Coal storage dynamic partition management and intelligent scheduling system and method

Abstract

The invention discloses a coal storage dynamic partition management and intelligent scheduling system and method, and relates to the technical field of intelligent storage and logistics management. According to the method, a coal yard live-action model is constructed through laser scanning and data integration, and an intelligent digital twinborn model is generated by combining historical operation data training; creating a virtual honeycomb partition based on coal attributes and storage requirements, and dynamically updating the state of the virtual honeycomb partition by fusing multi-source real-time data; task, resource and equipment agents are introduced, an optimal scheduling scheme is simulated and generated through a game coordination mechanism, and simulation verification is carried out in a digital twin environment; finally, the scheme is decomposed into control instructions to be issued and executed, a real-time monitoring and dynamic rescheduling mechanism is established, and closed-loop optimization is achieved. According to the system, the problems of partition stiffness, scheduling lag and the like in traditional coal warehouse management are effectively solved, and the efficiency, the safety and the economical efficiency of warehouse operation are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent warehousing and logistics management technology, and more specifically, to a dynamic zoning management and intelligent scheduling system and method for coal storage. Background Technology

[0002] With the continuous expansion of the energy supply chain and the increasing demands for operational efficiency, the traditional extensive coal storage management model is no longer adequate for modern logistics needs. It commonly suffers from low yard utilization, inefficient coal handling and distribution, delayed updates to inventory and consumption information, and losses and environmental pollution caused by coal spontaneous combustion and weathering. While traditional systems have achieved preliminary zoning management, their "dynamic" response capabilities are insufficient, with a core drawback being their reliance on static and outdated data. Systems typically rely on limited, historical, manually entered information for zoning planning and scheduling, failing to perceive real-time changes in coal pile temperature, quantity, and quality. This leads to rigid zoning strategies and a disconnect between scheduling instructions and actual on-site conditions. This results in suboptimal stacker-reclaimer operating paths, traffic congestion within the yard, mixing of different coal types, and the inability to promptly handle high-temperature coal, among other safety hazards. Essentially, this stems from a lack of a closed loop of comprehensive perception, intelligent prediction, and autonomous decision-making, significantly reducing the overall system's operational efficiency, safety, and economy. Summary of the Invention

[0003] The main objective of this invention is to provide a dynamic zoning management and intelligent scheduling system and method for coal storage, thereby addressing at least the problems of rigid zoning strategies and the disconnect between scheduling instructions and actual on-site conditions in existing technologies. This improves the operational efficiency, security, and economy of zoning management and intelligent scheduling, ensuring the safe and efficient operation of the national energy supply chain.

[0004] To achieve the above objectives, a dynamic zoning management and intelligent scheduling system and method for coal storage is provided.

[0005] In a first aspect, the present invention provides a dynamic zoning management and intelligent scheduling system for coal storage, the system comprising: The initialization and data preparation unit constructs a real-world model of the coal yard through laser scanning and data integration, sets the parameters of the coal yard operation equipment in the real-world model to obtain a digital twin model, and trains the digital twin model using the historical operation data of the coal yard operation equipment to obtain an intelligent digital twin model. The situational awareness and digital synchronization unit is connected to the initialization and data preparation unit. The situational awareness and digital synchronization unit is used to collect multi-source data, create virtual cells in the intelligent digital twin model according to the properties of coal, storage requirements and operation plans, update the status data of different virtual cells in real time according to multi-source data, obtain new tasks of the coal yard and create digital task work orders for each new task. The intelligent game and decision generation unit is connected to the situational awareness and digital synchronization unit. The intelligent game and decision generation unit is used to construct task intelligent agents, resource intelligent agents and equipment intelligent agents, and to simulate the scheduling scheme based on task intelligent agents, resource intelligent agents and equipment intelligent agents through the game coordinator to obtain the optimal candidate scheme. The optimal candidate scheme is determined by simulating and deducing the optimal candidate scheme using the intelligent digital twin model to obtain the optimal scheme. The precise execution and dynamic optimization unit is connected to the intelligent game and decision generation unit. The precise execution and dynamic optimization unit is used to decompose the optimal candidate solution into control commands and issue control commands, preset safety thresholds and compare multi-source data with the preset safety thresholds in real time to obtain comparison results, and issue corresponding scheduling requests based on the comparison results.

[0006] Specifically, the initialization and data preparation unit includes: The digital twin model building module is used to collect raw data through vehicle-mounted LiDAR, drones and GPS and use the raw data to generate a real-world model that is completely consistent with the physical geometry and physical properties of the coal yard in the same engineering coordinate system. The engine is deployed and the data interface is set to obtain the digital twin model. The rules and constraints configuration module is connected to the digital twin model building module. The rules and constraints configuration module is used to add physical constraint configurations and business rule configurations to the digital twin model to obtain a rule-based digital twin model. The historical data learning module is connected to the rule and constraint configuration module. The historical data learning module is used to input historical operation data into the rule digital twin model, process the historical operation data and extract features to obtain historical operation features, and use the historical operation features to train the rule digital twin model to obtain an intelligent digital twin model.

[0007] Specifically, the digital twin model building module includes: The data acquisition and processing submodule is used to collect laser point cloud data of the coal yard using vehicle-mounted lidar, collect multi-angle high-definition images of the coal yard using drones to generate a 3D real scene model of the coal yard, use GPS to map the precise coordinate positioning of the fixed facilities in the coal yard, obtain the 3D CAD model of the coal yard operation equipment, and integrate the laser point cloud data, 3D real scene model, precise coordinate positioning of fixed facilities, and 3D CAD model of coal yard operation equipment into a real scene model. The 3D engine deployment and integration submodule connects with the data acquisition and processing submodule. The 3D engine deployment and integration submodule imports the reality model into the project created in the 3D engine, assigns an identifier to each object in the reality model, and creates an interface for each object to obtain a digital twin model so that the attribute data of each object can be changed.

[0008] Specifically, the rules and constraints configuration module includes: The physical constraint configuration submodule is connected to the digital twin model construction module. The physical constraint configuration submodule is used to establish the kinematic model of the coal yard operation equipment and add collision body models to the coal yard operation equipment, and input the performance parameters of the coal yard operation equipment. The business rules configuration submodule is connected to the physical constraint configuration submodule. The business rules configuration submodule is used to set the stacking rules, path rules and business processes in the digital twin model to obtain the rule digital twin model.

[0009] Specifically, the historical data learning module includes: The data governance and preprocessing submodule is connected to the rules and constraints configuration module. The data governance and preprocessing submodule is used to access historical operation data into the rule digital twin model. The historical operation data includes historical operation work orders, monitoring data, metering data, test data and meteorological data of coal yard operation equipment. After data cleaning and timestamp alignment of the historical operation data, feature extraction is performed to obtain historical operation features. The model pre-training submodule is connected to the data governance and preprocessing submodule. The model pre-training submodule is used to train the rule-based digital twin model using historical job features to obtain the intelligent digital twin model.

[0010] Specifically, the situational awareness and digital synchronization unit includes: The multi-source data fusion module is connected to the initialization and data preparation unit. The multi-source data fusion module is used to fuse the UAV scanning data, the data collected by the sensors on the coal yard operation equipment, the GPS positioning data, and the business data of the coal yard operation equipment into multi-source data and to connect the multi-source data to the intelligent digital twin model in real time. The virtual cell status update module is connected to the multi-source data fusion module. The virtual cell status update module is used to dynamically adjust the geometry, inventory, physical attributes and chemical attributes of the virtual cells in the intelligent digital twin model according to the multi-source data, and mark the idle status, occupied status, waiting status and warning status of the virtual cells. The task instruction listening module is connected to the virtual cell status update module. The task instruction listening module is used to capture new tasks in the coal yard and create a digital task work order for each new task.

[0011] Specifically, the intelligent game and decision generation unit includes: The multi-agent market construction module is connected to the situational awareness and digital synchronization unit. The multi-agent market construction module is used to construct task agents, resource agents, and equipment agents. The task agent represents each work task, the resource agent represents each virtual cell, and the equipment agent represents each coal yard operation equipment. The distributed collaborative game module is connected to the multi-agent market construction module. The distributed collaborative game module is used to send price requests to the device agent in the virtual market using the task agent, and to collect all bids and evaluate conflicts using the game coordinator to obtain the optimal candidate solution. The simulation and scheme confirmation module is connected to the distributed collaborative game module. The simulation and scheme confirmation module is used to simulate and deduce the optimal candidate scheme in the intelligent digital twin model, verify the feasibility of the optimal candidate scheme, and confirm the optimal candidate scheme as the only executable optimal scheme.

[0012] Specifically, the precise execution and dynamic optimization unit includes: The instruction serialization and distribution module is connected to the intelligent game and decision generation unit. The instruction serialization and distribution module is used to decompose the optimal solution into control instructions that can be recognized by the coal yard operation equipment, and distribute the control instructions to the corresponding coal yard operation equipment through the industrial Internet of Things platform. The real-time monitoring and closed-loop feedback module is connected to the instruction serialization and issuance module. The real-time monitoring and closed-loop feedback module is used to continuously receive the actual execution status of the coal yard operation equipment and compare the actual execution status with the expected status when the intelligent digital twin model simulates the optimal solution to ensure that the optimal solution is implemented correctly. The anomaly detection and dynamic rescheduling module is connected to the real-time monitoring and closed-loop feedback module. The anomaly detection and dynamic rescheduling module is used to monitor abnormal events in real time and preset safety thresholds. It compares multi-source data with the preset safety thresholds in real time to obtain comparison results and issues corresponding scheduling requests based on the comparison results.

[0013] Secondly, this invention provides a method for dynamic zoning management and intelligent scheduling of coal storage, wherein the scheduling method applies the scheduling system of the first aspect, and the scheduling method includes: A digital twin model is obtained by constructing a real-world model of the coal yard through laser scanning and data integration, and setting the parameters of the coal yard operation equipment in the real-world model. The digital twin model is then trained to obtain an intelligent digital twin model. Collect multi-source data and create virtual cells, update the status data of virtual cells based on multi-source data, obtain new tasks and create digital task work orders for each new task; Construct task intelligent agents, resource intelligent agents, and equipment intelligent agents. Use a game coordinator to simulate scheduling schemes to obtain the optimal candidate scheme. Use an intelligent digital twin model to simulate and deduce the optimal candidate scheme, determine the feasibility of the optimal candidate scheme, and obtain the optimal scheme. The optimal solution is decomposed into control commands, which are then issued and a safety threshold is preset. During the execution of the control commands, the multi-source data is compared with the preset safety threshold in real time to obtain the comparison results. Based on the comparison results, the corresponding scheduling request is issued.

[0014] Specifically, a digital twin model is obtained by constructing a real-world model of the coal yard through laser scanning and data integration, and setting the parameters of the coal yard operation equipment in the real-world model. The digital twin model is then trained to obtain an intelligent digital twin model, including: Raw data of the coal yard is collected by vehicle-mounted LiDAR, drones and GPS, and a real-world model of the coal yard is generated using the raw data. The engine is deployed and the data interface is set for the real-world model to obtain a digital twin model. Add physical constraint configurations and business rule configurations to the digital twin model to obtain a rule-based digital twin model; Historical operational data of coal yard equipment is integrated into a rule-based digital twin model. Historical operational features are extracted and used to train the rule-based digital twin model to obtain an intelligent digital twin model.

[0015] This application provides a dynamic zoning management and intelligent scheduling system and method for coal storage. The system first constructs and trains an intelligent digital twin model that accurately maps the physical coal yard using laser scanning. Then, using this model, it senses the coal status in real time and creates dynamically updated virtual zones (virtual cells), while simultaneously converting new tasks into digital work orders. A multi-agent game decision-making mechanism is introduced, allowing agents representing tasks, resources, and equipment to collaboratively compete in a virtual market, simulating and generating the optimal scheduling scheme. After verification through digital twin simulation, execution instructions are finally issued, with continuous monitoring and closed-loop optimization during execution. This achieves global intelligence, dynamic response, and precise execution in storage management. This invention at least solves the problems of rigid zoning strategies and the disconnect between scheduling instructions and actual on-site conditions in existing technologies. It further improves the operational efficiency, security, and economy of zoning management and intelligent scheduling, ensuring the security and efficient operation of the national energy supply chain. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 A connection diagram of a dynamic zoning management and intelligent scheduling system for coal storage provided in this application; Figure 2 A flowchart illustrating a dynamic zoning management and intelligent scheduling method for coal storage provided in this application; Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0019] In this invention, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0020] This application provides a dynamic zoning management and intelligent scheduling system and method for coal storage. The system first creates a three-dimensional dynamic model that perfectly corresponds to the physical coal yard through laser scanning and data fusion. Based on this, the system divides the coal yard into dynamically adjustable "virtual cells" for refined management and synchronizes the operating data of various equipment and the coal status in real time. When a new task arises, the system initiates an innovative collaborative decision-making mechanism: virtual agents representing "tasks," "equipment," and "coal piles" autonomously negotiate and simulate bidding to quickly determine the optimal scheduling scheme. Before execution, the scheme is simulated and verified in the digital model to ensure its feasibility and efficiency, ultimately forming a closed-loop management process encompassing perception, decision-making, execution, and continuous optimization.

[0021] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0022] Figure 1 A connection diagram of a dynamic zoning management and intelligent scheduling system for coal storage provided in this application is shown below. Figure 1As shown in this embodiment, a dynamic zoning management and intelligent scheduling system for coal storage is provided. The system includes: The initialization and data preparation unit constructs a real-world model of the coal yard through laser scanning and data integration, sets the parameters of the coal yard operation equipment in the real-world model to obtain a digital twin model, and trains the digital twin model using the historical operation data of the coal yard operation equipment to obtain an intelligent digital twin model. The situational awareness and digital synchronization unit is connected to the initialization and data preparation unit. The situational awareness and digital synchronization unit is used to collect multi-source data, create virtual cells in the intelligent digital twin model according to the properties of coal, storage requirements and operation plans, update the status data of different virtual cells in real time according to multi-source data, obtain new tasks of the coal yard and create digital task work orders for each new task. The intelligent game and decision generation unit is connected to the situational awareness and digital synchronization unit. The intelligent game and decision generation unit is used to construct task intelligent agents, resource intelligent agents and equipment intelligent agents, and to simulate the scheduling scheme based on task intelligent agents, resource intelligent agents and equipment intelligent agents through the game coordinator to obtain the optimal candidate scheme. The optimal candidate scheme is determined by simulating and deducing the optimal candidate scheme using the intelligent digital twin model to obtain the optimal scheme. The precise execution and dynamic optimization unit is connected to the intelligent game and decision generation unit. The precise execution and dynamic optimization unit is used to decompose the optimal candidate solution into control commands and issue control commands, preset safety thresholds and compare multi-source data with the preset safety thresholds in real time to obtain comparison results, and issue corresponding scheduling requests based on the comparison results.

[0023] This application provides a dynamic zoning management and intelligent scheduling system for coal storage, which systematically solves many shortcomings of traditional management models through intelligent steps. First, the method constructs a digital twin model that perfectly corresponds to the physical coal yard through high-precision laser scanning and data integration, and trains it using historical data to make it a predictive "intelligent brain." Next, the system dynamically divides the model into "virtual cells" based on coal properties for refined management, and integrates multi-source data from drones, sensors, etc., to achieve real-time perception and synchronization of the yard's status, ensuring the unity of the digital and physical worlds. Subsequently, in the core decision-making stage, the system introduces a multi-agent game mechanism, allowing virtual agents representing tasks, resources, and equipment to autonomously negotiate and simulate bidding, thereby quickly generating the optimal scheduling scheme under various constraints. Simulations are then performed in the digital twin environment to predict risks and verify feasibility. Finally, the scheme is decomposed into specific instructions and precisely issued to the equipment for execution. Closed-loop feedback and dynamic rescheduling are achieved through real-time monitoring and preset safety thresholds to ensure safe and reliable operation. This system represents a leap from static experience-based management to dynamic real-time optimization, significantly improving yard utilization and operational efficiency. Through simulation and intelligent game theory, scheduling decisions are transformed from passive response to proactive optimization, effectively avoiding equipment conflicts and path detours, and reducing operating costs. A closed-loop process of "perception-decision-execution-optimization" is constructed, significantly enhancing the system's adaptability to abnormal operating conditions and comprehensively improving the intelligence, safety, and economy of coal storage management.

[0024] Specifically, the initialization and data preparation unit includes: The digital twin model building module is used to collect raw data through vehicle-mounted LiDAR, drones and GPS and use the raw data to generate a real-world model that is completely consistent with the physical geometry and physical properties of the coal yard in the same engineering coordinate system. The engine is deployed and the data interface is set to obtain the digital twin model. The rules and constraints configuration module is connected to the digital twin model building module. The rules and constraints configuration module is used to add physical constraint configurations and business rule configurations to the digital twin model to obtain a rule-based digital twin model. The historical data learning module is connected to the rule and constraint configuration module. The historical data learning module is used to input historical operation data into the rule digital twin model, process the historical operation data and extract features to obtain historical operation features, and use the historical operation features to train the rule digital twin model to obtain an intelligent digital twin model.

[0025] In this embodiment, the initialization and data preparation unit systematically addresses the core issues of insufficient accuracy, missing rules, and low intelligence in the construction of digital models for coal storage through three key steps. First, the digital twin model construction module integrates multi-source data from vehicle-mounted LiDAR, drones, and GPS to generate a high-precision real-world model in a unified coordinate system. Through engine deployment and interface settings, it solves the problems of inaccurate digital mapping of physical scenes and data silos. Second, the rule and constraint configuration module injects equipment physical constraints and stacking / reclaiming business rules into the model, overcoming the drawback of the disconnect between the virtual model and actual on-site operational rules. Finally, the historical data learning module manages and extracts features from massive amounts of operational data and uses this data to train the model, enabling the digital twin to learn from historical experience and predict and optimize, breaking through the bottleneck of static models' inability to autonomously evolve and make intelligent decisions. The common advantage of this series of steps is that it constructs a dynamic digital twin that is not only similar in form but also possesses on-site rules and an intelligent core, laying a solid and reliable data and model foundation for subsequent accurate perception and intelligent decision-making.

[0026] Specifically, the digital twin model building module includes: The data acquisition and processing submodule is used to collect laser point cloud data of the coal yard using vehicle-mounted lidar, collect multi-angle high-definition images of the coal yard using drones to generate a 3D real scene model of the coal yard, use GPS to map the precise coordinate positioning of the fixed facilities in the coal yard, obtain the 3D CAD model of the coal yard operation equipment, and integrate the laser point cloud data, 3D real scene model, precise coordinate positioning of fixed facilities, and 3D CAD model of coal yard operation equipment into a real scene model. The 3D engine deployment and integration submodule connects with the data acquisition and processing submodule. The 3D engine deployment and integration submodule imports the reality model into the project created in the 3D engine, assigns an identifier to each object in the reality model, and creates an interface for each object to obtain a digital twin model so that the attribute data of each object can be changed.

[0027] In this embodiment, the digital twin model construction module systematically solves the key problems of single data, static models, and poor interactivity in 3D modeling of coal storage through two core steps. First, a vehicle-mounted LiDAR is used to perform a high-precision, no-blind-spot scan of the entire coal yard area, acquiring point cloud data with centimeter-level accuracy. Second, a drone is used to capture multi-angle high-definition images, generating a 3D reality model with realistic textures. Third, GPS mapping is used to accurately locate the coordinates of key facilities (such as tracks, turntables, and pile foundations). Finally, CAD models of major equipment such as stacker-reclaimers and belt conveyors are obtained from CAD models provided by equipment manufacturers.

[0028] Laser point cloud data is registered and fused with oblique photogrammetry models to form a unified coal yard basic terrain and scene model with both geometric accuracy and realistic texture. Point cloud processing software (Terrasolid) is used to automatically identify and classify ground, buildings, equipment, existing coal piles, etc. The classified point cloud is converted into a 3D mesh model (OBJ format) and vector boundary lines, significantly reducing the model size and improving rendering and computational efficiency. The equipment CAD model is lightweighted, removing internal parts and unnecessary details while retaining its appearance and kinematic structure, and then converted into a game engine. In the 3D engine, the processed equipment model is precisely placed onto the corresponding coordinates of the fused scene model. The entire scene is unified under a single Engineering Coordinate System (UTM) to ensure a one-to-one correspondence with real-world coordinates. A project is created in the 3D engine, the generated lightweight model is imported, and the scene graph is organized in a hierarchical and modular manner, with each object assigned a unique ID. An API interface is created in the engine for each interactive object (equipment, stack), allowing its position, status, and other attributes to be read and driven by external programs. This module creates a dynamic digital twin base that combines geometric accuracy with data interactivity, providing a realistic, reliable, and flexibly controllable virtual space environment for subsequent real-time monitoring, simulation, and intelligent decision-making. It solves the bottleneck of previous 3D models being unable to update data in real time and difficult to link with business systems.

[0029] Specifically, the rules and constraints configuration module includes: The physical constraint configuration submodule is connected to the digital twin model construction module. The physical constraint configuration submodule is used to establish the kinematic model of the coal yard operation equipment and add collision body models to the coal yard operation equipment, and input the performance parameters of the coal yard operation equipment. The business rules configuration submodule is connected to the physical constraint configuration submodule. The business rules configuration submodule is used to set the stacking rules, path rules and business processes in the digital twin model to obtain the rule digital twin model.

[0030] The rules and constraints configuration module in this embodiment effectively solves the core problems of physical distortion and missing business logic in digital twin models through two progressive steps. Specifically, it defines the kinematic chain of the stacker-reclaimer (trolley travel range, slewing angle range, cantilever pitch angle range). It defines the operating speed range and start / stop acceleration curves of the belt conveyor. It adds simplified collision bodies (such as cuboids and cylinders) to all equipment, buildings, and even coal piles. It sets safe operating distances according to the safety operation manuals of each piece of equipment and enters the rated power, maximum stacking / reclaiming capacity (tons / hour), and maximum travel speed of the equipment.

[0031] The system configures stacking and retrieval rules, including: layered stacking method, defining the height increment for each stacking operation; diamond-shaped retrieval method, defining the swing pattern and depth of the cantilever during retrieval; and a rule prohibiting mixed stacking, specifying the minimum physical distance between different coal types. Path rules are configured: defining the connectivity logic of the main conveyor belt and establishing a relationship diagram of "equipment-conveyor belt path-coal pile". Business processes are configured: standard work process templates are configured in the system, such as the various stages and state transitions of "coal unloading by train - storage in warehouse - retrieval and loading - loading onto ship". This configuration process successfully upgrades a purely geometric model into a "rule-based digital twin model" that conforms to both physical reality and business rules. This provides a realistic, reliable, and directly applicable virtual testbed for all subsequent simulations, game theory decisions, and automated scheduling, fundamentally ensuring the executability and security of the decision-making scheme.

[0032] Specifically, the historical data learning module includes: The data governance and preprocessing submodule is connected to the rules and constraints configuration module. The data governance and preprocessing submodule is used to access historical operation data into the rule digital twin model. The historical operation data includes historical operation work orders, monitoring data, metering data, test data and meteorological data of coal yard operation equipment. After data cleaning and timestamp alignment of the historical operation data, feature extraction is performed to obtain historical operation features. The model pre-training submodule is connected to the data governance and preprocessing submodule. The model pre-training submodule is used to train the rule-based digital twin model using historical job features to obtain the intelligent digital twin model.

[0033] This embodiment's historical data learning module systematically addresses the core issues of digital twin models lacking experience transfer and intelligent decision-making capabilities through two key steps. Specifically, data source access includes: a production execution system (retrieving historical work orders, start / end time, equipment, coal type, and quantity); an equipment monitoring system (retrieving equipment operation logs, fault records, and sensor data); a metering system (retrieving precise data from belt scales and truck scales); a laboratory system (retrieving historical coal quality data); and meteorological data (retrieving historical temperature, humidity, wind speed, and precipitation data). After data access, data cleaning and alignment are performed. Data cleaning includes handling missing values ​​and outliers (such as obviously incorrect equipment location coordinates). Timestamp alignment unifies data from all different sources onto the same timeline. Feature extraction includes extracting features from the operational data, such as "equipment idle distance," "task switching frequency," and "operational efficiency under specific weather conditions." Unstructured data (such as fault text descriptions) is encoded into model-readable labels. The extracted features (coal sulfur content, volatile matter, stockpiling time, average ambient temperature, and historical temperature sequence of the coal pile interior) are input into a Long Short-Term Memory (LSTM) network for training, resulting in a coal spontaneous combustion prediction model, which is then imported into a rule-based digital twin model. Historical data is replayed, allowing the agent to simulate operations over the past year in the digital twin environment. A multi-agent reinforcement learning algorithm is employed to enable the agents to learn collaborative strategies, thus forming a preliminary scheduling strategy. This method successfully upgrades a static model with only rules and physical constraints into an "intelligent digital twin model" capable of learning from historical experience and possessing predictive and optimization capabilities. This provides data-driven insights for subsequent scheduling decisions, significantly improving the accuracy and intelligence of the system's decision-making.

[0034] Specifically, the situational awareness and digital synchronization unit includes: The multi-source data fusion module is connected to the initialization and data preparation unit. The multi-source data fusion module is used to fuse the UAV scanning data, the data collected by the sensors on the coal yard operation equipment, the GPS positioning data, and the business data of the coal yard operation equipment into multi-source data and to connect the multi-source data to the intelligent digital twin model in real time. The virtual cell status update module is connected to the multi-source data fusion module. The virtual cell status update module is used to dynamically adjust the geometry, inventory, physical attributes and chemical attributes of the virtual cells in the intelligent digital twin model according to the multi-source data, and mark the idle status, occupied status, waiting status and warning status of the virtual cells. The task instruction listening module is connected to the virtual cell status update module. The task instruction listening module is used to capture new tasks in the coal yard and create a digital task work order for each new task.

[0035] This embodiment's situational awareness and digital synchronization unit systematically solves the key problems of lagging information updates, opaque yard status, and slow task response in coal storage management through three core steps. Specifically, it includes real-time access to data from drone scanning, equipment sensors, GPS positioning, and business systems (such as MES and ERP). It correlates and integrates fragmented physical information (such as equipment location, coal pile volume, and coal temperature) with business information (such as waybills and quality inspection reports). Based on the latest scanning data, it dynamically adjusts the geometry, inventory, and physicochemical properties of each "virtual cell" in the digital twin, identifying cells in different states such as idle, occupied, pending retrieval, and warning (e.g., excessively high temperature). It automatically captures newly arrived stacking / retrieval tasks and creates a digital task order for each new task, serving as the starting point for triggering intelligent decision-making. This step constructs a digital mirror that is synchronized with the physical world in real time and has full-element visibility, laying a solid data foundation for achieving refined management of storage resources, rapid response to operational tasks, and proactive early warning of safety risks.

[0036] Specifically, the intelligent game and decision generation unit includes: The multi-agent market construction module is connected to the situational awareness and digital synchronization unit. The multi-agent market construction module is used to construct task agents, resource agents, and equipment agents. The task agent represents each work task, the resource agent represents each virtual cell, and the equipment agent represents each coal yard operation equipment. The distributed collaborative game module is connected to the multi-agent market construction module. The distributed collaborative game module is used to send price requests to the device agent in the virtual market using the task agent, and to collect all bids and evaluate conflicts using the game coordinator to obtain the optimal candidate solution. The simulation and scheme confirmation module is connected to the distributed collaborative game module. The simulation and scheme confirmation module is used to simulate and deduce the optimal candidate scheme in the intelligent digital twin model, verify the feasibility of the optimal candidate scheme, and confirm the optimal candidate scheme as the only executable optimal scheme.

[0037] The intelligent game theory and decision generation unit in this embodiment systematically solves the key problems of difficult multi-objective conflict coordination, insufficient global optimization, and inability to predict the feasibility of solutions in coal storage scheduling through three core steps. Specifically, it includes: multi-agent market construction: Task agents: representing each task and carrying its objective (e.g., "store 5000 tons of high-calorific-value coal within 1 hour"). Resource agents: representing each virtual cell and publishing its "commodity" information (e.g., location, inventory, coal quality). Equipment agents: representing each stacker-reclaimer and publishing its "service" information (e.g., current location, working status, efficiency). Distributed collaborative game theory: Task agents, acting as "buyers," issue "inquiry requests" to resource and equipment agents in the virtual market. Each agent calculates the "virtual cost" or "benefit" of the service and submits a bid based on its own state and global rules. The game coordinator, acting as the "market manager," collects all bids, evaluates conflicts (e.g., path intersections, resource competition), and facilitates the matching of transactions that achieve the global objective (e.g., shortest total time, optimal total path, highest space utilization). Simulation and Solution Confirmation: The optimal candidate solutions generated by the game are simulated at millisecond-level speeds in a digital twin. The feasibility and robustness of the solutions are verified, ensuring no physical collisions or logical deadlocks. Finally, the uniquely executable optimal solution is confirmed. This step guarantees both the global optimality and multi-party coordination of scheduling decisions, and ensures the feasibility of the solution through digital twin simulation, ultimately achieving a fundamental transformation in coal storage scheduling from experience-based judgment to intelligent optimization, and from static planning to dynamic optimization.

[0038] Specifically, the precise execution and dynamic optimization unit includes: The instruction serialization and distribution module is connected to the intelligent game and decision generation unit. The instruction serialization and distribution module is used to decompose the optimal solution into control instructions that can be recognized by the coal yard operation equipment, and distribute the control instructions to the corresponding coal yard operation equipment through the industrial Internet of Things platform. The real-time monitoring and closed-loop feedback module is connected to the instruction serialization and issuance module. The real-time monitoring and closed-loop feedback module is used to continuously receive the actual execution status of the coal yard operation equipment and compare the actual execution status with the expected status when the intelligent digital twin model simulates the optimal solution to ensure that the optimal solution is implemented correctly. The anomaly detection and dynamic rescheduling module is connected to the real-time monitoring and closed-loop feedback module. The anomaly detection and dynamic rescheduling module is used to monitor abnormal events in real time and preset safety thresholds. It compares multi-source data with the preset safety thresholds in real time to obtain comparison results and issues corresponding scheduling requests based on the comparison results.

[0039] This embodiment's precise execution and dynamic optimization unit systematically solves the core problems of instruction execution deviation, lack of process monitoring, and delayed anomaly response in coal storage operations through three key steps. Specifically, these include: instruction serialization and issuance, which decomposes the verified scheduling scheme into a series of low-level control instructions recognizable by the equipment (e.g., starting the stacker conveyor; tilting the boom to 15°; moving the trolley to X=105, Y=300). These instructions are then securely and reliably issued to the corresponding stacker-reclaimers, conveyor belts, and other equipment via an industrial IoT platform. Real-time monitoring and closed-loop feedback continuously receive the actual execution status of the equipment (e.g., real-time location, current, flow rate) and compare it with the expected status in the digital twin. This forms a real-time closed loop of "perception-decision-execution-feedback," ensuring that operations proceed as planned. Anomaly detection and dynamic rescheduling monitor various abnormal events in real time (e.g., sudden drop in equipment efficiency, emergency order insertion, sudden weather changes). Once a deviation sufficient to affect the original plan is detected, a "rescheduling request" is immediately sent to the intelligent game and decision generation unit. The intelligent game and decision generation unit will quickly initiate a new, small-scale game based on the latest real-time situation, generate adjustment plans, and execute them, thereby achieving dynamic optimization and self-healing of the production process. This step ensures that the optimization plan can be implemented accurately and flawlessly, and establishes a state monitoring and adaptive adjustment mechanism throughout the entire process, thus significantly improving the accuracy, reliability, and immediate response capability to unexpected situations of the system.

[0040] Figure 2 A flowchart illustrating a dynamic zoning management and intelligent scheduling method for coal storage provided in this application is shown below. Figure 2 As shown, this embodiment provides a method for dynamic zoning management and intelligent scheduling of coal storage. This method applies... Figure 1 The embodiment describes a dynamic zoning management and intelligent scheduling method system for coal storage, the scheduling method comprising: S101: Construct a real-world model of the coal yard through laser scanning and data integration, and set the parameters of the coal yard operation equipment in the real-world model to obtain a digital twin model. Train the digital twin model to obtain an intelligent digital twin model. S102: Collect multi-source data and create virtual cells, update the status data of virtual cells according to multi-source data, obtain new tasks and create digital task work orders for each new task; S103: Construct task intelligent agents, resource intelligent agents, and equipment intelligent agents. Use a game coordinator to simulate scheduling schemes to obtain the optimal candidate scheme. Use an intelligent digital twin model to simulate and deduce the optimal candidate scheme, determine the feasibility of the optimal candidate scheme, and obtain the optimal scheme. S104: Decompose the optimal solution into control commands and issue control commands and preset safety thresholds. During the execution of control commands, compare multi-source data with the preset safety thresholds in real time to obtain comparison results, and issue corresponding scheduling requests based on the comparison results.

[0041] The proposed method for dynamic zoning management and intelligent scheduling of coal storage facilities establishes a complete intelligent management solution. This method first establishes a high-precision digital twin model, training it with laser scanning and historical data to form a predictive digital image. Then, it creates virtual cellular zones based on coal characteristics, combining multi-source real-time data to achieve precise monitoring and dynamic updates of the storage yard status. In the decision-making stage, a multi-agent game mechanism is employed, generating the optimal scheduling scheme through collaborative computation by virtual agents, and verifying it through simulation in a digital environment. Finally, the scheme is transformed into execution commands issued to equipment, establishing a closed-loop optimization mechanism based on real-time monitoring and safety thresholds. This innovative system realizes the transformation of storage management from experience-driven to data-driven, significantly improving storage yard utilization and operational efficiency. It effectively reduces operating costs through intelligent decision-making and process optimization, while establishing a closed-loop control mechanism throughout the entire process, comprehensively enhancing the system's adaptability and reliability.

[0042] Specifically, a digital twin model is obtained by constructing a real-world model of the coal yard through laser scanning and data integration, and setting the parameters of the coal yard operation equipment in the real-world model. The digital twin model is then trained to obtain an intelligent digital twin model, including: Raw data of the coal yard is collected by vehicle-mounted LiDAR, drones and GPS, and a real-world model of the coal yard is generated using the raw data. The engine is deployed and the data interface is set for the real-world model to obtain a digital twin model. Add physical constraint configurations and business rule configurations to the digital twin model to obtain a rule-based digital twin model; Historical operational data of coal yard equipment is integrated into a rule-based digital twin model. Historical operational features are extracted and used to train the rule-based digital twin model to obtain an intelligent digital twin model.

[0043] This solution systematically constructs an intelligent digital twin model for coal storage through three key steps. First, it integrates multi-source data from vehicle-mounted LiDAR, drones, and GPS to build a high-precision real-world model. An interactive digital twin model is then established through engine deployment, overcoming the problems of single data sets and lack of interactivity inherent in traditional modeling methods. Second, it configures equipment kinematic constraints and operational rules for the model, forming a rule-based digital twin model, overcoming the drawback of virtual models being disconnected from actual on-site conditions. Finally, by accessing historical operational data and extracting features for model training, the digital twin acquires the ability to learn autonomously and predictively optimize, overcoming the limitations of static models that lack intelligent decision-making capabilities. The key advantage of this series of steps is that it constructs a digital twin system that combines geometric accuracy, rule compliance, and intelligent predictive capabilities, providing a reliable digital foundation for subsequent precise perception and intelligent decision-making, and achieving a fundamental shift from traditional experience-based management to data-driven intelligent management.

Claims

1. A dynamic zoning management and intelligent scheduling system for coal storage, characterized in that, include: An initialization and data preparation unit is used to construct a real-world model of the coal yard through laser scanning and data integration, and to set the parameters of the coal yard operation equipment in the real-world model to obtain a digital twin model. The digital twin model is then trained using the historical operation data of the coal yard operation equipment to obtain an intelligent digital twin model. The situational awareness and digital synchronization unit is connected to the initialization and data preparation unit. The situational awareness and digital synchronization unit is used to collect multi-source data, create virtual cells in the intelligent digital twin model according to the properties of coal, storage requirements and operation plans, update the status data of different virtual cells in real time according to the multi-source data, obtain new tasks of the coal yard and create digital task work orders for each new task. The intelligent game and decision generation unit is connected to the situational awareness and digital synchronization unit. The intelligent game and decision generation unit is used to construct task intelligent agents, resource intelligent agents and equipment intelligent agents, and to obtain the optimal candidate solution by simulating the scheduling scheme based on the task intelligent agents, resource intelligent agents and equipment intelligent agents through the game coordinator. The optimal candidate solution is determined by simulating and deducing the optimal candidate solution using the intelligent digital twin model to obtain the optimal solution. The precise execution and dynamic optimization unit is connected to the intelligent game and decision generation unit. The precise execution and dynamic optimization unit is used to decompose the optimal candidate solution into control instructions and issue the control instructions, preset a safety threshold and compare the multi-source data with the preset safety threshold in real time to obtain the comparison result, and issue a corresponding scheduling request based on the comparison result.

2. The dynamic zoning management and intelligent scheduling system for coal storage according to claim 1, characterized in that, The initialization and data preparation unit includes: A digital twin model building module is used to collect raw data through vehicle-mounted lidar, drones and GPS and use the raw data to generate a real-scene model that is completely consistent with the physical geometry and physical properties of the coal yard in the same engineering coordinate system. The real-scene model is then used to deploy the engine and set the data interface to obtain the digital twin model. The rules and constraints configuration module is connected to the digital twin model construction module. The rules and constraints configuration module is used to add physical constraint configuration and business rule configuration to the digital twin model to obtain a rule-based digital twin model. The historical data learning module is connected to the rule and constraint configuration module. The historical data learning module is used to input the historical operation data into the rule digital twin model, perform data processing and feature extraction on the historical operation data to obtain historical operation features, and use the historical operation features to train the rule digital twin model to obtain the intelligent digital twin model.

3. The dynamic zoning management and intelligent scheduling system for coal storage according to claim 2, characterized in that, The digital twin model construction module includes: The data acquisition and processing submodule is used to acquire laser point cloud data of the coal yard using vehicle-mounted lidar, acquire multi-angle high-definition images of the coal yard using drones to generate a 3D real-scene model of the coal yard, use GPS to map the precise coordinates of the fixed facilities of the coal yard, obtain the 3D CAD model of the coal yard operation equipment, and integrate the laser point cloud data, the 3D real-scene model, the precise coordinates of the fixed facilities, and the 3D CAD model of the coal yard operation equipment into the real-scene model. The 3D engine deployment and integration submodule is connected to the data acquisition and processing submodule. The 3D engine deployment and integration submodule imports the real scene model into the project created in the 3D engine, assigns an identifier to each object in the real scene model, and creates an interface for each object to obtain the digital twin model so as to modify the attribute data of each object.

4. The dynamic zoning management and intelligent scheduling system for coal storage according to claim 2, characterized in that, The rules and constraints configuration module includes: The physical constraint configuration submodule is connected to the digital twin model construction module. The physical constraint configuration submodule is used to establish the kinematic model of the coal yard operation equipment and add a collision body model to the coal yard operation equipment, and input the performance parameters of the coal yard operation equipment. The business rule configuration submodule is connected to the physical constraint configuration submodule. The business rule configuration submodule is used to set the stacking rules, path rules and business processes in the digital twin model to obtain the rule digital twin model.

5. A dynamic zoning management and intelligent scheduling system for coal storage according to claim 2, characterized in that, The historical data learning module includes: The data governance and preprocessing submodule is connected to the rule and constraint configuration module. The data governance and preprocessing submodule is used to access the historical operation data into the rule digital twin model. The historical operation data includes historical operation work orders, monitoring data, metering data, test data and meteorological data of the coal yard operation equipment. After data cleaning and timestamp alignment of the historical operation data, the feature extraction is performed to obtain the historical operation features. The model pre-training submodule is connected to the data governance and preprocessing submodule. The model pre-training submodule is used to train the rule-based digital twin model using the historical job features to obtain the intelligent digital twin model.

6. The dynamic zoning management and intelligent scheduling system for coal storage according to claim 1, characterized in that, The situational awareness and digital synchronization unit includes: A multi-source data fusion module is connected to the initialization and data preparation unit. The multi-source data fusion module is used to fuse the UAV scanning data, the data collected by the sensors on the coal yard operation equipment, the GPS positioning data, and the business data of the coal yard operation equipment into the multi-source data and to connect the multi-source data to the intelligent digital twin model in real time. The virtual cell status update module is connected to the multi-source data fusion module. The virtual cell status update module is used to dynamically adjust the geometry, inventory, physical properties and chemical properties of the virtual cells in the intelligent digital twin model according to the multi-source data, and mark the idle status, occupied status, pending status and warning status of the virtual cells. The task instruction monitoring module is connected to the virtual cellular status update module. The task instruction monitoring module is used to capture the new tasks in the coal yard and create the digital task work order for each new task.

7. A dynamic zoning management and intelligent scheduling system for coal storage according to claim 1, characterized in that, The intelligent game and decision generation unit includes: A multi-agent market construction module is connected to the situational awareness and digital synchronization unit. The multi-agent market construction module is used to construct the task agent, the resource agent, and the device agent. The task agent represents each work task, the resource agent represents each virtual cell, and the device agent represents each coal yard operation device. A distributed collaborative game module is connected to the multi-agent market construction module. The distributed collaborative game module is used to send a price inquiry request to the device agent in the virtual market using the task agent, and to collect all bids and evaluate conflicts using the game coordinator to obtain the optimal candidate solution. The simulation and scheme confirmation module is connected to the distributed collaborative game module. The simulation and scheme confirmation module is used to perform the simulation of the optimal candidate scheme in the intelligent digital twin model, verify the feasibility of the optimal candidate scheme, and confirm that the optimal candidate scheme is the only executable optimal scheme.

8. A dynamic zoning management and intelligent scheduling system for coal storage according to claim 1, characterized in that, The intelligent game and decision generation unit includes: The instruction serialization and distribution module is connected to the intelligent game and decision generation unit. The instruction serialization and distribution module is used to decompose the optimal solution into control instructions that can be recognized by the coal yard operation equipment, and distribute the control instructions to the corresponding coal yard operation equipment through the industrial Internet of Things platform. The real-time monitoring and closed-loop feedback module is connected to the instruction serialization and issuance module. The real-time monitoring and closed-loop feedback module is used to continuously receive the actual execution status of the coal yard operation equipment and compare the actual execution status with the expected status when the intelligent digital twin model simulates the optimal solution to ensure that the optimal solution is implemented correctly. The anomaly detection and dynamic rescheduling module is connected to the real-time monitoring and closed-loop feedback module. The anomaly detection and dynamic rescheduling module is used to monitor abnormal events in real time and preset the safety threshold, compare the multi-source data with the preset safety threshold in real time to obtain the comparison result, and issue the corresponding scheduling request based on the comparison result.

9. A method for dynamic zoning management and intelligent scheduling of coal storage, characterized in that, The scheduling method employs the scheduling system according to any one of claims 1 to 8, and the scheduling method includes: The digital twin model is obtained by integrating the laser scan with the data to construct the real-scene model of the coal yard and setting the parameters of the coal yard operation equipment in the real-scene model. The digital twin model is then trained to obtain the intelligent digital twin model. Collect the multi-source data and create the virtual cell, update the status data of the virtual cell according to the multi-source data, obtain the new task and create the digital task work order for each new task; The task agent, resource agent, and device agent are constructed. The optimal candidate scheme is obtained by simulating the scheduling scheme through the game coordinator. The optimal candidate scheme is then simulated and deduced using the intelligent digital twin model. The feasibility of the optimal candidate scheme is determined to obtain the optimal scheme. The optimal solution is decomposed into the control command and the control command is issued and the safety threshold is preset. During the execution of the control command, the multi-source data is compared with the preset safety threshold in real time to obtain the comparison result. Based on the comparison result, the corresponding scheduling request is issued.

10. The method for dynamic zoning management and intelligent scheduling of coal storage according to claim 9, characterized in that, The process of constructing a real-world model of the coal yard by integrating the laser scan with the data, setting the parameters of the coal yard operating equipment in the real-world model to obtain the digital twin model, and training the digital twin model to obtain the intelligent digital twin model includes: The raw data of the coal yard is collected by the vehicle-mounted lidar, the drone and the GPS, and the raw data is used to generate the real scene model of the coal yard. The engine is deployed and the data interface is set on the real scene model to obtain the digital twin model. Add the physical constraint configuration and the business rule configuration to the digital twin model to obtain the rule-based digital twin model; The historical operation data of the coal yard operation equipment is input into the rule-based digital twin model, the historical operation features are extracted, and the rule-based digital twin model is trained using the historical operation features to obtain the intelligent digital twin model.

Images