Intelligent mine ore conveying management system

Abstract

The invention relates to the technical field of ore conveying, and discloses an intelligent mine ore conveying management system, which comprises a deep reinforcement learning scheduling module used for realizing optimal scheduling through a deep reinforcement learning algorithm based on an ore conveying equipment state, a conveying path condition and ore queue information; the digital twinning system is used for creating a digital twinning model of the conveying equipment and carrying out multi-physics field simulation, real-time data synchronization and state prediction so as to manage the full life cycle of the equipment; and the edge calculation module is used for carrying out data acquisition, real-time processing and lightweight AI reasoning on a mine site so as to realize equipment anomaly detection and early warning. By introducing deep reinforcement learning scheduling, digital twinning, edge computing and block chain data management technologies, adaptive optimization scheduling, accurate state monitoring and fault prediction of the ore conveying system are realized, an efficient safety protection and intelligent user interface is provided, and the operation efficiency, reliability and safety of the system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ore transportation, and in particular to an intelligent mine ore transportation management system. Background Art

[0002] The mine ore transportation management system is an important part of the mine production process. It is responsible for the efficient transportation of ore from the mining site to the processing or storage site. Traditional ore transportation mainly relies on belt conveyors, truck transportation and other methods. The transportation process is optimized through the dispatching and path management of transportation equipment. In modern mine operations, the transportation management system not only needs to efficiently complete the transfer of ore, but also needs to take into account the comprehensive optimization of energy consumption, maintenance, safety and other aspects to improve the overall operation efficiency of the mine;

[0003] Existing mine ore transportation management systems usually adopt rule-driven or fixed parameter-based scheduling methods to achieve automatic transportation of ore by setting equipment operating parameters and path selection rules. Such systems rely on preset control logic and data monitoring, and can perform basic monitoring and control of the operating status of transportation equipment, such as detecting equipment failures, adjusting transportation speed and path, etc. At the same time, some advanced systems have introduced simple data analysis and early warning functions in order to optimize equipment maintenance and energy consumption control.

[0004] However, the existing mine ore transportation management systems still have some shortcomings in practical applications. First, these systems mostly rely on static rules and fixed parameters, lack adaptive scheduling capabilities, and are difficult to cope with dynamic changes in mining environment and transportation needs. Secondly, the accuracy of equipment status monitoring and fault prediction is not high, and accurate predictive maintenance cannot be achieved, resulting in high maintenance costs and insufficient equipment reliability. In addition, the data security and network protection capabilities of traditional systems are limited, and they cannot effectively prevent network attacks and data tampering risks. Finally, the existing user interfaces and control platforms lack intelligence and visualization, and the operation is complex and non-intuitive, affecting management efficiency and decision-making effectiveness. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an intelligent mine ore transportation management system. By introducing deep reinforcement learning scheduling, digital twins, edge computing and blockchain data management technology, the present invention realizes adaptive optimization scheduling, precise status monitoring and fault prediction of the ore transportation system, and provides efficient safety protection and intelligent user interface, thereby improving the operating efficiency, reliability and safety of the system, and solving the problems of traditional systems being unable to be dynamically adjusted, high maintenance costs, poor data security and complex operations.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent mine ore transportation management system, comprising:

[0007] Deep reinforcement learning scheduling module, which is used to achieve optimized scheduling through deep reinforcement learning algorithm based on the status of ore conveying equipment, conveying path conditions and ore queue information;

[0008] Digital twin system, used to create digital twin models of conveying equipment, perform multi-physics simulation, real-time data synchronization and status prediction to manage the entire life cycle of the equipment;

[0009] Edge computing module, used for data collection, real-time processing and lightweight AI reasoning at the mine site, to achieve equipment anomaly detection and early warning;

[0010] Blockchain data management module, used to record and store equipment operation logs and maintenance records, and ensure data security and immutability through blockchain;

[0011] The network security protection module is used to monitor the mine network security in real time through the AI-driven network detection and response system to protect against potential network attacks and data leaks;

[0012] User interface and control platform for operators to monitor and control the ore conveying system in real time, providing status visualization, alarm management and operation interface.

[0013] Preferably, the deep reinforcement learning scheduling module includes:

[0014] State space modeling: by converting equipment status, path status and ore queue information into multi-dimensional vectors and inputting them into the deep reinforcement learning model;

[0015] Reward function construction: Establish a multi-objective optimization reward function based on delivery time, energy consumption, maintenance cost and safety;

[0016] Strategy and value network training: PPO algorithm is used for strategy optimization, and training is performed through the loss function consisting of strategy ratio, value function loss and strategy entropy;

[0017] Strategy execution and optimization: Use distributed architecture to deploy scheduling strategies, and adjust and optimize strategies through real-time feedback data.

[0018] Preferably, the digital twin system includes:

[0019] Simulation model establishment: Use multi-physics field coupling method to build dynamic, thermal and electrical characteristic models of conveying equipment;

[0020] Real-time data synchronization: Synchronize sensor data with the digital twin in real time through the OPC UA protocol to ensure that the model reflects the real-time status of the equipment;

[0021] State estimation and updating: Use Kalman filter to perform state estimation and error correction on real-time data;

[0022] Full life cycle management: Analyze the equipment operation status through the Markov chain model, and combine the LSTM prediction model to perform fault prediction and maintenance optimization;

[0023] Redundancy design and fault management: Design equipment redundancy systems to automatically switch to backup equipment or systems when equipment or sensors fail to ensure continuous operation of the system.

[0024] Preferably, the edge computing module includes:

[0025] Edge hardware configuration: Use NVIDIA Jetson Nano for edge computing and deploy the EdgeX Foundry platform for data management;

[0026] Data collection and preprocessing: Collect sensor data through the device service layer and perform standardization and denoising;

[0027] Lightweight AI model deployment: Use TensorFlow Lite to deploy quantized convolutional neural networks for real-time anomaly detection;

[0028] Real-time reasoning and feedback: Use batch processing and parallel computing technology to achieve efficient real-time reasoning and feedback to the central control system;

[0029] Benefit Analysis: Integrated benefit analysis module to evaluate equipment performance and system efficiency in real time, including response time, energy consumption and maintenance cost savings.

[0030] Preferably, the blockchain data management module includes:

[0031] Build a blockchain network: Build a blockchain network based on Hyperledger Fabric, with nodes including device controllers and management servers;

[0032] Smart contract implementation: Deploy smart contracts to record and verify device operation logs and maintenance records, and use the PBFT protocol to reach consensus;

[0033] Data storage and verification: Hash calculation is performed on each operation and stored in the blockchain to ensure data integrity and immutability;

[0034] Multi-copy storage and permission management: Implement multi-copy data storage and access permission control to ensure the system's fault tolerance and data security;

[0035] Data tracing and analysis: Through the chain structure of blockchain, historical operation data can be traced and analyzed to provide data support for equipment maintenance decisions.

[0036] Preferably, the network security protection module includes:

[0037] Real-time network monitoring: Use AI-based network detection systems to monitor data traffic, identify and warn of abnormal traffic;

[0038] Anomaly detection model: The autoencoder model is used to perform unsupervised learning on normal traffic and identify anomalies based on reconstruction errors.

[0039] Dynamic security threshold adjustment: Based on the real-time network status, the threshold of anomaly detection is adaptively adjusted to improve the detection accuracy and response speed;

[0040] Security response mechanism: When abnormal traffic is detected, isolation, alarm and protection operations are automatically triggered;

[0041] Redundant path design: Configure redundant network paths to ensure automatic switching to backup paths in the event of network failure or attack, thus ensuring system continuity.

[0042] Preferably, the user interface and control platform include:

[0043] Data visualization: Provides real-time visualization of conveyor equipment status, path conditions, and ore queue information;

[0044] Scheduling control: Control the scheduling module through the interface and adjust the transportation strategy and parameters;

[0045] Alarm management: Real-time display of system abnormality and fault information, and provision of alarm management and emergency operation functions;

[0046] Performance analysis and reporting: Provides system performance analysis and benefit reporting interface, showing indicators of response time, energy savings and maintenance costs;

[0047] User authority configuration: Configure access rights for different user levels to ensure system operation security and data confidentiality.

[0048] Preferably, the reward function construction includes:

[0049] Time efficiency calculation: The time efficiency bonus is calculated by the ratio of actual delivery time to target delivery time;

[0050] Energy consumption assessment: Calculate energy consumption rewards based on the ratio of current energy consumption to target energy consumption to encourage energy efficiency optimization;

[0051] Maintenance cost assessment: Calculate the ratio of current maintenance cost to preset maintenance budget to optimize maintenance plan;

[0052] Safety assessment: Positive rewards are given when equipment operation meets safety standards;

[0053] Queue management: monitor and control the length of the ore queue to avoid reduced transportation efficiency due to excessively long queues.

[0054] Preferably, the real-time data synchronization includes:

[0055] Data filtering: Use filtering algorithms to remove noise and outliers in data and improve data quality;

[0056] Timestamp correction: Ensure the synchronization of sensor data and digital twins through timestamp correction;

[0057] Predictive status update: status prediction based on short-term historical data to optimize and adjust equipment parameters in real time;

[0058] Two-way feedback mechanism: Establish a two-way data feedback mechanism between the digital twin and the actual device to ensure that the model is consistent with the device status;

[0059] Fault warning and response: Use predictive models to provide early warning of potential faults and trigger corresponding maintenance responses.

[0060] Preferably, the lightweight AI model deployment includes:

[0061] Data preprocessing: standardize the collected sensor data to ensure the stability and accuracy of the model input;

[0062] Model compression and optimization: reduce model size and computational complexity through pruning and quantization techniques;

[0063] Real-time inference optimization: Use batch processing and parallel computing technology to achieve efficient real-time inference on edge devices;

[0064] System integration and control: Integrate AI reasoning results with system control modules to achieve automated decision-making;

[0065] Benefit analysis: Real-time evaluation of the contribution of AI models to system benefits, including equipment response time, energy consumption optimization, and maintenance cost control.

[0066] The present invention provides an intelligent mine ore transportation management system, which has the following beneficial effects:

[0067] 1. The present invention realizes the intelligence and efficiency of ore transportation management by introducing deep reinforcement learning scheduling, digital twin system and edge computing module. At the same time, it can optimize the scheduling strategy in real time, accurately predict the equipment status, and quickly respond to abnormalities, significantly improve the ore transportation efficiency, reduce energy consumption and maintenance costs, and ensure the safe operation of the equipment.

[0068] 2. The present invention realizes the tamper-proof storage and real-time network security monitoring of equipment operation data by adopting blockchain data management module and network security protection module, and ensures data integrity and security through smart contracts and multi-copy storage. Combined with AI-driven anomaly detection and automatic protection mechanism, it effectively improves the data security and operational stability of the mine transportation system.

[0069] 3. The present invention combines data visualization, automatic control and benefit analysis functions through the user interface and control platform to provide operators with intuitive status monitoring, flexible scheduling control and accurate benefit evaluation. It also improves the transparency and operational efficiency of mine management through real-time data-driven decision-making and control optimization, and supports the continuous improvement and development of smart mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a system structure diagram of the present invention. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] Please see attached Figure 1 The embodiment of the present invention provides an intelligent mine ore transportation management system, including:

[0073] Deep reinforcement learning scheduling module, which is used to optimize scheduling through deep reinforcement learning algorithm based on the status of ore conveying equipment, conveying path conditions and ore queue information. Through the introduction of deep reinforcement learning scheduling module, the system can adaptively adjust the conveying strategy, maximize equipment utilization, reduce energy consumption and maintenance costs, and improve the overall efficiency and safety of ore conveying;

[0074] Digital twin system, which is used to create digital twin models of conveying equipment, conduct multi-physics field simulation, real-time data synchronization and status prediction to manage the entire life cycle of equipment. The digital twin system not only improves the monitoring accuracy of equipment status, but also can provide early fault warning, optimize equipment maintenance strategies, reduce unplanned downtime, extend equipment life, and improve the reliability and operation efficiency of the overall system;

[0075] Edge computing module, which is used for data collection, real-time processing and lightweight AI reasoning at the mine site to realize equipment anomaly detection and early warning. Through the edge computing module, the system can respond quickly at the point where data is generated, improving the real-time and accuracy of anomaly detection, reducing data processing delays and network transmission costs, thereby improving the intelligence and efficiency of mine operations;

[0076] Blockchain data management module, used to record and store equipment operation logs and maintenance records, and ensure the security and immutability of data through blockchain. This module provides a safe and transparent solution for the operation records and maintenance management of mining equipment, ensuring that data is not tampered with, and improving the credibility of equipment management and the accuracy of decision-making;

[0077] The network security protection module is used to monitor the mine network security in real time through the AI-driven network detection and response system to protect against potential network attacks and data leaks. The network security protection module improves the security of the mine network, prevents potential network attacks and data leaks, and ensures the stable operation of the system and the security of data transmission;

[0078] The user interface and control platform are used for operators to monitor and control the ore conveying system in real time, and provide status visualization, alarm management and operation interface. The user interface and control platform integrate system monitoring, control, analysis and management functions to provide operators with all-round support and convenience, thereby improving the operator's control ability over the system and transparency of the equipment's operating status, optimizing the management and maintenance of the ore conveying process, reducing human errors, and improving the overall operating efficiency and safety of the system.

[0079] Please see attached Figure 1 In a preferred embodiment of the present invention, the deep reinforcement learning scheduling module includes:

[0080] State space modeling: By converting the equipment status, path status and ore queue information into multi-dimensional vectors and inputting them into the deep reinforcement learning model, data is collected on the operating status of the ore conveying equipment (such as current speed, temperature, vibration, etc.), the status of the conveying path (such as path load, congestion, wear degree) and the information of the ore queue (such as ore type, quantity, priority, etc.), and these information are integrated into a multi-dimensional vector form. State space modeling comprehensively captures and models the key factors of the ore conveying system, providing an accurate description of the environment for the deep reinforcement learning algorithm, helping the system to accurately reflect the dynamic changes in the ore conveying process, and providing high-quality input data for the scheduling algorithm, thereby improving the accuracy and adaptability of the scheduling strategy and achieving more efficient ore conveying scheduling;

[0081] Reward function construction: Establish a multi-objective optimization reward function based on delivery time, energy consumption, maintenance cost and safety. During the reward function construction phase, the system integrates multiple optimization objectives such as delivery time, energy consumption, maintenance cost and safety into a comprehensive reward function. The constructed reward function takes the following form:

[0082]

[0083] Where: T actual and T optimal E represents the actual delivery time and the optimal delivery time respectively; consumed and E t arget is the actual and target energy consumption; C maint is the current maintenance cost, C m ax is the maximum allowable maintenance cost; I safety is a safety indicator. When the operation meets the safety standards, rewards are given. Parameters α, β, γ, and δ are used to weigh the importance of each optimization goal. This reward function integrates multiple optimization goals together to guide the scheduling strategy toward high transportation efficiency, low energy consumption, low maintenance cost, and high safety. It can find the optimal balance between different operating goals and achieve comprehensive scheduling optimization, which not only improves the efficiency of ore transportation, but also reduces the system operation cost and risk, and improves the overall efficiency and safety of the system.

[0084] Strategy and value network training: The PPO algorithm is used for strategy optimization. Training is performed through the loss function composed of strategy ratio, value function loss and strategy entropy. This can not only improve the efficiency of ore transportation, but also reduce the system operation cost and risk, and improve the overall efficiency and safety of the system. During training, the loss function is used:

[0085]

[0086] in: is the strategy ratio; is the advantage function, which is used to measure the superiority of the current action relative to the average strategy; ∈ is the clipping range, which prevents the strategy update from being too large, resulting in unstable training. The PPO algorithm avoids the instability in the policy gradient method by limiting the strategy update range, while maintaining a high training efficiency. The joint training of the strategy network and the value network enables the system to continuously optimize the scheduling strategy in a complex environment and quickly adapt to environmental changes, thereby ensuring that the scheduling strategy is continuously optimized in a complex and changeable mining environment, improving the accuracy and robustness of scheduling decisions, and improving the intelligence level and operation efficiency of the ore transportation system;

[0087] Strategy execution and optimization: Distributed architecture is used to deploy scheduling strategies, and real-time feedback data is used to adjust and optimize strategies. During the strategy execution and optimization stage, the system deploys the trained scheduling strategies in the distributed architecture for execution in actual mining operations. The distributed architecture supports multi-node parallel computing to ensure that scheduling strategies can quickly respond to changes in field data. Strategy execution and optimization achieve dynamic scheduling through closed-loop control. Even when the mining environment is constantly changing, the system can automatically adjust the scheduling strategy to ensure the optimal transportation path and equipment utilization, thereby improving ore transportation efficiency, reducing operating costs, and effectively responding to the uncertainty and complexity in mining operations.

[0088] Please see attached Figure 1 In a preferred embodiment of the present invention, the digital twin system includes:

[0089] Simulation model establishment: The multi-physics field coupling method is used to construct the dynamic, thermal and electrical characteristic models of the conveying equipment. The system first uses CAD modeling and finite element analysis (FEA) to model the physical characteristics of the equipment in detail, such as the stiffness, vibration mode and heat conduction path of the conveying equipment. The dynamic simulation simulates the motion characteristics of the conveying equipment, including belt tension, speed change and load distribution. The thermal model is used to predict the temperature distribution of the equipment under different loads and environmental conditions to ensure that the equipment will not fail due to overheating during operation. The electrical characteristic model focuses on the working status of the motor and control system, such as current, voltage fluctuations and power consumption. Through the establishment of the simulation model, the digital twin system can accurately simulate and predict the operation of the equipment in a virtual environment, providing data support for the optimal design and operation decisions of the equipment, effectively reducing the actual trial and error costs, and improving the design and operation efficiency of the equipment;

[0090] Real-time data synchronization: The sensor data is synchronized with the digital twin in real time through the OPC UA protocol to ensure that the model reflects the real-time status of the equipment. The OPC UA protocol provides a standardized data exchange method that can support data transmission and interoperability between multiple devices. Through real-time data synchronization, the digital twin can obtain the latest operating data of the equipment in a timely manner, such as vibration, temperature, speed and other parameters, and update the status in the simulation model. Real-time data synchronization ensures the synchronous operation of the digital twin and the actual equipment, so that the digital twin system can reflect the real-time status and operation trend of the equipment, enhance the accuracy of equipment status monitoring, and provide a reliable data foundation for predictive maintenance and intelligent decision-making of equipment;

[0091] State estimation and update: Use Kalman filter to perform state estimation and error correction on real-time data. Kalman filter can effectively filter out noise and outliers in measurement data by iterating between state prediction and observation update, improve the accuracy of state estimation, and improve the prediction accuracy and response speed of digital twins to equipment status, so that the system can quickly detect anomalies and deviations in equipment operation, and provide timely and reliable support for safe operation and maintenance of equipment;

[0092] Full life cycle management: The Markov chain model is used to analyze the equipment operation status, and the LSTM prediction model is used to predict faults and optimize maintenance. The Markov chain model uses the equipment's historical operation data to estimate the transition probability between different states, thereby predicting the equipment's future state distribution. The LSTM model uses a deep learning algorithm to train the equipment's historical data, capturing the long-term and short-term dependencies of the equipment's operation status over time, and accurately predicting the equipment's future state and potential failure risks. Through dynamic analysis and prediction of the equipment's operation status, the equipment's maintenance plan is optimized, the number of unplanned downtimes is reduced, the equipment's availability and operating efficiency are improved, the equipment's service life is extended, and the operating and maintenance costs are reduced;

[0093] Redundant design and fault management: Design equipment redundant system to automatically switch to backup equipment or system when equipment or sensor fails to ensure continuous operation of the system. Redundant design uses backup equipment and redundant sensors to form a multi-level protection mechanism to enhance the system's recovery capability when equipment or sensor fails. The fault management module can quickly identify faults and perform corresponding switching and recovery operations by real-time monitoring of the operating status of equipment and sensors, reducing the impact of faults on system operation, thereby improving system reliability and robustness, ensuring that when equipment or sensor fails, the system can quickly resume operation, reducing downtime and production losses, and ensuring the continuity and efficiency of the ore transportation management system.

[0094] Please see attached Figure 1 In a preferred embodiment of the present invention, the edge computing module includes:

[0095] Edge hardware configuration: Use NVIDIA Jetson Nano for edge computing and deploy EdgeX Foundry platform for data management. This module deploys EdgeX Foundry platform as a data management and service integration framework to support data access from multiple sensors and devices. EdgeX Foundry is an open and modular edge computing platform that provides device service layer, middleware and application service layer, and can effectively manage the complete process from data acquisition, processing to reasoning. Jetson Nano provides powerful parallel computing capabilities with its embedded GPU, which can accelerate the reasoning process of AI models. EdgeX Foundry integrates data from various sensors and devices through a loosely coupled microservice architecture, and provides data management, rule engine and edge analysis functions, providing a flexible and efficient solution for on-site data processing in smart mines, thereby reducing data transmission delays and dependence on cloud computing resources, and enhancing the system's response speed and autonomous decision-making capabilities.

[0096] Data collection and preprocessing: Sensor data is collected through the device service layer and standardized and denoised. Denoising improves the quality and credibility of data by filtering out irrelevant noise and abnormal data, ensuring the accuracy of subsequent AI reasoning. This not only improves the accuracy of the system's anomaly detection, but also reduces misjudgments and delays caused by data problems, and improves the reliability and real-time nature of on-site decision-making in mines.

[0097] Lightweight AI model deployment: Use TensorFlow Lite to deploy quantized convolutional neural networks for real-time anomaly detection. TensorFlow Lite provides a lightweight inference engine that is optimized for edge devices and can achieve efficient model inference with limited computing resources. Through quantization, the size and computational complexity of the model are greatly reduced, allowing the convolutional neural network to run smoothly on edge computing devices such as Jetson Nano, achieving real-time anomaly detection. By quickly responding to abnormal situations, the system can take corrective measures in a timely manner to avoid downtime and losses caused by equipment failures.

[0098] Real-time reasoning and feedback: Batch processing and parallel computing technologies are used to achieve efficient real-time reasoning and feedback to the central control system. The system uses batch processing and parallel computing technologies to quickly input the collected pre-processed data into the AI ​​model for reasoning, and the reasoning results are immediately transmitted to the central control system to provide real-time decision support for mine management. In addition, the system will automatically generate feedback reports based on the reasoning results and trigger corresponding alarms and processing measures to ensure the safe operation of mining equipment. Parallel computing uses the multi-core architecture of Jetson Nano to further accelerate the data processing and model reasoning process. By feeding back the reasoning results to the central control system in real time, the system can respond quickly when an abnormality occurs, reducing risks and losses, so that the system can respond at the moment of abnormality, provide effective decision support, improve the continuity and reliability of mine operations, and reduce unplanned downtime and production interruptions caused by equipment abnormalities;

[0099] Benefit analysis: The integrated benefit analysis module evaluates equipment performance and system efficiency in real time, including response time, energy consumption and maintenance cost savings. The system uses big data analysis technology to process historical data and identify potential optimization opportunities and efficiency improvement points. Through regular benefit reports, management can see the system's benefit change trends more clearly and adjust strategies and optimization measures in a timely manner. Through accurate benefit evaluation and trend analysis, the system can discover weak links and optimization space in equipment operation, continuously improve equipment performance, reduce operating costs, and improve the overall benefits of mining operations.

[0100] Please see attached Figure 1 In a preferred embodiment of the present invention, the blockchain data management module includes:

[0101] Build a blockchain network: Build a blockchain network based on Hyperledger Fabric. The nodes include device controllers and management servers. Hyperledger Fabric is an open source enterprise-level blockchain platform with a modular architecture and flexible permission management mechanism. The blockchain network of Hyperledger Fabric uses distributed ledger technology to achieve data sharing and collaboration among multiple nodes. Each node saves a complete copy of the ledger. When new device operation data is written to the ledger, each node confirms and synchronizes it through a consensus protocol to ensure the consistency and integrity of all node ledgers. This enables safe and reliable data sharing and collaboration between the mine's device controllers and management servers, and enhances the transparency and credibility of the data, thereby effectively preventing data tampering and loss, and providing a solid foundation for the operation and management of mining equipment.

[0102] Smart contract implementation: Smart contracts are deployed to record and verify equipment operation logs and maintenance records, and the PBFT protocol is used to reach a consensus. Smart contracts automatically process equipment operation and maintenance events through predefined logic, eliminating the possibility of manual intervention and reducing human errors. The PBFT protocol is an efficient consensus algorithm suitable for enterprise-level blockchain applications. It can ensure the normal operation and data consensus of the blockchain network when some nodes have untrustworthy behaviors, so that the blockchain network can efficiently record equipment operation logs and maintenance records, ensuring the real-time and integrity of the data. At the same time, it improves the system's anti-interference ability and data consistency in complex mining environments, reduces errors and omissions in operation data, and provides reliable data support for equipment maintenance decisions;

[0103] Data storage and verification: Each operation is hashed and stored in the blockchain to ensure data integrity and immutability. Hash calculation converts operation data into a unique digital fingerprint, so that even a small change in data will cause a change in the hash value, thereby quickly detecting data tampering and ensuring data immutability. Any modification of historical data will destroy the integrity of the entire chain;

[0104] Multi-copy storage and permission management: Implement multi-copy data storage and access permission control to ensure the system's fault tolerance and data security. Multi-copy storage uses the distributed characteristics of blockchain to save a complete copy of the ledger at each node to achieve redundant backup of data, improve the system's fault tolerance and data recovery capabilities, thereby ensuring data security and system operational compliance. At the same time, multi-copy storage enhances the system's reliability and risk resistance. Even if some nodes fail, the system can ensure data integrity and availability;

[0105] Data tracing and analysis: Through the chain structure of blockchain, historical operation data can be traced and analyzed to provide data support for equipment maintenance decisions.

[0106] Please see attached Figure 1 In a preferred embodiment of the present invention, the network security protection module includes:

[0107] Real-time network monitoring: Use an AI-based network detection system to monitor data traffic, identify and warn of abnormal traffic. The system deploys traffic monitoring sensors at key network nodes, obtains data flow information at the network layer and application layer through deep packet inspection (DPI) technology, and uses AI algorithms to analyze data flows to identify common network behavior characteristics and traffic patterns, such as data packet size, transmission frequency, and communication protocols. Once abnormal traffic characteristics are found, the system will immediately generate an alarm and record abnormal events, thereby ensuring the visibility and controllability of mine network security, and being able to identify and warn in a timely manner at the early stages of an attack, reducing the potential threat of network attacks to mine intelligent conveying systems and improving the security and stability of the system;

[0108] Anomaly detection model: The autoencoder model is used to perform unsupervised learning on normal traffic and identify anomalies based on reconstruction errors. The system collects normal network traffic data and uses the autoencoder for model training. The autoencoder consists of an encoder and a decoder. The encoder compresses the input data into a low-dimensional feature space, while the decoder attempts to reconstruct the original data. After training, the autoencoder can accurately reconstruct normal traffic data, but when encountering abnormal traffic, the reconstruction error will increase. By detecting anomalies through reconstruction errors, the system can effectively respond to various unknown or new network threats and avoid the limitations of traditional rules and signature detection methods. Through unsupervised learning, the model can adaptively update normal traffic patterns, maintain high sensitivity to dynamically changing network environments, and improve overall detection accuracy and response speed;

[0109] Dynamic safety threshold adjustment: Based on the real-time network status, the threshold of anomaly detection is adaptively adjusted to improve the detection accuracy and response speed. The system continuously evaluates the current network traffic pattern and the reconstruction error distribution of the autoencoder, and dynamically adjusts the detection threshold according to the error change trend to avoid false positives and false negatives. Dynamic safety threshold adjustment improves the accuracy and response speed of anomaly detection, reduces false positives and false negatives caused by fixed thresholds, and enables the system to run stably and reliably under different network loads and environmental conditions, enhancing the security protection capabilities of the mine network.

[0110] Security response mechanism: When abnormal traffic is detected, isolation, alarm and protection operations are automatically triggered. The system first quickly isolates the abnormal traffic to prevent it from spreading further and affecting other parts of the system. Then, the system sends an alarm message to the network administrator through the interface or SMS, email, etc., detailing the nature and scope of the abnormal event. The security response mechanism shortens the response time from threat detection to taking protective measures, reducing the losses that may be caused by network attacks. At the same time, the system's rapid isolation and alarm functions improve the efficiency of security incident processing, providing stronger protection for mine network security management;

[0111] Redundant path design: Configure redundant network paths to ensure automatic switching to backup paths in the event of network failure or attack to ensure system continuity. The system uses network path monitoring and traffic distribution mechanisms to evaluate the status and load of each path in real time and select the optimal transmission path. The redundant path design also combines load balancing technology to disperse traffic under normal circumstances and improve the transmission efficiency and stability of the overall network. The system uses real-time monitoring technology to determine the health status of the main path. When a path failure or abnormality is detected, it quickly switches to the backup path to avoid data interruption. The redundant path design effectively improves the reliability and risk resistance of the mine network, ensures the continuity and stability of data transmission, and allows the system to maintain normal operation even in the event of a network failure or attack.

[0112] Please see attached Figure 1 In a preferred embodiment of the present invention, the user interface and control platform include:

[0113] Data visualization: Provides real-time visualization of the status of conveying equipment, path conditions, and ore queue information. Data visualization helps users quickly understand the operating status of the system by converting complex operating data into easy-to-understand visual elements. Visualization technology combines real-time data flow and graphics rendering algorithms to achieve dynamic update and accurate display of data. Through real-time status display, users can promptly discover potential problems, thereby improving the management level and operational efficiency of the ore conveying system.

[0114] Scheduling control: The scheduling module can be controlled through the interface to adjust the transportation strategy and parameters. The scheduling control function allows users to directly intervene and adjust the transportation strategy and related parameters through the interface. Users can adjust the equipment operation mode, priority setting, path selection and other scheduling parameters through the control panel on the interface, and control the ore transportation process in real time. Through direct operation of the user interface, the system can quickly perform scheduling adjustments, optimize the transportation path and equipment utilization, and improve the efficiency and safety of mine transportation;

[0115] Alarm management: Real-time display of system abnormality and fault information, and provides alarm management and emergency operation functions. The system receives abnormal events and fault alarms from various modules (such as network security protection module, edge computing module). The alarm information will be displayed in the form of red highlight, pop-up window, etc. on the interface to attract user attention. Users can view the detailed information, occurrence time and impact range of the alarm through the interface, and perform corresponding measures according to the preset emergency operation guide, such as remotely restarting the device, adjusting operating parameters or notifying on-site personnel to handle. The alarm management function improves the system's response speed and processing efficiency to abnormal events. Through real-time alarm prompts and emergency operation guidance, users can take measures quickly to reduce losses caused by equipment failures or abnormal operations;

[0116] Performance analysis and reporting: Provides a system performance analysis and benefit report interface to display indicators of response time, energy savings, and maintenance costs. The performance analysis module uses big data analysis and machine learning algorithms to process and analyze system operation data. The system compares and analyzes real-time data and historical data based on preset performance indicators to identify performance bottlenecks and improvement opportunities. Through detailed benefit reports, users can quantify the economic benefits brought about by equipment and strategy adjustments, support decision optimization and continuous improvement, and improve the overall efficiency and profitability of the mine conveying system;

[0117] User authority configuration: configure access rights for different user levels to ensure system operation security and data confidentiality. User authority configuration is implemented through the role-based access control (RBAC) model to strictly control user access behavior. The user authority configuration function improves the operational security and data protection level of the mine transportation management system, prevents unauthorized access and operation, and ensures the confidentiality and integrity of sensitive information. At the same time, through strict authority management and audit mechanisms, the system can quickly identify and respond to abnormal operations to ensure the safe operation of the mine transportation system.

[0118] Please see attached Figure 1 In a preferred embodiment of the present invention, the reward function construction includes:

[0119] Time efficiency calculation: The time efficiency bonus is calculated by the ratio of the actual delivery time to the target delivery time. The time efficiency bonus is calculated by the following formula:

[0120]

[0121] Among them, α is the weight coefficient, which is used to adjust the impact of time efficiency in the overall reward.

[0122] Through this formula, the system can motivate the scheduling strategy to minimize the gap between the actual transportation time and the target time. This time efficiency calculation method quantifies the time performance of the transportation task by comparing the actual and target times. The reward function takes shortening the actual transportation time as the goal to encourage the scheduling system to optimize the use of equipment and path selection, thereby improving the overall transportation efficiency and improving the response speed and operating efficiency of the ore transportation system. By reducing the transportation time, the system can speed up the turnover and processing speed of the ore, optimize the utilization rate of the equipment, and improve the overall operating efficiency of the mine;

[0123] Energy consumption assessment: Energy consumption rewards are calculated based on the ratio of current energy consumption to target energy consumption to encourage energy efficiency optimization. Energy consumption rewards are calculated using the following formula:

[0124]

[0125] Among them, β is the weight coefficient used to adjust the impact of energy consumption on rewards.

[0126] This formula motivates the dispatching system to reduce energy consumption and control the actual energy consumption within the target energy consumption range or even lower, thereby achieving energy-saving optimization. The energy consumption evaluation function effectively reduces the operating cost of the ore transportation system. By optimizing energy consumption management, the system can reduce electricity consumption and energy costs, improve the economic benefits of the mine, and also help reduce carbon emissions and support green mine construction.

[0127] Maintenance cost assessment: Calculate the ratio of current maintenance cost to preset maintenance budget to optimize maintenance plan. Maintenance cost reward is calculated by the following formula:

[0128]

[0129] Among them, γ is the weight coefficient, which adjusts the impact of maintenance cost on reward.

[0130] Through this formula, the system can encourage scheduling strategies to meet transportation needs while keeping maintenance costs within budget. The maintenance cost assessment function helps mines reduce unplanned downtime and high maintenance costs. Through intelligent scheduling and maintenance optimization, the system can extend equipment life, reduce maintenance frequency and costs, and achieve economical and efficient management of equipment maintenance.

[0131] Safety assessment: positive rewards are given when equipment operations meet safety standards. Safety assessment continuously monitors the operating status of the equipment to ensure that all operations are performed within a safe range. Positive rewards encourage the system to prioritize safer operating modes and reduce risks in equipment operations. The safety assessment function improves the safety management level of the ore conveying system. Through positive incentives for safe operations, the system can reduce accidents and injuries, ensuring the safety of mine operations and the lives of employees.

[0132] Queue management: monitor and control the length of the ore queue to avoid the reduction of transportation efficiency due to the long queue. Queue management is used to monitor and control the length of the ore queue to avoid the reduction of transportation efficiency due to the long queue. The queue reward is calculated by the following formula:

[0133]

[0134] Among them, ∈ is the weight coefficient of the influence of adjusting the queue length on the reward. This formula encourages the scheduling system to keep the ore queue within the optimal length range, thereby avoiding waste of resources caused by queue backlog or shortening. The queue management function improves the scheduling efficiency of the ore conveying system. By reasonably controlling the length of the ore queue, the system can optimize equipment utilization and conveying path selection, avoid efficiency decline caused by queue problems, and achieve smooth and efficient operation of the conveying operation.

[0135] Please see attached Figure 1 In a preferred embodiment of the present invention, real-time data synchronization includes:

[0136] Data filtering: Use filtering algorithms to remove noise and outliers in data and improve data quality. The Kalman filter algorithm is an optimal estimator based on the state space model. By combining the current measurement value and the predicted state, the impact of noise is minimized to achieve accurate estimation of the real device state. Through data filtering, the system can effectively remove random noise and outliers in sensor data, ensuring that the real-time data transmitted to the digital twin is of high quality and reliability, thereby enhancing the accuracy of equipment status monitoring and providing a solid data foundation for subsequent status prediction and fault warning.

[0137] Timestamp correction: Timestamp correction is used to ensure the synchronization between sensor data and digital twins. Since sensor data may come from different devices and time sources, there may be delays and time deviations. The system uses the timestamp correction mechanism to adjust the unified time base of all sensor data. Time synchronization protocols such as NTP and PTP provide high-precision time alignment functions, so that each sensor data can be seamlessly integrated into the digital twin model and maintain the synchronization between the model and the actual device operation status. This function improves the accuracy of the digital twin's reflection of the actual status of the device, making the linkage between the simulation model and the real device closer and more reliable;

[0138] Predictive status update: Based on short-term historical data, status prediction is performed to optimize the real-time adjustment of equipment parameters. The system uses time series prediction models, such as ARIMA (autoregressive integrated moving average model) or LSTM (long short-term memory network), to predict the status parameters of the equipment. The model uses the equipment operation data of the past period of time to predict the state changes at the next moment or in the short term, such as temperature, vibration or load changes. According to the prediction results, the system can adjust the equipment's operating parameters in advance, such as adjusting the speed or load, to avoid possible abnormalities or overloads. Through accurate prediction of future states, the system can adjust equipment parameters in advance to avoid potential abnormalities and overloads, improve equipment stability and operating efficiency, and extend equipment life.

[0139] Bidirectional feedback mechanism: Establish a bidirectional data feedback mechanism between the digital twin and the actual equipment to ensure that the model is consistent with the equipment status. Through the edge computing node or industrial Internet of Things gateway, the simulation results of the digital twin and the status data of the actual equipment are transmitted and compared with each other. The bidirectional feedback mechanism can timely feedback the predictions and suggestions of the digital twin to the equipment control system, and at the same time transmit the actual operation data of the equipment back to the digital twin for model correction and update. The bidirectional feedback mechanism forms a closed-loop control system through real-time status comparison and data exchange, thereby enhancing the control and response speed of the digital twin over the operation of the equipment. Through real-time data interaction, the system can quickly identify deviations and make corrections to ensure that the equipment operates in the optimal state and improve the accuracy and reliability of the ore conveying system.

[0140] Fault warning and response: Predictive models are used to provide early warning of potential faults and trigger corresponding maintenance responses. Fault warning and response uses deep learning models to perform complex pattern recognition and trend analysis on equipment operation data, which can capture potential equipment fault signals in advance, thereby reducing the occurrence of equipment failures and unplanned downtime. By identifying and handling potential problems in advance, the system can improve equipment availability and operational efficiency, reduce maintenance costs and equipment losses, and enhance the stability and sustainable operation capabilities of the ore conveying system.

[0141] Please see attached Figure 1 In a preferred embodiment of the present invention, lightweight AI model deployment includes:

[0142] Data preprocessing: Standardize the collected sensor data to ensure the stability and accuracy of the model input. Data standardization and outlier removal ensure the stability and accuracy of the model when processing input data by adjusting data distribution and cleaning invalid data. The data preprocessing function improves the robustness of the model to input data, enabling the model to work stably under different data sources and different working conditions, reducing the impact of data quality issues on reasoning results. At the same time, the preprocessed data provides high-quality input for the AI ​​model, improves the accuracy and reliability of reasoning, and thus improves the intelligence level of the ore transportation system;

[0143] Model compression and optimization: Reduce model size and computational complexity through pruning and quantization techniques. The system compresses the model through pruning and quantization techniques. Pruning technology reduces the redundant calculation of the model by removing neuron connections that contribute less to reasoning, while quantization technology converts model parameters from 32-bit floating point numbers to 8-bit integers to reduce the storage and computing requirements of the model. By reducing the model size and computational complexity, the system reduces the computing load and energy consumption of edge devices, extends the service life of the equipment, and ensures the stability of physical performance.

[0144] Real-time reasoning optimization: Using batch processing and parallel computing technology, efficient real-time reasoning is achieved on edge devices. The system divides the real-time collected data into small batches and uses the multi-core processor or GPU of the device for parallel reasoning. The batch processing mode not only improves the reasoning speed, but also reduces the computational overhead of each reasoning process. Parallel computing technology combined with asynchronous task scheduling ensures that the system can maintain stable real-time reasoning capabilities under high load conditions. The real-time reasoning optimization function greatly improves the system's response speed to changes in equipment status, enabling ore transportation management to make intelligent decisions in a short time and adjust transportation strategies and parameters in a timely manner, providing strong support for efficient operation and intelligent decision-making of mines;

[0145] System integration and control: Integrate AI reasoning results with system control modules to achieve automated decision-making. System integration achieves data-driven automated decision-making by seamlessly connecting AI reasoning results with control modules. The analysis results of AI models on real-time data directly affect the control strategy of the equipment, forming an adaptive, self-optimizing intelligent control system, thereby reducing the risk of human intervention and operational errors. Through real-time AI reasoning and automatic control, the system can continuously optimize the equipment operating status and transportation strategy, improve the efficiency and reliability of ore transportation, and at the same time improve the productivity and economic benefits of the mine;

[0146] Benefit analysis: Real-time evaluation of the contribution of AI models to system benefits, including equipment response time, energy consumption optimization, and maintenance cost control. Benefit analysis quantifies the specific contribution of AI to system performance improvement by comparing actual data with the optimization suggestions of the AI ​​model. The benefit analysis function provides a quantitative evaluation of the actual value of the AI ​​model in the mine conveying system, helping management to fully understand the economic benefits and operational improvement effects of AI deployment. At the same time, through detailed benefit reports, the system can identify key points of optimization, further enhance the intelligence and economy of mine operations, and provide data support and decision-making support for continuous improvement.

[0147] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent mine ore transportation management system, characterized in that: include: Deep reinforcement learning scheduling module, which is used to achieve optimized scheduling through deep reinforcement learning algorithm based on the status of ore conveying equipment, conveying path conditions and ore queue information; Digital twin system, used to create digital twin models of conveying equipment, perform multi-physics simulation, real-time data synchronization and status prediction to manage the entire life cycle of the equipment; Edge computing module, used for data collection, real-time processing and lightweight AI reasoning at the mine site, to achieve equipment anomaly detection and early warning; Blockchain data management module, used to record and store equipment operation logs and maintenance records, and ensure data security and immutability through blockchain; The network security protection module is used to monitor the mine network security in real time through the AI-driven network detection and response system to protect against potential network attacks and data leaks; User interface and control platform for operators to monitor and control the ore conveying system in real time, providing status visualization, alarm management and operation interface.

2. The intelligent mine ore transportation management system according to claim 1 is characterized in that: The deep reinforcement learning scheduling module includes: State space modeling: by converting equipment status, path status and ore queue information into multi-dimensional vectors and inputting them into the deep reinforcement learning model; Reward function construction: Establish a multi-objective optimization reward function based on delivery time, energy consumption, maintenance cost and safety; Strategy and value network training: PPO algorithm is used for strategy optimization, and training is performed through the loss function consisting of strategy ratio, value function loss and strategy entropy; Strategy execution and optimization: Use distributed architecture to deploy scheduling strategies, and adjust and optimize strategies through real-time feedback data.

3. The intelligent mine ore transportation management system according to claim 1 is characterized in that: The digital twin system includes: Simulation model establishment: Use multi-physics field coupling method to build dynamic, thermal and electrical characteristic models of conveying equipment; Real-time data synchronization: Synchronize sensor data with the digital twin in real time through the OPC UA protocol to ensure that the model reflects the real-time status of the equipment; State estimation and updating: Use Kalman filter to perform state estimation and error correction on real-time data; Full life cycle management: Analyze the equipment operation status through the Markov chain model, and combine the LSTM prediction model to perform fault prediction and maintenance optimization; Redundancy design and fault management: Design equipment redundancy systems to automatically switch to backup equipment or systems when equipment or sensors fail to ensure continuous operation of the system.

4. The intelligent mine ore transportation management system according to claim 1 is characterized in that: The edge computing module includes: Edge hardware configuration: Use NVIDIA Jetson Nano for edge computing and deploy the EdgeX Foundry platform for data management; Data collection and preprocessing: Collect sensor data through the device service layer and perform standardization and denoising; Lightweight AI model deployment: Use TensorFlow Lite to deploy quantized convolutional neural networks for real-time anomaly detection; Real-time reasoning and feedback: Use batch processing and parallel computing technology to achieve efficient real-time reasoning and feedback to the central control system; Benefit Analysis: Integrated benefit analysis module to evaluate equipment performance and system efficiency in real time, including response time, energy consumption and maintenance cost savings.

5. The intelligent mine ore transportation management system according to claim 1 is characterized in that: The blockchain data management module includes: Build a blockchain network: Build a blockchain network based on Hyperledger Fabric, with nodes including device controllers and management servers; Smart contract implementation: Deploy smart contracts to record and verify device operation logs and maintenance records, and use the PBFT protocol to reach consensus; Data storage and verification: Hash calculation is performed on each operation and stored in the blockchain to ensure data integrity and immutability; Multi-copy storage and permission management: Implement multi-copy data storage and access permission control to ensure the system's fault tolerance and data security; Data tracing and analysis: Through the chain structure of blockchain, historical operation data can be traced and analyzed to provide data support for equipment maintenance decisions.

6. The intelligent mine ore transportation management system according to claim 1, characterized in that: The network security protection module includes: Real-time network monitoring: Use AI-based network detection systems to monitor data traffic, identify and warn of abnormal traffic; Anomaly detection model: Use the autoencoder model to perform unsupervised learning on normal traffic and identify anomalies based on reconstruction errors; Dynamic security threshold adjustment: Based on the real-time network status, the threshold of anomaly detection is adaptively adjusted to improve the detection accuracy and response speed; Security response mechanism: When abnormal traffic is detected, isolation, alarm and protection operations are automatically triggered; Redundant path design: Configure redundant network paths to ensure automatic switching to backup paths in the event of network failure or attack, thus ensuring system continuity.

7. The intelligent mine ore transportation management system according to claim 1 is characterized in that: The user interface and control platform include: Data visualization: Provides real-time visualization of conveyor equipment status, path conditions, and ore queue information; Scheduling control: Control the scheduling module through the interface and adjust the transportation strategy and parameters; Alarm management: Real-time display of system abnormality and fault information, and provision of alarm management and emergency operation functions; Performance analysis and reporting: Provides system performance analysis and benefit reporting interface, showing indicators of response time, energy savings and maintenance costs; User authority configuration: Configure access rights for different user levels to ensure system operation security and data confidentiality.

8. The intelligent mine ore transportation management system according to claim 2 is characterized in that: The reward function construction includes: Time efficiency calculation: The time efficiency bonus is calculated by the ratio of actual delivery time to target delivery time; Energy consumption assessment: Calculate energy consumption rewards based on the ratio of current energy consumption to target energy consumption to encourage energy efficiency optimization; Maintenance cost assessment: Calculate the ratio of current maintenance cost to preset maintenance budget to optimize maintenance plan; Safety assessment: Positive rewards are given when equipment operation meets safety standards; Queue management: monitor and control the length of the ore queue to avoid reduced transportation efficiency due to excessively long queues.

9. The intelligent mine ore transportation management system according to claim 3 is characterized in that: The real-time data synchronization includes: Data filtering: Use filtering algorithms to remove noise and outliers in data and improve data quality; Timestamp correction: Ensure the synchronization of sensor data and digital twins through timestamp correction; Predictive status update: status prediction based on short-term historical data to optimize and adjust equipment parameters in real time; Two-way feedback mechanism: Establish a two-way data feedback mechanism between the digital twin and the actual device to ensure that the model is consistent with the device status; Fault warning and response: Use predictive models to provide early warning of potential faults and trigger corresponding maintenance responses.

10. The intelligent mine ore transportation management system according to claim 4, characterized in that: The lightweight AI model deployment includes: Data preprocessing: standardize the collected sensor data to ensure the stability and accuracy of the model input; Model compression and optimization: reduce model size and computational complexity through pruning and quantization techniques; Real-time inference optimization: Use batch processing and parallel computing technology to achieve efficient real-time inference on edge devices; System integration and control: Integrate AI reasoning results with system control modules to achieve automated decision-making; Benefit analysis: Real-time evaluation of the contribution of AI models to system benefits, including equipment response time, energy consumption optimization, and maintenance cost control.

Images