim 1, characterized in that, The order data includes at least one of the following: order volume, order distribution, estimated delivery time, cargo type, and customer location; the transportation capacity data includes at least one of the following: number of vehicles, vehicle location data, and employee shift schedule.

3. The method according to claim 1, characterized in that, The reference logistics backlog data includes the total number of backlogged parcels. Obtaining the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data and storing them in the shared storage area includes: Obtain the target evacuation time corresponding to the total amount of backlogged packages; Obtain the current total resources of the logistics nodes corresponding to the reference logistics backlog period; Calculate the total amount of current resources and the total number of evacuable packages within the target evacuation time. The number of reference scheduling resources is obtained based on the difference between the total backlog of packages and the total number of packages that can be evacuated, and the number of packages that can be evacuated corresponding to a single scheduling resource. Obtain the target policy rule corresponding to the reference scheduling resource from the preset policy rule base to obtain the reference scheduling method corresponding to the reference scheduling resource.

4. The method of claim 1, wherein, The step of determining the target scheduling resource and target scheduling method corresponding to the current logistics backlog data based on the reference scheduling resource, the reference scheduling method, and the reference logistics backlog data includes: Calculate the ratio between the current logistics backlog amount in the current logistics backlog data and the reference logistics backlog amount in the reference logistics backlog data; The target scheduling resource is determined based on the ratio and the scheduling resource.

5. The method of claim 1, wherein, The determination of the target reference logistics backlog data based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period includes: Calculate the difference between the current logistics backlog data and the reference logistics backlog data corresponding to each reference logistics backlog period; The smallest gap is selected from all the gaps. If the smallest gap meets the preset gap conditions, the reference logistics backlog data corresponding to the smallest gap is determined as the target reference logistics backlog data.

6. An intelligent logistics scheduling apparatus, characterized by comprising: The device includes: The communication module is used to acquire order data and transportation capacity data in response to resource scheduling trigger events; the order data includes the order quantity, and the transportation capacity data includes data related to available transportation roads and available transportation manpower. The simulation module is used to acquire a disturbance configuration template, obtain a first multiple interval corresponding to the order data and a second multiple interval corresponding to the transportation capacity data from the disturbance configuration template; the first multiple in the first multiple interval is greater than one, and the second multiple in the second multiple interval is less than one; a first multiple is randomly selected from the first multiple interval, and the order quantity is expanded based on the first multiple to obtain the corresponding target order quantity; a second multiple is randomly selected from the second multiple interval, and the number of available transportation roads and the number of available transportation personnel are reduced based on the second multiple to obtain the corresponding target road data and target capacity data. The system calculates and distributes the target manpower data; it obtains disturbance combination data based on the target order volume, target road data, and target manpower data; based on the disturbance combination data, it predicts a reference logistics backlog period and reference logistics backlog data corresponding to the reference logistics backlog period, including: determining the logistics transportation period based on the resource scheduling trigger event; simulating the disturbance combination data to predict the simulation index data of each logistics node at each moment within the logistics transportation period, and determining the simulated logistics backlog information of each logistics node at each moment based on the corresponding simulation index data; the simulated logistics backlog information is used to characterize the logistics node at each moment. Whether it is in a logistics backlog state; determine the reference logistics backlog period based on the simulated logistics backlog information of each logistics node at each time; determine the corresponding reference logistics backlog data based on the simulation index data corresponding to the reference logistics backlog period; or determine the logistics transportation period based on the resource scheduling trigger event; perform a preset number of simulations on the disturbance combination data, and in each simulation, predict the simulation index data of each logistics node at each time within the logistics transportation period, and determine the simulated logistics backlog information of each logistics node at each time based on the corresponding simulation index data, so as to obtain the simulation results of each simulation; Based on each simulation result, a heatmap of logistics nodes corresponding to the logistics transportation period is obtained; the heatmap of logistics nodes is used to characterize the correlation between the probabilities of logistics backlog at each moment within the logistics transportation period; the target logistics backlog probability that meets the preset probability conditions is queried in the heatmap of logistics nodes, and the reference logistics backlog period related to the target logistics backlog probability is determined; the corresponding reference logistics backlog data is determined based on the statistical data corresponding to the simulation index data corresponding to the reference logistics backlog period; the reference scheduling resources and reference scheduling methods corresponding to the reference logistics backlog data are obtained and stored in the shared storage area; The scheduling module, located within the same physical host as the simulation module, interacts with the simulation module via a shared memory area within the physical host. During the reference logistics backlog period, if a target reference logistics backlog is determined based on the difference between the current logistics backlog data and the reference logistics backlog data corresponding to the reference logistics backlog period, the module retrieves the reference scheduling resource and reference scheduling method corresponding to the target reference logistics backlog data from the shared memory area. Based on the reference scheduling resource, the reference scheduling method, and the reference logistics backlog data, the module determines the target scheduling resource and target scheduling method corresponding to the current logistics backlog data and sends the corresponding target scheduling method to the target scheduling resource for resource scheduling.

7. The apparatus of claim 6, wherein, The simulation module and the scheduling module are connected via a high-speed bus.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images