Intelligent Material Allocation Management System Based on Industrial Internet of Things

Abstract

This invention discloses an intelligent material allocation and management system based on the Industrial Internet of Things (IIoT), relating to the field of IIoT technology. It includes a central collaborative controller, an edge computing gateway cluster, an IIoT sensing node array, heterogeneous material handling equipment groups, and an infrastructure linkage unit. The central collaborative controller converts production plans into standardized allocation instructions for issuance; the edge computing gateway cluster, combined with real-time status feedback from sensing nodes, performs dynamic task decomposition, path planning, and equipment assignment; the infrastructure linkage unit is responsible for the automatic control of facilities such as elevators and access control systems, cooperating with the heterogeneous material handling equipment groups to complete cross-regional operations. This invention, through a three-layer cloud-edge-device architecture, solves the problems of system fragmentation, feedback lag, and lack of cross-regional logistics coordination in existing technologies, achieving millisecond-level synchronization of production and logistics, end-to-end closed-loop feedback, and highly flexible intelligent allocation.

Description

Technical Field

[0001] This invention belongs to the field of industrial Internet of Things (IoT) technology, and more specifically, relates to an intelligent material allocation and management system based on industrial IoT. Background Technology

[0002] Against the backdrop of the accelerated evolution of modern manufacturing towards intelligence and flexibility, material allocation, as a crucial link connecting production planning and logistics execution, directly determines the responsiveness and resource utilization level of the entire manufacturing system through its operational efficiency and accuracy. Especially in discrete manufacturing, electronic assembly, and high-end equipment manufacturing scenarios involving diverse product types and small batches, materials must be delivered to the correct workstations at the right time and in the right quantities, posing unprecedented challenges to traditional logistics management models. The rise of Industrial Internet of Things (IIoT) technology offers a new possibility for reconstructing the logic of material allocation. Its core value lies in achieving a fundamental shift from "experience-driven" to "data-driven" through equipment interconnection, data integration, and intelligent decision-making.

[0003] For a long time, manufacturing enterprises have generally adopted a material allocation model that relies primarily on manual scheduling, supplemented by paper or semi-electronic documents. This model typically depends on independently operating warehouse management systems and manufacturing execution systems. While it effectively supported the stable operation of large-scale production in a specific historical period, its architecture is inherently fragmented: production plans are generated by the enterprise resource planning system and then manually broken down and transmitted to the warehousing and logistics links, lacking real-time feedback and dynamic calibration mechanisms. Furthermore, key operations such as material inbound / outbound, route selection, and vehicle scheduling heavily depend on the experience and judgment of on-site personnel, leading to frequent problems such as discrepancies between inventory records and actual stock, delivery delays, and incorrect or missing shipments. More importantly, due to inconsistent data standards and closed interfaces between systems, typical information silos have been formed, resulting in significant time lags and semantic discrepancies between production instructions and logistics execution, making it difficult to support the high-frequency, fast-paced flexible production demands.

[0004] However, with the deepening implementation of Industry 4.0 and the continuous improvement of customer requirements for delivery cycles and customization levels, the inherent limitations of the aforementioned technical architecture at the principle level are becoming increasingly apparent. Specifically, while traditional allocation models can solve material flow problems in static, deterministic environments, they cannot cope with the real-time coordination needs under dynamic disturbances. The reason for this is that its underlying logic is based on a linear closed loop of "planning-execution-feedback," and the feedback cycle of this closed loop is too long and the granularity is too coarse. As a result, the system lacks the ability to quickly replan and adaptively adjust when faced with common operating conditions such as abnormal production line shutdowns, emergency order insertions, and fluctuations in material availability. Furthermore, even if some companies introduce automated handling equipment such as automated guided vehicles (AGVs), without a deep linkage mechanism with infrastructure such as elevators, hoists, and access control systems, logistics bottlenecks across floors and regions are still difficult to overcome. Automated equipment often becomes "partial automation," and the overall logistics network still exhibits fragmented characteristics. Furthermore, the architecture of uploading all sensing data to the cloud for centralized processing is susceptible to network jitter and bandwidth limitations in high-concurrency, low-latency application scenarios. This not only increases system response latency but may also lead to global scheduling failure due to single-point failures, thereby weakening the robustness and reliability of the system.

[0005] The core of the aforementioned contradiction lies in the fact that while pursuing partial automation, existing technological systems have failed to construct a real-time collaborative architecture that integrates perception, decision-making, and execution. This results in a structural disconnect between the flexible requirements of the "planning layer" and the rigid constraints of the "execution layer." This disconnect is not only reflected in the lag in response time but also in the lack of coordination in space—that is, the inability to perform global optimization and dynamic coordination for multiple types of handling equipment and multiple constraints (such as path conflicts, energy consumption limitations, and equipment status) in complex physical environments. Therefore, how to construct a material intelligent allocation mechanism that deeply integrates industrial IoT technology, possesses edge intelligence and cloud-edge collaboration capabilities, and can achieve seamless data fusion and closed-loop feedback across multiple systems, while balancing scheduling efficiency, energy consumption, system reliability, and management flexibility, has become a key challenge and an urgent technical problem for those skilled in the art. Summary of the Invention

[0006] This invention provides a material intelligent allocation and management system based on the Industrial Internet of Things, aiming to solve the technical problems of low scheduling efficiency, response delay, difficulty in overcoming cross-regional logistics bottlenecks, and insufficient overall system robustness caused by system fragmentation, delayed feedback, and lack of coordination in the prior art.

[0007] To address the aforementioned problems, the technical solution adopted by this invention is as follows: a material intelligent allocation management system based on the Industrial Internet of Things (IIoT), characterized in that it includes: a central collaborative controller, an edge computing gateway cluster, an IIoT sensing node array, a heterogeneous handling equipment group, and an infrastructure linkage unit; the central collaborative controller is deployed in the cloud and is used to receive the master production plan data stream from the enterprise resource planning system, convert it into a standardized allocation instruction set, and distribute it through a secure encrypted channel; the edge computing gateway cluster is deployed according to physical regions, with each edge computing gateway responsible for an independent logistics operation subdomain, used to receive the allocation instruction set and combine it with local real-time status information for dynamic... The system performs task decomposition, path planning, and task allocation, generating control commands which are then sent to the heterogeneous material handling equipment group. The industrial IoT sensing node array is distributed across the logistics operation site to collect material location and environmental status information in real time, updating it to the edge computing gateway cluster at a frequency no less than a preset frequency. The heterogeneous material handling equipment group receives and executes control commands issued by the edge computing gateway cluster and periodically reports its own status through a unified communication interface. The infrastructure linkage unit connects to the building automation system in a non-intrusive manner, automatically controlling the operation of elevators, hoists, and access control facilities according to requests from the edge computing gateway cluster during the execution of tasks by the heterogeneous material handling equipment group.

[0008] Furthermore, the central collaborative controller is equipped with an instruction conversion engine. The instruction conversion engine converts the master production plan data stream, which includes material codes, target workstation identifiers, expected arrival time windows, and priority labels, into a structured allocation instruction set based on a preset mapping rule library. Each allocation instruction includes a unique task ID, source location coordinates, target location coordinates, material type identifier, load weight, time window constraints, priority level, and task timeout threshold.

[0009] Furthermore, each edge computing gateway in the edge computing gateway cluster integrates a local state cache module, a real-time scheduling engine, a conflict resolution module, and a device driver adapter. The real-time scheduling engine adopts a hierarchical decision-making mechanism, including a task decomposition module, a path planning module, and a task allocation module. The task decomposition module breaks down the allocation instruction into several atomic operation units. The path planning module generates a trajectory for each atomic operation unit that meets energy consumption constraints and obstacle avoidance requirements. The task allocation module uses a greedy strategy based on capability matching degree to assign equipment to the heterogeneous handling equipment group.

[0010] Furthermore, the path planning module uses an improved A algorithm combined with a dynamic window method for trajectory generation. The cost function of the improved A algorithm is:

[0011]

[0012] in, This represents the actual cost from the starting point to the current node. This represents the heuristic cost from the current node to the destination. This represents the estimated energy consumption from the current node to the target node. This indicates the probability of the path segment being occupied in the next thirty seconds. Energy consumption weighting factor The congestion prediction factor is used; the occupancy probability is calculated in real time from historical traffic data using a sliding window statistical model.

[0013] Furthermore, when assigning devices, the task allocation module selects the execution device based on the capability matching degree M, and the formula for calculating the capability matching degree M is:

[0014]

[0015] in, This is the Euclidean distance between the current position of the device and the starting point of the mission. The preset distance smoothing factor, This represents the device's remaining battery life. The estimated power consumption for the mission, This is the compatibility coefficient between the device's current load status and the task load type; and Both are maximum-minimum normalization functions, used to map the distance factor and energy consumption factor to the [0,1] interval, respectively; For the preset weighting coefficients, satisfy .

[0016] Furthermore, the conflict resolution module adopts a dynamic reservation mechanism based on a spatiotemporal resource graph, mapping the trajectory output by the path planning module to spatiotemporal occupancy blocks. When an impending path conflict is detected, the priorities of the conflicting tasks are compared: the transport equipment corresponding to the high-priority task maintains its original trajectory; the transport equipment corresponding to the low-priority task triggers a local adjustment strategy, calculating the cost function of waiting in place or detouring at the current node. If the waiting time is less than a preset threshold, a waiting instruction is inserted and the trajectory timestamp is updated; otherwise, the real-time scheduling engine is triggered to replan.

[0017] Furthermore, the infrastructure linkage unit includes an elevator control interface module, a hoist coordination module, and an access control linkage module. The elevator control interface module is used to listen for floor request signals, read the elevator status via the Modbus TCP protocol, and send a call command and lock the right to use the elevator after confirming that it is idle or about to arrive. The hoist coordination module is used to receive hoisting requests and send a start signal to the hoist controller after verifying that the weight of the material is within the rated load range and that the pallet has completely entered the platform area. The access control linkage module is used to verify the validity of the task ID and access permission when the handling equipment approaches the controlled channel, and send an unlock command after successful verification.

[0018] Furthermore, the industrial IoT sensing node array consists of a fixed RFID reader, a lidar, an infrared sensor, and a pressure sensing pad. The fixed RFID reader uses a backscatter communication mechanism to read the electronic tags on the materials. The lidar generates a two-dimensional point cloud map of the local environment at a frequency of ten frames per second. The pressure sensing pad is installed on the surface of the loading and unloading platform to monitor the material pressure value. If the pressure value does not reach the threshold within a preset time, it is determined as a loading failure and a retry process is triggered.

[0019] Furthermore, the system also includes a digital twin simulation platform running on the server where the central collaborative controller is located. The digital twin simulation platform includes an entity model and a logical model. The digital twin simulation platform periodically receives actual operating data from the edge computing gateway cluster, including task execution time, path deviation, and device energy consumption, and compares the deviation with the simulation results. When the task completion time deviation exceeds a preset threshold or the path deviation exceeds the allowable range, the strategy adjustment mechanism is automatically triggered to retrain the heuristic rule base in the scheduling engine.

[0020] Furthermore, after detecting the device failure event flag, the edge computing gateway executes a relay takeover mechanism: removing the faulty device from the available resource pool, searching for the nearest available device near the fault point and assigning it as the relay executor, updating the starting coordinates of the relay task to the real-time location at the time of the fault, and having the relay executor continue to complete the remaining journey.

[0021] Compared with the prior art, the beneficial effects of the present invention are:

[0022] (1) This invention constructs a three-layer architecture of “cloud collaboration - edge decision-making - device execution”. The central collaborative controller solves the problem of information silos and transforms the upper-layer plan into standardized instructions; the edge computing gateway cluster realizes the localization of task decomposition and path planning, which greatly reduces network latency and achieves millisecond-level scheduling response; the infrastructure linkage unit realizes seamless interaction between the handling equipment and elevators and access control through non-intrusive access, effectively breaking through the logistics bottlenecks across regions and floors and ensuring the automated execution of the entire link.

[0023] (2) This invention introduces an improved A algorithm that includes energy consumption weights and congestion prediction factors in path planning, and adopts a capability matching model that comprehensively considers distance, range, and load compatibility when allocating tasks. This not only avoids local congestion caused by simply pursuing the shortest path, but also achieves a globally optimal balance between equipment energy consumption and operating efficiency, thereby improving the intelligence level of the system.

[0024] (3) Through the dynamic reservation mechanism and priority preemption strategy based on the spatiotemporal resource map, the system can resolve path conflicts and deadlock risks between multiple devices in real time. In particular, the introduction of the "relay takeover" mechanism enables the rapid assignment of nearby devices to take over the task when a device fails midway, ensuring the continuity of material flow and significantly enhancing the system's fault tolerance and robustness.

[0025] (4) By combining with a digital twin simulation platform, this invention can compare the actual operating data of the physical world with the simulation model in real time. Through the deviation-triggered strategy adjustment mechanism, the system can automatically optimize the heuristic rule base, realizing the transformation from "preset rule execution" to "data-driven self-evolution", and continuously improving the accuracy of the scheduling strategy.

[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, embodiments of the present invention are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of the overall architecture of the intelligent material allocation management system based on the Industrial Internet of Things of this invention.

[0029] Figure 2 This is a schematic diagram of the hierarchical decision-making mechanism of the real-time scheduling engine in this invention.

[0030] Figure 3 This is a schematic diagram illustrating the execution flow of the intelligent material allocation and management method of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0032] This invention provides a material intelligent allocation and management system based on the Industrial Internet of Things. Its core lies in constructing a three-layer architecture consisting of a perception layer, an edge decision layer, and a cloud collaboration layer. Through system-level integration and data closed-loop mechanism, it achieves millisecond-level response and dynamic optimization from Enterprise Resource Planning (ERP) instructions to the execution of material handling in the physical world.

[0033] In a preferred embodiment of the present invention, the above three-layer architecture can be summarized into three collaboratively operating platforms in terms of logic and function: management platform, sensor network platform, and perception control platform.

[0034] Specifically, the management platform serves as the system's brain, comprising a central collaborative controller and an edge computing gateway cluster. The management platform is responsible for end-to-end data processing and in-depth analysis, undertaking core computing tasks such as master production schedule parsing, global scheduling, local dynamic task allocation, and path planning.

[0035] The perception and control platform serves as the system's backbone and senses, comprising an industrial IoT sensing node array, heterogeneous material handling equipment groups, and infrastructure linkage units. The platform is primarily responsible for the real-time acquisition of raw data such as the underlying physical environment, equipment status, and material location at the logistics operation site, and ultimately for executing the physical handling and linkage control tasks issued by the management platform.

[0036] The sensor network platform is the neural network connecting the system. As the data transmission hub between the management platform and the perception and control platform, it is responsible for transmitting the massive amount of low-level state data collected by the perception and control platform to the management platform for unified analysis and processing at high speed and with low latency. It also accurately sends the control commands and scheduling schemes generated by the management platform to each execution terminal of the perception and control platform.

[0037] The following will provide a detailed description of the specific implementation of this invention, including its system composition, functional modules, communication protocols, scheduling logic, and exception handling mechanisms.

[0038] like Figure 1As shown, the central collaborative controller is deployed in an enterprise private cloud or hybrid cloud environment, serving as the top-level decision-making hub of the management platform. Its main functions include master production schedule parsing, allocation instruction generation, global status monitoring, and dynamic strategy adjustment. The central collaborative controller receives master production schedule data streams from the enterprise resource planning system through a standard interface. This data stream contains fields such as material codes, target workstation identifiers, expected arrival time windows, priority tags, and associated work order numbers in a structured format. The controller has an internal instruction conversion engine that converts high-level production instructions into standardized allocation instruction sets based on a preset mapping rule base. Each allocation instruction is a structured message body containing a unique task ID, source location coordinates, target location coordinates, material type identifier, load weight, time window constraints, priority level, and task timeout threshold. The generated allocation instruction sets are distributed to each edge computing gateway via a secure encrypted channel based on TLS 1.3, ensuring data integrity, confidentiality, and replay attack prevention during transmission.

[0039] Edge computing gateway clusters are deployed according to physical regions, with each gateway responsible for an independent logistics operation subdomain, such as the raw material warehouse area, assembly line A section, and finished product temporary storage area.

[0040] Each edge computing gateway, acting as an edge processing node of the management platform, integrates a local state cache module, a real-time scheduling engine, a conflict resolution module, and a device driver adapter. This allows for localized analysis and processing of data within its jurisdiction. The local state cache module stores the real-time status of all key entities within the current subdomain in key-value pairs, including the current location of materials, the power level of handling equipment, path occupancy status, elevator car location, access control permission status, and infrastructure availability indicators. This module continuously receives status updates from the industrial IoT sensing node array via a subscription mechanism, updating at least twice per second to ensure the timeliness of the local state view.

[0041] like Figure 2 As shown, the real-time scheduling engine adopts a hierarchical decision-making mechanism divided into three levels: task decomposition, path planning, and task allocation. The task decomposition module receives high-order allocation instructions from the central collaborative controller and breaks them down into several atomic operation units. Each atomic operation unit defines precise starting point geographical coordinates, ending point geographical coordinates, load type, maximum allowable time, and completion flags for dependent preceding operations. The path planning module, based on a high-precision topology map of the plant area and dynamic obstacle information, generates an optimal trajectory for each atomic operation unit that meets energy consumption constraints and obstacle avoidance requirements. This module uses an improved A algorithm combined with a dynamic window method for trajectory generation. The improved A algorithm improves upon the traditional cost function... Introducing an energy consumption weighting factor based on Congestion predictors This forms a new cost function:

[0042]

[0043] in This represents the estimated energy consumption from the current node to the target node. This indicates the probability of occupancy of the path segment within the next 30 seconds, calculated in real-time from historical traffic data using a sliding window statistical model. The dynamic window rule is used to correct the trajectory online, ensuring that the generated trajectory meets the dynamic constraints of the equipment, such as maximum speed, acceleration, and turning radius.

[0044] The initial trajectory set generated by the path planning module will be output as verification data to the conflict resolution module in the gateway. Only after the trajectory passes the spatiotemporal compliance verification can it enter the next level of task allocation process.

[0045] The task allocation module employs a greedy strategy based on capability matching degree for equipment assignment. To eliminate dimensional differences and prevent computational overflow, the capability matching degree M is calculated using a weighted average of three normalized indicators:

[0046]

[0047] in, ε is the Euclidean distance between the current position of the device and the starting point of the task, and ε is a preset distance smoothing factor (ranging from 0.01 to 0.1 meters) used to prevent calculation anomalies caused by the denominator being zero when devices overlap. This represents the device's remaining battery life. Estimated power consumption for the mission; This is the compatibility coefficient between the device's current load status and the task load type (1 for compatible, 0 for incompatible). and Both are maximum-minimum normalization functions, used to map the distance factor and energy consumption factor to the [0,1] interval, respectively. For the preset weighting coefficients, satisfy Each time a task is assigned, the scheduling engine iterates through all available devices, calculates their capability matching score, and selects the one with the highest score to execute the task. If multiple devices have the same matching score, they are selected in ascending order of device ID.

[0048] The conflict resolution module, as a functional unit independent of the scheduling engine within the edge computing gateway, is responsible for verifying the physical feasibility of the planned trajectory. This module employs a dynamic reservation mechanism based on a spatiotemporal resource graph, mapping the trajectory output by the path planning module into a series of spatiotemporal occupancy blocks. The system checks in real-time whether there are overlapping occupancy blocks in the global spatiotemporal resource graph: if an impending path conflict is detected (i.e., different devices occupy the same road network node or road segment within the same time window), the module first compares the priorities of the conflicting tasks. The transport equipment corresponding to the high-priority task maintains its original trajectory; the transport equipment corresponding to the low-priority task triggers a local adjustment strategy, and the system calculates the cost function for it to "wait in place" or "detour" at the current node. If the waiting time is less than a preset threshold (e.g., 5 seconds), a waiting instruction is inserted into the device and its trajectory timestamp is updated; otherwise, a signal is sent to the scheduling engine to trigger replanning. Through this mechanism, the system completes deadlock prevention and traffic flow management within milliseconds, ensuring the continuity of multi-machine collaboration.

[0049] The heterogeneous material handling equipment group includes Automated Guided Vehicles (AGVs), Autonomous Mobile Robots (AMRs), rail-mounted conveyors, and forklift-type handling devices. Each type of equipment is equipped with a unified communication interface and a status reporting protocol. The unified communication interface uses the Lightweight Message Queuing Telemetry Transport Protocol (MQTT), running on top of the TCP / IP protocol stack, and supports bidirectional heartbeat detection and disconnection reconnection mechanisms. The heartbeat packet sending period is 300 milliseconds. If no heartbeat response is received three times consecutively, the equipment is considered offline and a fault handling process is triggered. The status reporting protocol stipulates that the equipment must periodically upload its own position coordinates (accuracy better than ±5 cm), direction of movement (expressed in radians), remaining range (unit: km), current task ID, and abnormal event flags (such as low battery, collision alarm, load drop, etc.), with a reporting period of no more than 500 milliseconds.

[0050] The infrastructure linkage unit consists of an elevator control interface module, a hoist coordination module, and an access control linkage module. Each module connects to the existing building automation system in a non-intrusive manner, without requiring modification to the original control logic. The elevator control interface module listens for floor request signals from the edge computing gateway, which include the target floor, estimated dwell time, and task priority. The module communicates with the elevator controller via the Modbus TCP protocol, reading the status register to obtain the elevator's current direction of travel, current floor, and door open / close status. When it confirms that the target elevator is idle or about to arrive at the target floor, it sends a call command to the elevator controller and locks the elevator's usage rights for a specified time period in the elevator control system to prevent interference from other unauthorized requests. When multiple handling devices simultaneously request the same elevator, the elevator control interface module prioritizes them based on task priority and estimated waiting time, sending call commands sequentially to ensure that higher-priority tasks are responded to first.

[0051] The aforementioned communication protocols (including MQTT, Modbus TCP, and secure encrypted channels based on TLS 1.3) collectively form the foundation of the sensor network platform of this invention. In actual operation, data flow follows a closed-loop logic of acquisition-transmission-analysis and processing-execution: the sensing and control platform (sensing node array, handling equipment, etc.) does not perform complex logical operations on-site, but instead uploads basic data such as the acquired equipment location, power consumption, environmental point cloud, and access control elevator status to the management platform (edge ​​computing gateway and central controller) in real time through the sensor network platform; after receiving this data, the management platform performs complex data analysis and processing such as path planning, collision detection, and energy consumption prediction, and then sends the generated action commands back through the sensor network platform to drive the sensing and control platform to complete the physical process of material allocation.

[0052] The aforementioned hoist coordination module ensures that materials are not overloaded or unevenly loaded during vertical transfer through preset start / stop logic and load verification mechanisms. This module receives hoisting requests from the edge computing gateway, verifies whether the material weight is within the hoist's rated load range, and checks whether the pallet has fully entered the hoisting platform area. Only after both verifications pass does it send a start signal to the hoist controller. The access control module automatically verifies the legitimacy of the task when the transport equipment approaches the controlled channel. Verification logic includes: whether the current task ID is valid, whether the target area matches, and whether the equipment has access permissions. After successful verification, the module sends an unlock command to the access control controller and automatically closes the access control after the equipment has completely passed through; the entire process requires no manual intervention.

[0053] The Industrial Internet of Things (IIoT) sensing node array consists of fixed RFID readers, LiDAR, infrared sensors, and pressure sensing pads, distributed across warehouse shelves, production line workstations, transfer nodes, and critical path intersections. The fixed RFID readers operate in the 860 MHz to 960 MHz frequency band, employing a backscatter communication mechanism with a reading distance of at least five meters. The readers periodically poll the passive UHF electronic tags within their coverage area to obtain the unique identification of each material. Each electronic tag stores four core fields: material code, batch number, expiration date, and project number. All fields are written once before shipment and cannot be changed. The LiDAR scans the local environment at a frequency of ten frames per second, generating a two-dimensional point cloud map to assist handling equipment in precise positioning and dynamic obstacle avoidance. Infrared sensors detect personnel entering dangerous areas; upon detection of unauthorized entry, a safety alarm signal is immediately sent to the edge computing gateway. Pressure sensing pads are installed on the loading / unloading platform surface to monitor whether materials are fully loaded or unloaded, preventing them from falling during transport. If the pressure value does not reach the threshold within the preset time, it is determined as a loading failure, and a retry or manual intervention process is triggered.

[0054] The digital twin simulation platform runs on the server hosting the central collaborative controller. Its physical model and logical model respectively map the physical world's equipment layout and business rules. The physical model is constructed using 3D modeling tools, including the shape parameters of all handling equipment, path width, turning radius, and obstacle contours, with a spatial accuracy better than ±10 cm. The logical model describes various business processes in the form of finite state machines, including normal allocation processes, emergency order processing processes, equipment failure emergency processes, and energy replenishment scheduling processes. Each state transition condition is defined by a Boolean expression, and the action is implemented by a script function. The digital twin simulation platform periodically receives actual operating data from the edge computing gateway, including task execution time, path deviation, equipment energy consumption, and abnormal event records, and compares the deviation with the simulation results. When the task completion time deviation exceeds a preset threshold or the path deviation exceeds the allowable range (e.g., one meter), the strategy adjustment mechanism is automatically triggered to retrain the heuristic rule base in the scheduling engine.

[0055] The central coordinating controller and edge computing gateways interact via a publish-subscribe model. The central coordinating controller, acting as the publisher, broadcasts allocation instructions as topics, named according to the convention "region / {region ID} / task", such as "region / warehouse_A / task". Each edge computing gateway, acting as a subscriber, only subscribes to topics relevant to its assigned region, significantly reducing unnecessary data traffic. After completing local scheduling, the edge computing gateway sends an acknowledgment message back to the central coordinating controller, containing the task ID, actual start time, actual completion time, executing device ID, and exception code. This mechanism forms a complete closed-loop feedback chain, supporting performance evaluation and rolling plan adjustments by the upper-layer system.

[0056] The central collaborative controller also integrates a low-code business process configuration interface, allowing administrators to define new allocation rules or modify existing rules through a graphical drag-and-drop interface. Each component in the configuration interface corresponds to a predefined functional block, including conditional judgment blocks (such as "if task priority is greater than 5"), action execution blocks (such as "allocate AMR type equipment"), waiting delay blocks (such as "wait 30 seconds"), and exception jump blocks (such as "if battery level is below 20%, jump to charging station"). All configuration operations are ultimately compiled into Structured Query Language (SQL) scripts, stored in the rule table of the central database, and dynamically loaded and executed by the central collaborative controller at runtime, enabling flexible iteration of business logic.

[0057] The edge computing gateway employs a dual-buffer mechanism to handle sudden surges in concurrent requests. The first buffer, with a capacity of one hundred newly arrived allocation instructions, temporarily stores copies of instructions currently being processed; the second buffer, with a capacity of fifty, stores copies of instructions being processed. When the first buffer is full, a priority preemption mechanism is activated: if a newly arrived instruction has a higher priority than the lowest priority instruction in the buffer, the storage space of the lower-priority instruction is overwritten, and it is marked as "cancelled." This mechanism ensures that critical tasks are not blocked, maintaining the system's responsiveness under high load.

[0058] like Figure 3 As shown, the specific execution process of the intelligent material allocation management system of the present invention is as follows: First, the central collaborative controller receives the master production plan issued by the enterprise resource planning system, extracts the material demand list, and generates an initial allocation instruction set; second, the allocation instruction set is distributed to the corresponding edge computing gateways according to the target area; third, each edge computing gateway, combined with local sensing data, generates a specific task sequence through a real-time scheduling engine and issues it to the heterogeneous handling equipment group; then, during the execution of the task, the handling equipment continuously reports its location and status information through the industrial Internet of Things sensing node array; simultaneously, the infrastructure linkage unit automatically activates relevant facilities according to the task progress; finally, the edge computing gateway sends the task execution result back to the central collaborative controller, completing a complete allocation closed loop.

[0059] In emergency task insertion scenarios, the central coordination controller immediately generates a high-priority allocation command and forcibly interrupts ongoing low-priority tasks within the affected area, requeuing them. The remaining tasks of the interrupted tasks are re-routed by the scheduling engine to ensure that the overall system throughput is not significantly affected. During the re-scheduling process, the scheduling engine prioritizes nearby available devices and uses a digital twin platform for rapid simulation to verify the feasibility of the new solution.

[0060] In the event of a device failure, once the edge computing gateway detects the abnormal event flag, it immediately removes the faulty device from the available resource pool and reassigns its currently assigned tasks. If the failure occurs midway through a task, the scheduling engine initiates a "relay takeover" mechanism: it searches for the nearest available device near the failure point and assigns it as the relay executor to continue the remaining journey. The starting coordinates of the relay task are updated to the real-time location at the time of the failure, ensuring that materials do not remain in uncontrolled areas.

[0061] Furthermore, the digital twin simulation platform underwent two weeks of offline simulation verification before the system went live. The input consisted of three months of historical production plans and logistics data, and the simulation model reproduced all equipment behavior and infrastructure interaction logic. Simulation results showed that the new scheduling strategy could reduce the average task completion time by 23% and the equipment idle rate by 17%. The online simulation mode was executed every ten minutes during daily operation, successfully predicting and avoiding seven potential path congestion events, validating its forward-looking control capabilities.

[0062] In summary, this invention, through the deep integration of industrial IoT sensing technology, edge intelligent decision-making, and cloud-based collaborative control, constructs a highly integrated, adaptive, and scalable intelligent material allocation and management system. This system not only achieves real-time mapping and closed-loop optimization between the physical and digital worlds, but also supports rapid deployment and continuous evolution in different manufacturing scenarios through standardized interfaces and modular design, providing a solid and reliable intelligent logistics foundation for modern flexible manufacturing.

[0063] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A material intelligent allocation and management system based on the Industrial Internet of Things, characterized in that, include: Central collaborative controller, edge computing gateway cluster, industrial IoT sensing node array, heterogeneous handling equipment group, and infrastructure linkage unit; The central collaborative controller is deployed in the cloud and is used to receive the master production plan data stream from the enterprise resource planning system, convert it into a standardized allocation instruction set, and distribute it through a secure and encrypted channel. The edge computing gateway cluster is deployed according to physical regions. Each edge computing gateway is responsible for an independent logistics operation subdomain, which is used to receive the allocation instruction set, combine local real-time status information to perform dynamic task decomposition, path planning and task allocation, and generate control instructions to send to the heterogeneous handling equipment group. The industrial IoT sensing node array is distributed at the logistics operation site to collect material location and environmental status information in real time, and update it to the edge computing gateway cluster at a frequency of not less than a preset frequency. The heterogeneous handling equipment group is used to receive and execute control commands issued by the edge computing gateway cluster, and periodically report its own status through a unified communication interface; The infrastructure linkage unit is connected to the building automation system in a non-intrusive manner and is used to automatically control the operation of elevators, hoists and access control facilities according to the requests of the edge computing gateway cluster during the execution of tasks by the heterogeneous handling equipment group.

2. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 1, characterized in that, The central collaborative controller is equipped with an instruction conversion engine. The instruction conversion engine converts the master production plan data stream, which includes material codes, target workstation identifiers, expected arrival time windows and priority labels, into a structured allocation instruction set based on a preset mapping rule library. Each allocation instruction includes a unique task ID, source location coordinates, target location coordinates, material type identifier, load weight, time window constraints, priority level and task timeout threshold.

3. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 1, characterized in that, Each edge computing gateway in the edge computing gateway cluster integrates a local state cache module, a real-time scheduling engine, a conflict resolution module, and a device driver adapter. The real-time scheduling engine adopts a hierarchical decision-making mechanism, including a task decomposition module, a path planning module, and a task allocation module. The task decomposition module breaks down the allocation command into several atomic operation units; the path planning module generates a trajectory for each atomic operation unit that meets energy consumption constraints and obstacle avoidance requirements; and the task allocation module uses a greedy strategy based on capability matching degree to assign equipment to the heterogeneous handling equipment group.

4. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 3, characterized in that, The path planning module uses an improved A algorithm combined with a dynamic window method to generate the trajectory. The cost function of the improved A algorithm is: in, This represents the actual cost from the starting point to the current node. This represents the heuristic cost from the current node to the destination. This represents the estimated energy consumption from the current node to the target node. This indicates the probability of the path segment being occupied in the next thirty seconds. Energy consumption weighting factor The congestion prediction factor is used; the occupancy probability is calculated in real time from historical traffic data using a sliding window statistical model.

5. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 3, characterized in that, When assigning devices, the task allocation module selects the execution device based on the capability matching degree M. The formula for calculating the capability matching degree M is as follows: in, This is the Euclidean distance between the current position of the device and the starting point of the mission. The preset distance smoothing factor, This represents the device's remaining battery life. The estimated power consumption for the mission, This is the compatibility coefficient between the device's current load status and the task load type. and Both are maximum-minimum normalization functions, used to map the distance factor and energy consumption factor to the [0,1] interval, respectively; For the preset weighting coefficients, satisfy .

6. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 3, characterized in that, The conflict resolution module adopts a dynamic reservation mechanism based on a spatiotemporal resource map, which maps the trajectory output by the path planning module to a spatiotemporal occupancy block; when an impending path conflict is detected, the priorities of the conflicting tasks are compared: the handling equipment corresponding to the higher priority task keeps its original trajectory unchanged; For low-priority tasks, the handling equipment triggers a local adjustment strategy, calculating the cost function of waiting in place or detouring at the current node. If the waiting time is less than a preset threshold, a waiting instruction is inserted and the trajectory timestamp is updated; otherwise, the real-time scheduling engine is triggered to replan.

7. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 1, characterized in that, The infrastructure linkage unit includes an elevator control interface module, a hoist coordination module, and an access control linkage module. The elevator control interface module is used to listen for floor request signals, read the elevator status through the Modbus TCP protocol, and send a call command and lock the right to use the elevator after confirming that the elevator is free or about to arrive. The hoist coordination module is used to receive hoisting requests and, after verifying that the material weight is within the rated load range and the pallet has completely entered the platform area, sends a start signal to the hoist controller. The access control linkage module is used to verify the validity of the task ID and access permission when the moving equipment approaches the controlled channel, and sends an unlocking command after successful verification.

8. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 1, characterized in that, The industrial IoT sensing node array consists of a fixed RFID reader, a lidar, an infrared sensor, and a pressure sensing pad. The fixed RFID reader uses a backscatter communication mechanism to read the electronic tags of the materials; the lidar generates a two-dimensional point cloud map of the local environment at a frequency of ten frames per second; the pressure sensing pad is installed on the surface of the loading and unloading platform to monitor the material pressure value. If the pressure value does not reach the threshold within a preset time, it is determined as a loading failure and a retry process is triggered.

9. The intelligent material allocation management system based on the Industrial Internet of Things as described in claim 1, characterized in that, The system also includes a digital twin simulation platform running on the server where the central collaborative controller is located. The digital twin simulation platform includes an entity model and a logical model. The digital twin simulation platform periodically receives actual operating data from the edge computing gateway cluster, including task execution time, path deviation, and device energy consumption, and compares the deviation with the simulation results. When the task completion time deviation exceeds a preset threshold or the path deviation exceeds the allowable range, the strategy adjustment mechanism is automatically triggered to retrain the heuristic rule base in the scheduling engine.

10. The intelligent material allocation management system based on the Industrial Internet of Things according to claim 3, characterized in that, After detecting the device fault abnormal event flag, the edge computing gateway executes a relay takeover mechanism: it removes the faulty device from the available resource pool, searches for the nearest available device near the fault point and assigns it as the relay executor, updates the starting coordinates of the relay task to the real-time location when the fault occurred, and the relay executor continues to complete the remaining journey.

Images