A production material precise allocation method under an industrial internet of things architecture

Abstract

This invention discloses a method for precise material allocation in an industrial Internet of Things (IIoT) architecture. The method first constructs a multi-source data acquisition link by deploying edge computing nodes to acquire real-time workstation operating status, product codes, and material reserves. Next, a lightweight dynamic scheduling engine preprocesses the heterogeneous data and calculates production cycle deviation and material consumption trends in real-time based on product completion timestamps. Then, when a material shortage risk is identified, the priority of delivery tasks is dynamically increased, and global collaborative optimization calculations are performed based on the geographical location, load status, and task queue of automated guided vehicles (AGVs). Finally, conflict-free shortest paths are planned, and control commands are issued to drive the delivery operation. This invention achieves real-time closed-loop response between production demand and material supply, effectively reducing production line downtime due to material shortages, significantly improving the efficiency of logistics resource allocation and the adaptability of flexible production, and ensuring the real-time performance and reliability of material allocation in complex environments.

Description

Technical Field

[0001] This invention belongs to the field of industrial Internet of Things (IIoT), specifically, it relates to a method for precise allocation of production materials under an IIoT architecture. Background Technology

[0002] With the rapid development of Industrial Internet of Things (IIoT) technology, its application in modern manufacturing has become a key driving force for promoting the transformation and upgrading of intelligent manufacturing. By integrating sensing, edge computing, and network communication technologies, IIoT can achieve comprehensive interconnection and real-time interaction of production factors, significantly improving the transparency and operational efficiency of manufacturing systems. Especially in highly automated and intelligent production environments, material allocation, as a key hub connecting the supply and production ends, directly affects the continuity of production and the scientific nature of resource allocation due to its level of intelligence.

[0003] Among them, the precise allocation method for production materials under the Industrial Internet of Things (IIoT) architecture aims to achieve deep coupling between production demand and logistics supply by sensing the real-time operating status of the production line and material consumption trends. This technology relies on distributed sensing nodes and efficient scheduling algorithms to transform the real-time production cycle and work-in-process inventory of each workstation into precise logistics instructions to support the flexible operation requirements in complex production scenarios.

[0004] However, existing technologies, primarily relying on pre-set fixed plans or localized automated control logic, struggle to cope with the high-frequency fluctuations in production lines caused by multi-variety and small-batch production. Traditional material scheduling systems suffer from severe information silos, failing to capture real-time dynamic deviations between actual production pace and buffer capacity at workstations, resulting in significant time lags between delivery instructions and actual production demands. Furthermore, traditional path planning and task allocation lack global collaborative optimization of all logistics resources, making adaptive adjustments difficult in the event of sudden changes in production rates or equipment malfunctions. In addition, data processing capabilities are limited by the communication latency of centralized architectures, failing to meet the stringent requirements of highly flexible production lines for real-time feedback and closed-loop control of material allocation. Summary of the Invention

[0005] The purpose of this invention is to provide a method for precise allocation of production materials under an industrial Internet of Things (IoT) architecture. This method mainly addresses the problem of severe information silos in traditional material scheduling systems, which cannot obtain the dynamic deviation between the actual production cycle and buffer capacity of workstations in real time, resulting in a significant time lag between delivery instructions and actual production needs.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for precise allocation of production materials under an industrial Internet of Things (IIoT) architecture includes the following steps:

[0008] S1 acquires the real-time operating status bit of each workstation equipment, the specific model code of the currently processed product, and the original data of the timestamp of each process completion through the industrial communication link.

[0009] S2 preprocesses the massive heterogeneous data collected, transforming the real-time quantity of work-in-process at each workstation, the precise remaining amount of the material buffer, and the instantaneous cycle time of equipment operation into a structured data stream, which serves as the input feature vector for subsequent scheduling algorithms.

[0010] S3 calculates the actual production cycle time based on the completion timestamp of continuously ordered products, and compares it with the theoretical production cycle time in real time to obtain the deviation value of the actual cycle time of each workstation; combined with the material buffer reserve change rate, the future material consumption trend and material shortage risk level of the workstation are calculated through the prediction algorithm.

[0011] S4. When it is detected that the actual cycle time of the workstation is faster than the theoretical cycle time and the material buffer reserve is lower than the safety stock warning line, the dynamic scheduling engine automatically increases the priority of the corresponding material delivery and performs global collaborative optimization in combination with AGV position, load and task queue.

[0012] Based on the electronic map of the entire site and real-time traffic flow, S5 dynamically plans the shortest conflict-free path for high-priority delivery tasks and sends instructions to the AGV scheduling system through the communication protocol to drive the AGV to deliver in advance or at an accelerated pace, thereby achieving a real-time closed-loop response between production demand and material supply.

[0013] Furthermore, in step S1, an industrial gateway and edge controller with high-performance computing capabilities are deployed near the edge of the production line. At the same time, a bidirectional communication connection is established between the industrial gateway and the field programmable logic controller of each workstation using an industrial communication link to achieve real-time data acquisition. Meanwhile, photoelectric sensors and weighing modules are installed in the material buffer to monitor the real-time remaining amount of materials.

[0014] Furthermore, in step S2, the data preprocessing includes cleaning, normalizing, and aligning the heterogeneous data based on a unified time reference.

[0015] Further, in step S3, the calculation process of the actual production cycle time is as follows: record the time points when a predetermined number of products pass through the end detection point of the workstation, and define them as the first time point, the second time point, and the third time point respectively; calculate the difference between the second time point and the first time point, and calculate the difference between the third time point and the second time point; take the arithmetic mean of the differences as the current actual production cycle time.

[0016] The actual production cycle deviation is calculated as follows: actual production cycle minus theoretical production cycle. When the actual production cycle deviation is negative, the production progress is determined to be ahead of schedule, and the system automatically shortens the expected trigger time for material delivery. When the actual production cycle deviation is positive, the production progress is determined to be behind schedule, and the system delays the delivery of non-urgent materials to optimize the utilization rate of logistics resources.

[0017] Furthermore, in this invention, the prediction of the material consumption trend is based on the quotient of the current buffer balance and the actual production cycle time, where the quotient represents the remaining production time that the current inventory can support. When the remaining production time is less than the sum of the estimated travel time of the AGV from the warehouse to the workstation and the predetermined safety buffer balance, the system immediately triggers the highest level of allocation warning. The prediction algorithm also combines the real-time change rate of the material buffer balance to establish a prediction model based on the remaining available time. The prediction model divides the total amount of remaining material in the current buffer by the real-time calculated average material consumption per unit time to obtain the remaining production time that the current inventory can support.

[0018] Furthermore, in this invention, the logic for prioritizing delivery adopts a multi-factor weighted evaluation model. The weighting factors of the multi-factor weighted evaluation model include cycle time deviation weight, inventory balance weight, workstation importance weight, and material scarcity weight. The inventory balance weight occupies a first preset proportion, and the cycle time deviation weight occupies a second preset proportion, ensuring that the workstation receives priority in logistics resource allocation when inventory is insufficient and production is accelerated. The multi-factor weighted evaluation model calculates priority evaluation indicators to respond to the material needs of the workstation when inventory is critically low and production is continuously accelerating.

[0019] Furthermore, in this invention, the specific process of global collaborative optimization is as follows: finding the optimal matching pair in the candidate task queue, and under the premise of satisfying the high-priority delivery task, selecting an automated guided vehicle that is within a preset distance from the target workstation, has sufficient power, and matches the load for task assignment.

[0020] Furthermore, in this invention, the shortest conflict-free path planning employs an improved heuristic search algorithm. When calculating path costs, it comprehensively considers physical distance, the distribution density of other mobile devices on the path, turning radius costs, and congestion nodes to ensure that the planned path is in an optimal state in the time dimension. The improved heuristic search algorithm introduces a path heat factor, which is dynamically adjusted based on the current distribution density of other mobile devices on the road segment, the speed reduction costs caused by turning radii, and the potential congestion probability of intersections. The path generated by the system consists of a series of coordinate point sequences and action commands, used to guide the AGV to avoid temporary obstacles during travel.

[0021] Furthermore, in this invention, the communication protocol includes a message queue telemetry transmission protocol or an object-oriented unified process control architecture; the message queue telemetry transmission protocol adopts a publish-subscribe model, and the scheduling instructions are encapsulated in a payload of a specific format, the payload including the target workstation number, material type code, delivery priority value, and suggested driving path point sequence; the communication link has an automatic reconnection mechanism to ensure the reliability of instruction issuance when the wireless network signal fluctuates; the object-oriented unified process control architecture adopts secure encrypted transmission and supports digital certificate authentication to ensure network security of scheduling instructions during transmission; after receiving the instruction, the automated guided vehicle scheduling system completes the task allocation logic within a preset processing time limit and sends motion control instructions to the designated automated guided vehicle terminal via a wireless local area network.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] (1) This invention achieves millisecond-level linkage between workstation production status, material balance and delivery instructions by real-time acquisition of multi-source data on the edge side and dynamic calculation of cycle deviation. It completely solves the time lag problem of mismatch between delivery demand and actual production rhythm, effectively improves the on-time delivery rate of materials and greatly reduces the risk of production line shutdown due to material shortage.

[0024] (2) This invention constructs a multi-factor weighted priority evaluation and AGV global collaborative scheduling mechanism, which dynamically adjusts the delivery priority based on factors such as cycle deviation, inventory balance, workstation importance, and material shortage. At the same time, it plans the optimal conflict-free path based on the overall traffic flow status. In the high-frequency fluctuation scenario of multi-variety small-batch production, it can effectively improve the utilization rate of logistics resources and perfectly adapt to the dynamic adjustment needs of flexible production.

[0025] (3) The present invention adopts an edge-side distributed computing architecture to replace the traditional centralized scheduling mode, and pushes the core logic such as data preprocessing, cycle time calculation, and path planning down to the edge nodes of the production line, compressing the overall system response delay to within 500ms; at the same time, it supports self-learning and iterative adaptation to the long-term changes in the production line cycle time, without the need for frequent manual adjustment of configuration parameters, which significantly reduces the system operation and maintenance costs.

[0026] (4) The invention supports multi-protocol secure encrypted transmission and automatic link reconnection mechanism, adapts to complex electromagnetic environment and network fluctuation scenarios in industrial sites, improves the reliability of instruction transmission by two orders of magnitude compared with traditional solutions, can ensure the stable operation of material distribution links under high-load production scenarios, and can still maintain stable performance output in flexible assembly lines with more than 20 workstations and in scenarios with more than 10 AGVs operating concurrently. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0028] Figure 2 This is a flowchart of the material consumption trend prediction algorithm in an embodiment of the present invention. Detailed Implementation

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0030] like Figure 1 As shown, this invention discloses a method for precise allocation of production materials under an Industrial Internet of Things (IIoT) architecture. The method first deploys IIoT edge computing nodes and constructs a multi-source data acquisition link. An industrial gateway and edge controller with high-performance computing capabilities are deployed near the edge of the production line. The industrial gateway uses a multi-core high-performance processor with a clock speed of at least 1.5 GHz and a memory capacity of at least 8 GB to meet the real-time processing needs of massive amounts of data at the edge. The industrial gateway establishes a physical layer connection with the field-programmable logic controllers (FPGAs) at each workstation via shielded twisted-pair cables or optical fibers, ensuring that the data transmission error rate is below a preset extremely low threshold even in complex industrial electromagnetic interference environments. At the communication protocol level, the industrial gateway establishes a bidirectional communication handshake with the FPGAs at each workstation through fieldbus technologies such as Industrial Ethernet, Profinet, EtherCAT, or Modbus TCP. Through an open user communication protocol, the industrial gateway polls or subscribes to the operating status bits of each workstation's equipment in real time, including but not limited to equipment readiness signals, processing signals, fault alarm signals, and the specific model code of the currently processed product. Meanwhile, the industrial gateway accurately captures the original timestamp data of each process completion. This timestamp is generated by the internal millisecond-level timer of the field programmable logic controller and is encapsulated into a data packet the instant the process completion signal is triggered.

[0031] In this embodiment, the data acquisition sampling period is set within a preset range of 10ms to 50ms to ensure the ability to capture instantaneous state changes on the high-frequency production line. A multi-dimensional sensing array is deployed in the material buffer zone. The photoelectric sensors installed in the material buffer zone utilize the principle of infrared diffuse reflection, with highly integrated transmitter and receiver, and a response time of less than or equal to 1ms, used for non-contact detection of the material tray's position and departure status. Simultaneously, a weighing module is embedded below the material storage support. The weighing module uses a 4-bridge high-precision pressure sensor, achieving a measurement accuracy of 10g and a range covering a predetermined interval of 0 to 500kg. The analog voltage signal output by the weighing module is converted into a 16-bit or 24-bit digital signal via a high-bit-width analog-to-digital converter and transmitted to the edge controller via an RS485 bus or CAN bus. The edge controller, through real-time monitoring of the weight data and combining it with a preset weight benchmark for individual materials, achieves real-time digital representation of the remaining material quantity.

[0032] In this embodiment, a lightweight dynamic scheduling engine is run and data preprocessing is performed. The lightweight dynamic scheduling engine is deployed in a Linux containerized computing environment on an industrial gateway or edge controller. This engine is designed based on a microservice architecture, with each functional module running independently as a Docker container, supporting dynamic horizontal scaling according to production line load. After the engine starts, its primary task is to clean the massive amounts of heterogeneous data from different workstations and using different protocols. The data cleaning process includes using median filtering or moving average filtering algorithms to remove abnormal spikes caused by sensor electromagnetic jitter. To address the issue of inconsistent time bases, the engine adopts an alignment mechanism based on a network time protocol, using the edge controller's master clock as the reference, to timestamp normalize all sensor data and field-programmable logic controller status data.

[0033] In this embodiment, data preprocessing further includes converting unstructured equipment logs into structured data streams. The real-time quantity of work-in-process at each workstation, the precise remaining amount in the material buffer, and the instantaneous cycle time of equipment operation are transformed into high-dimensional feature vectors. These feature vectors serve as input to subsequent scheduling algorithms, and their dimensions cover multiple key evaluation indicators such as production progress, inventory level, and equipment efficiency. Through offline analysis of massive historical data, the system pre-trains a normalization operator to map data of different dimensions to a numerical range of 0 to 1, eliminating the impact of magnitude differences on scheduling decisions.

[0034] In this embodiment, the system calculates the production cycle deviation and material consumption dynamic trend in real time, and obtains the product completion time sequence in real time by listening to the trigger signal of the detection point at the end of the workstation. Specifically, the calculation logic of the actual production cycle is as follows: record the time points when three consecutive products pass through the detection point at the end of the workstation as the first time T1, the second time T2, and the third time T3, and calculate the current actual production cycle T using the following formula. act :

[0035]

[0036] After obtaining the actual production cycle time, the system compares it with the theoretical production cycle time T preset in the edge node memory. std Real-time comparison is performed to obtain the actual cycle time deviation value of each workstation. T act -T std When the deviation value When the value is negative, it indicates that the current production progress is ahead of schedule. The system will automatically shorten the expected trigger time for material delivery to prevent the risk of material shortages caused by accelerated production. When the deviation value is negative... When the value is positive, it indicates that the production schedule is lagging behind, and the system will appropriately delay the delivery of non-urgent materials in order to optimize logistics route resources.

[0037] like Figure 2 As shown, in this embodiment, the prediction of material consumption trends relies not only on the current cycle time deviation but also on the real-time change rate of the material buffer reserve. The system establishes a prediction model based on remaining available time. This model divides the total remaining material in the current buffer by the real-time calculated average material consumption per minute to obtain the remaining production time that the current inventory can support. When this remaining time is less than the estimated average travel time of the automated guided vehicle (AGV) from the central warehouse to the workstation plus the predetermined safety buffer reserve, the system immediately marks the workstation as having a high material shortage risk level.

[0038] When the lightweight dynamic scheduling engine detects that the actual production cycle time of a certain workstation is faster than the theoretical cycle time, and the corresponding material buffer reserve is lower than the preset 15% safety stock warning line, a priority reconfiguration mechanism is triggered. The task priority improvement logic adopts a multi-factor weighted evaluation model, and the calculation logic of its priority evaluation index P is as follows:

[0039]

[0040] in, For the rhythm deviation weight, Weighted by inventory balance. As the weight of workstation importance, Assigning a weight to the scarcity level of materials, satisfying In this embodiment, inventory balance weighting The first preset proportion, accounting for 40%, is the beat deviation weight. The second preset proportion occupies 30%. The nonlinear mapping function for the beat deviation, This is the inverse function of the remaining inventory percentage. Using this model, the system can ensure that the material requirements of a workstation receive the highest level of response during critical moments when inventory is running low and production is accelerating. I represents the workstation's importance metric, ranging from 0 to 1, with higher values ​​for core production workstations. M represents the material scarcity metric, ranging from 0 to 1, with higher values ​​for irreplaceable / scarce materials.

[0041] In this embodiment, via an industrial wireless local area network, edge nodes acquire the current geographical coordinates, real-time remaining battery power, current speed, steering angle, and type and weight of materials carried by all automated guided vehicles (AGVs) every 100ms. The system maintains a local dynamic map at the edge, updating the motion vector of each AGV in the global coordinate system in real time. Through a global collaborative optimization algorithm, the system searches for the optimal matching pair in the candidate task queue; that is, under the premise of satisfying high-priority delivery tasks, it selects the AGV closest to the target workstation, with sufficient battery power and matching load for task assignment.

[0042] The system then plans the shortest, conflict-free path and issues control commands. Based on the overall electronic map information and real-time traffic flow conditions, the system dynamically plans the shortest, conflict-free path for the selected automated guided vehicles (AGVs). The planning algorithm employs an improved heuristic search algorithm, which considers not only physical distance but also a path heat factor when calculating path costs. The path heat factor is dynamically adjusted based on the current distribution density of other mobile devices on the road segment, the speed reduction cost caused by turning radii, and the potential congestion probability at intersections. The path generated by the system consists of a series of coordinate point sequences and action commands, ensuring that the AGV can avoid temporary obstacles and reduce unnecessary acceleration and deceleration during operation.

[0043] In this embodiment, the dispatching instructions are issued using a message queue telemetry transmission protocol or an object-oriented unified process control architecture. The dispatching instructions are encapsulated in a specific format payload, including the target workstation number, material type code, delivery priority value, and suggested route point sequence. The communication link has an automatic reconnection mechanism and a message acknowledgment mechanism to ensure the reliability of instruction issuance even in environments with fluctuating wireless network signals or multipath fading. After receiving the instruction, the automated guided vehicle (AGV) dispatching system completes the conversion of the underlying motion control logic within a preset processing time of 100ms and sends a drive signal to the designated AGV terminal via radio.

[0044] Furthermore, the lightweight dynamic scheduling engine possesses self-learning capabilities. By accumulating over 30 days of historical production data, the system automatically corrects the theoretical production cycle time baseline values ​​for different product models using machine learning algorithms. When a new model is introduced to the production line or the baseline cycle time shifts due to equipment aging, the system can detect this long-term trend and automatically update configuration parameters, enabling the system to adapt to frequent switching in multi-variety, small-batch production modes. This method is applied to flexible automated assembly lines with more than 20 processing stations, supporting concurrent collaborative operation of more than 10 automated guided vehicles. The overall task response latency of the system is less than 500ms, and the on-time material distribution rate reaches over 99.9%.

[0045] In specific application scenarios, such as automotive engine assembly lines, small parts like bolts and seals are consumed extremely quickly at each workstation, with large fluctuations in production cycle time. This embodiment achieves real-time sensing of material consumption by installing high-precision weighing sensors under the material boxes at each workstation. When the edge controller detects that only 50 bolts of a certain model remain in stock, and the current workstation's production cycle time increases from the preset 60 seconds / piece to 55 seconds / piece, the system immediately calculates that the remaining materials can only support approximately 45 minutes of production. Considering that the entire process of an automated guided vehicle (AGV) retrieving goods from the warehouse, traveling, avoiding obstacles, and unloading takes an average of 20 minutes, the system automatically prioritizes this task to the highest level and instructs the nearest empty AGV to perform the delivery, thereby avoiding production line downtime.

[0046] In summary, this invention constructs a closed-loop material distribution system from real-time sensing and intelligent prediction to precise execution by introducing edge computing and a dynamic scheduling engine into the industrial IoT architecture. This system not only solves the response lag problem in traditional distribution models but also achieves a high degree of synergy between production logistics and manufacturing processes through multi-factor weighted scheduling, dynamic path planning, and self-learning algorithms. In practical applications, this solution significantly improves the flexibility and adaptability of production lines, reduces work-in-process inventory, and minimizes the need for manual intervention, providing crucial technical support for building highly automated smart factories.

[0047] The above embodiments are merely one of the preferred embodiments of the present invention and should not be used to limit the scope of protection of the present invention. Any modifications or refinements made to the main design concept and spirit of the present invention that are not of substantial significance, but solve the same technical problem as the present invention, should be included within the scope of protection of the present invention.

Claims

1. A method for precise allocation of production materials under an industrial Internet of Things (IIoT) architecture, characterized in that, Includes the following steps: S1 acquires the real-time operating status bit of each workstation equipment, the specific model code of the currently processed product, and the original data of the timestamp of each process completion through the industrial communication link. S2 preprocesses the massive heterogeneous data collected, transforming the real-time quantity of work-in-process at each workstation, the precise remaining amount of the material buffer, and the instantaneous cycle time of equipment operation into a structured data stream, which serves as the input feature vector for subsequent scheduling algorithms. S3 calculates the actual production cycle time based on the completion timestamp of continuously ordered products, and compares it with the theoretical production cycle time in real time to obtain the deviation value of the actual cycle time of each workstation; combined with the material buffer reserve change rate, the future material consumption trend and material shortage risk level of the workstation are calculated through the prediction algorithm. S4. When it is detected that the actual cycle time of the workstation is faster than the theoretical cycle time and the material buffer reserve is lower than the safety stock warning line, the dynamic scheduling engine automatically increases the priority of the corresponding material delivery and performs global collaborative optimization in combination with AGV position, load and task queue. Based on the electronic map of the entire site and real-time traffic flow, S5 dynamically plans the shortest conflict-free path for high-priority delivery tasks and sends instructions to the AGV scheduling system through the communication protocol to drive the AGV to deliver in advance or at an accelerated pace, thereby achieving a real-time closed-loop response between production demand and material supply.

2. The method for precise allocation of production materials under an industrial Internet of Things (IoT) architecture according to claim 1, characterized in that, In step S1, an industrial gateway and edge controller with high-performance computing capabilities are deployed near the edge of the production line. At the same time, a bidirectional communication connection is established between the industrial gateway and the field programmable logic controllers of each workstation using an industrial communication link to achieve real-time data acquisition. Meanwhile, photoelectric sensors and weighing modules are installed in the material buffer to monitor the real-time remaining amount of materials.

3. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 2, characterized in that, In step S2, the data preprocessing includes cleaning, normalizing, and aligning heterogeneous data based on a unified time reference.

4. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 3, characterized in that, In step S3, the calculation process of the actual production cycle time is as follows: record the time points when a predetermined number of products pass through the end detection point of the workstation, and define them as the first moment, the second moment, and the third moment, respectively; calculate the difference between the second moment and the first moment, and calculate the difference between the third moment and the second moment; take the arithmetic mean of the differences as the current actual production cycle time. The actual production cycle deviation is calculated as follows: actual production cycle minus theoretical production cycle. When the actual production cycle deviation is negative, the production progress is determined to be ahead of schedule, and the system automatically shortens the expected trigger time for material delivery. When the actual production cycle deviation is positive, the production progress is determined to be behind schedule, and the system delays the delivery of non-urgent materials to optimize the utilization rate of logistics resources.

5. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 4, characterized in that, The prediction of material consumption trends is based on the quotient of the current buffer balance and the actual production cycle time. The quotient represents the remaining production time that the current inventory can support. When the remaining production time is less than the sum of the estimated travel time of the AGV from the warehouse to the workstation and the predetermined safety buffer balance, the system immediately triggers the highest level of allocation warning. The prediction algorithm also combines the real-time change rate of the material buffer balance to establish a prediction model based on the remaining available time. The prediction model divides the total amount of remaining material in the current buffer by the real-time calculated average material consumption per unit time to obtain the remaining production time that the current inventory can support.

6. The method for precise distribution of production materials under an industrial Internet of Things (IoT) architecture according to claim 5, characterized in that, The logic for prioritizing delivery adopts a multi-factor weighted evaluation model. The weighting factors of the multi-factor weighted evaluation model include cycle time deviation weight, inventory balance weight, workstation importance weight, and material scarcity weight. Among them, the inventory balance weight occupies a first preset proportion, and the cycle time deviation weight occupies a second preset proportion, which is used to ensure that the workstation receives priority logistics resource allocation when inventory is insufficient and production is accelerated. The multi-factor weighted evaluation model responds to the material demand of the workstation when inventory is critically low and production continues to accelerate by calculating priority evaluation indicators.

7. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 6, characterized in that, The specific process of global collaborative optimization is as follows: find the optimal matching pair in the candidate task queue, and under the premise of satisfying the high-priority delivery task, select an automated guided vehicle that is within a preset distance from the target workstation, has sufficient power and a matching load for task assignment.

8. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 7, characterized in that, The shortest conflict-free path planning employs an improved heuristic search algorithm. When calculating path costs, it comprehensively considers physical distance, the distribution density of other mobile devices along the path, turning radius costs, and congestion nodes to ensure that the planned path is optimal in the time dimension. The improved heuristic search algorithm introduces a path heat factor, which is dynamically adjusted based on the current distribution density of other mobile devices on the road segment, the speed reduction costs caused by turning radii, and the potential congestion probability at intersections. The path generated by the system consists of a series of coordinate point sequences and action commands, used to guide the AGV to avoid temporary obstacles during travel.

9. The method for precise allocation of production materials under an industrial Internet of Things architecture according to claim 8, characterized in that, The communication protocol includes a message queue telemetry transmission protocol or an object-oriented unified process control architecture. The message queue telemetry transmission protocol adopts a publish-subscribe model, and the scheduling instructions are encapsulated in a specific format payload, which includes the target workstation number, material type code, delivery priority value, and suggested driving route point sequence. The communication link has an automatic reconnection mechanism to ensure the reliability of instruction issuance when the wireless network signal fluctuates. The object-oriented unified process control architecture uses secure encrypted transmission and supports digital certificate authentication to ensure network security of scheduling instructions during transmission. After receiving the instruction, the automated guided vehicle (AGV) dispatching system completes the task allocation logic within a preset processing time limit and sends motion control instructions to the designated AGV terminal via a wireless local area network.

Images