Double-flywheel system, intelligent mine field system and method

Abstract

The invention discloses a double-flywheel system and an intelligent mine field system and method.The double-flywheel system comprises a sensing system and a planning system, the sensing system fuses acquired multi-dimensional environment data of a mine field through the bird's-eye view angle technology and a deep learning algorithm, a mine field 3D environment model is established, and environment sensing data is output; the planning system performs planning path prediction on each mine car by using a deep learning algorithm and a Monte Carlo tree search algorithm based on vehicle management information, task scheduling data and environment perception data fed back by the motorcade management platform, and generates an optimal driving path. The vehicle management information comprises vehicle posture and health information, vehicle position state information, vehicle operation task priority information and mine field safety standard information. According to the invention, through deep cooperation of sensing and planning and integration of motorcade management information and safety standards, the vehicles can work according to the optimal path more accurately, the working yield of the mine field is improved, and the development of the intelligent mine field is promoted.

Description

Technical Field

[0001] The present invention relates to the field of unmanned driving technology, and in particular to a dual flywheel system, an intelligent mine system and a method. Background Art

[0002] With the rapid development of unmanned driving technology and the growing demand for automation in the mining industry, intelligent driving systems have gradually become a key technology in mines and other complex industrial environments.

[0003] Traditional intelligent driving systems usually rely on a single perception or planning model, or even some end-to-end models. Although end-to-end models have received widespread attention in unmanned driving technology in recent years, they reduce the data processing chain by directly generating driving control instructions from sensor inputs, and show certain flexibility and efficiency in public road scenarios. However, the application of existing end-to-end models still faces technical bottlenecks such as huge data requirements, reasoning "hallucination" phenomena, and difficulty in real-time response to dynamic environments. These problems can directly affect the robustness and safety of the end-to-end model, and the end-to-end model cannot achieve deep adaptation to the physical environment in the highly complex environment of the mine, especially in slopes, gravel and other complex terrains, and cannot accurately plan the path, resulting in reduced operating productivity in the mine. Summary of the invention

[0004] Based on this, it is necessary to address the above technical problems. The embodiments of the present invention provide a dual flywheel system, an intelligent mine system and a method, which can effectively solve the problems of insufficient robustness and safety in the end-to-end model, thereby achieving accurate path planning and improving the operating productivity of the mine.

[0005] A first aspect of an embodiment of the present application provides an intelligent mine system, wherein the dual flywheel system includes a perception system and a planning system; The perception system includes a multimodal sensor and a data processing module. The multimodal sensor is used to obtain multi-dimensional environmental data of the mine. The data processing module is used to fuse the multi-dimensional environmental data using bird's-eye view technology and deep learning algorithm to establish a 3D environmental model of the mine to output environmental perception data, and transmit the environmental perception data to the planning system. The planning system includes a path planning module, which is used to predict the planning path of each mine car based on the vehicle management information, task scheduling data and the environmental perception data fed back by the fleet management platform, and generate the optimal driving path by using a deep learning algorithm and a Monte Carlo tree search algorithm, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information.

[0006] Further, the multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module and a visual sensor module; The laser radar sensing module detects the surrounding environment by emitting a laser beam; The millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; The visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

[0007] Further, the planning system includes a task decision module and a safety monitoring module; The task decision module is used to adjust the driving path and operation sequence of each mining vehicle in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information; The safety monitoring module is used to monitor the driving path of each mining vehicle and the potential risks in the operation tasks in real time based on the mine safety standard information.

[0008] A second aspect of an embodiment of the present application provides an intelligent mine system, the intelligent mine system comprising a fleet management platform, a dual flywheel system and a vehicle control system, the dual flywheel system comprising a perception system and a planning system; The fleet management platform includes a task scheduling module and a data sharing module. The task scheduling module is used to assign the operation tasks corresponding to each vehicle based on the acquired vehicle management information to obtain the task scheduling data, and transmit the vehicle management information and the task scheduling data to the planning system in the dual flywheel system. The data sharing module is used to provide the operation scheduling instructions and transmit the operation scheduling instructions to the vehicle control system, wherein the vehicle management information includes the vehicle posture and health information, the vehicle position status information, the vehicle operation task priority information and the mine safety standard information; The perception system includes a multimodal sensor and a data processing module. The multimodal sensor is used to obtain multi-dimensional environmental data of the mine. The data processing module is used to fuse the multi-dimensional environmental data using bird's-eye view technology and deep learning algorithm to establish a 3D environmental model of the mine to output environmental perception data, and transmit the environmental perception data to the planning system. The planning system comprises a path planning module, wherein the path planning module is used to predict the planning path of each mining vehicle based on the vehicle management information, the task scheduling data and the environmental perception data by using a deep learning algorithm and a Monte Carlo tree search algorithm, generate an optimal driving path, and transmit the optimal driving path to the vehicle control system; The vehicle control system is used to control the vehicle to travel along the optimal driving path based on the operation scheduling instruction.

[0009] Further, the multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module and a visual sensor module; The laser radar sensing module detects the surrounding environment by emitting a laser beam; The millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; The visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

[0010] Further, the planning system includes a task decision module and a safety monitoring module; The task decision module is used to adjust the driving path and operation sequence of each mining vehicle in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information; The safety monitoring module is used to monitor the driving path of each mining vehicle and the potential risks in the operation tasks in real time based on the mine safety standard information.

[0011] Furthermore, the intelligent mine system further comprises an environment adaptation system, and the environment adaptation system comprises an adaptation module and an adjustment module; The adaptation module is used to adapt to the driving requirements of the vehicle in different physical environments; The adjustment module is used to adjust the travel speed and travel mode of each mine car in real time according to the environmental status information of the mine.

[0012] Furthermore, the intelligent mine system also includes an artificial closed-loop system, which includes a manual instruction module and a closed-loop feedback module; The manual instruction module is used to input manual instructions through a manual interface and transmit the manual instructions to the vehicle control system; The closed-loop feedback module is used to monitor and feedback the execution status of the manual instruction in real time.

[0013] Further, the vehicle control system includes an energy management module; The energy management module is used to dynamically adjust the power output of each mining vehicle according to the battery status and route requirements.

[0014] A third aspect of an embodiment of the present application provides a path planning control method, which is applied to the intelligent mine system according to any one of the second aspects above, and includes: Based on the acquired vehicle management information, the operation tasks corresponding to each vehicle are assigned to obtain task scheduling data, wherein the vehicle management information includes vehicle posture and health information, vehicle location status information, vehicle operation task priority information and mine safety standard information; Sending the task scheduling data and the vehicle management information to the dual flywheel system, so that the dual flywheel system uses a deep learning algorithm and a Monte Carlo tree search algorithm to predict the planning path of each mining car based on the vehicle management information, the task scheduling data and the environmental perception data, generates an optimal driving path, and transmits the optimal driving path to the vehicle control system; Sending a work scheduling instruction to the vehicle control system so that the vehicle control system responds to the work scheduling instruction and controls the vehicle to travel according to the optimal driving path: In summary, the present invention provides a dual flywheel system, an intelligent mine system and method, the dual flywheel system includes a perception system and a planning system, the perception system is used to use bird's-eye view technology and deep learning algorithm to fuse the multi-dimensional environmental data of the mine, establish a 3D environment model of the mine, output environmental perception data, and transmit the environmental perception data to the planning system, the planning system is used to predict the planning path of each mine car based on the vehicle management information, task scheduling data and environmental perception data fed back by the fleet management platform, and generate the optimal driving path by using deep learning algorithm and Monte Carlo tree search algorithm, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information. It can be seen that the present application can accurately generate the optimal driving path by combining the collaborative architecture of the perception system and the planning system, and integrating the fleet management information and safety standards, so that the vehicle can operate more accurately according to the optimal path, reduce unnecessary waiting time and path conflicts, not only improve the operation yield of the mine, but also lay a solid technical foundation for the development of future intelligent mines. In addition, the dual flywheel system of the present invention has significant advantages in target optimization and modeling complexity. By introducing rule-based enhancement mechanisms (deep learning algorithms and Monte Carlo tree search algorithms) in the decision-making and planning links, it effectively solves the problems of insufficient robustness and security in the end-to-end model. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technical businesses in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0016] Figure 1 It is a structural schematic diagram of an intelligent mine system provided by an embodiment of the present invention; Figure 2 is another structural schematic diagram of an intelligent mine system provided by an embodiment of the present invention; Figure 3is a structural schematic diagram of a dual flywheel system provided by one embodiment of the present invention; Figure 4 It is a flow chart of a path planning control method provided by an embodiment of the present invention; DETAILED DESCRIPTION The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0017] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0018] It should also be understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0019] As used in the present specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]", depending on the context.

[0020] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0021] References to "one embodiment" or "some embodiments" etc. described in the present specification mean that one or more embodiments of the present invention include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0022] It should be understood that the order of execution of the steps in the following embodiments does not imply a precedence of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0023] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.

[0024] Figure 1 is a structural diagram of an intelligent mine system according to an exemplary embodiment of the present application. Figure 1 As shown, the intelligent mine system includes a fleet management platform, a dual flywheel system and a vehicle control system, wherein the dual flywheel system includes a perception system and a planning system; The fleet management platform includes a task scheduling module and a data sharing module. The task scheduling module is used to assign the corresponding operation tasks to each vehicle based on the acquired vehicle management information to obtain the task scheduling data, and transmit the vehicle management information and the task scheduling data to the planning system in the dual flywheel system. The data sharing module is used to provide the operation scheduling instructions and transmit the operation scheduling instructions to the vehicle control system, wherein the vehicle management information includes the vehicle posture and health information, the vehicle position status information, the vehicle operation task priority information and the mine safety standard information; The perception system includes a multimodal sensor and a data processing module. The multimodal sensor is used to obtain multi-dimensional environmental data of the mine. The data processing module is used to fuse the multi-dimensional environmental data using bird's-eye view technology and deep learning algorithms, establish a 3D environmental model of the mine, output environmental perception data, and transmit the environmental perception data to the planning system; A planning system, including a path planning module, which is used to predict the planning path of each mining car based on vehicle management information, task scheduling data and environmental perception data using a deep learning algorithm and a Monte Carlo tree search algorithm, generate an optimal driving path, and transmit the optimal driving path to the vehicle control system; The vehicle control system is used to control the vehicle to travel along the optimal driving path based on the job scheduling instructions.

[0025] In the present application embodiment, Figure 2 As shown, the intelligent mine system includes a fleet management platform, a dual flywheel system and a vehicle control system. The fleet management platform includes a task scheduling module and a data sharing module. The task scheduling module is used to assign the corresponding operation tasks of each vehicle based on the acquired vehicle management information to obtain the task scheduling data, and transmit the vehicle management information and task scheduling data to the planning system in the dual flywheel system to ensure the optimal path planning for vehicle operation and charging. The data sharing module is used to provide unified operation scheduling instructions through the connection with all mine vehicles, charging piles and perception modules, and transmit the operation scheduling instructions to the vehicle control system, thereby realizing global information sharing. Among them, the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information. The vehicle posture information refers to the vehicle posture, which refers to the overall state and appearance shape of the vehicle, including the contact with the ground, the tilt angle of the vehicle body and whether the vehicle is vertical, including normal driving posture, tilt posture and abnormal posture, etc.; vehicle health information refers to information related to the vehicle operation status, component health status, etc., including: engine health information, chassis health information and body health information, etc. Vehicle location status information refers to the real-time location, driving direction, speed and current status (such as empty or fully loaded) of the vehicle in the mine area. It is usually obtained through GPS and sensors, which can help the dispatch center monitor the operation of the vehicle in real time to ensure operational efficiency and safety. Vehicle operation task priority information refers to the order and priority of the vehicle's execution of tasks. For example, some tasks may be more urgent or important than other tasks, so they need to be handled first, which helps to optimize resource allocation within the mine, improve operational efficiency, and ensure that key tasks can be completed in a timely manner. Mine safety standard information refers to the various safety standards and regulations that need to be followed during mine operations. This includes operating procedures, accident prevention measures, emergency response procedures, etc. These standards are designed to ensure the safety of operators and the safety of equipment and the environment, and ensure that mine operations comply with relevant laws and regulations. By effectively integrating and managing the three types of information, and combining the system with vehicle management information and task scheduling data, the subsequent dual flywheel system can predict through multi-agent interaction, so that the driving path and operation priority of each mine car do not interfere with each other, and achieve collaborative operation effects. The collaborative management of the mine fleet greatly reduces path conflicts and unnecessary waiting time, achieves optimal utilization of mine resources, and thus improves overall operational efficiency and productivity.

[0026] In this embodiment, the dual flywheel system includes a perception system and a planning system, wherein the perception system includes a multimodal sensor and a data processing module, wherein the multimodal sensor is used to obtain multidimensional environmental data of the mine, and the data processing module is used to pre-filter, denoise and correct the multidimensional environmental data to improve the quality and accuracy of the data, and then fuse the multidimensional environmental data through bird's-eye view technology and deep learning algorithm to form a more complete and reliable 3D environment model of the mine, to provide real-time detection and recognition of dynamic and static objects, and thus output environmental perception data, that is, feature extraction is performed on each modal data, and a multimodal feature fusion structure is designed to effectively fuse the features of different modalities together to construct a 3D environment model of the mine, including convolutional layers, fully connected layers, etc. for extracting environmental features, and the output layer is used for classification, and the features after the fusion of multimodal data are classified into environmental conditions, wherein the bird's-eye view technology is a visualization technology that presents spatial information through a bird's-eye view, and is often used in fields such as autonomous driving, drones, and map services; deep learning is a branch of machine learning, which uses multi-layer neural networks for data learning and pattern recognition for automatic feature extraction. Through the above steps, the vehicle can effectively identify slopes, obstacles and other dynamic entities in complex mines, improving the accuracy of environmental perception. This reduces operation interruptions and safety hazards caused by environmental misjudgment, significantly improves the safety of mine operations, and transmits environmental perception data to the planning system through the information interaction interface to facilitate decision-making and path planning.

[0027] In this embodiment, the planning system includes a path planning module, wherein the path planning module is used to predict the planning path of each mine car based on vehicle management information, task scheduling data and environmental perception data, using deep learning algorithms and Monte Carlo tree search algorithms, generate the optimal driving path, and quickly respond to changes in task priority, vehicle status and other subject behaviors to ensure that the unmanned vehicle can maintain continuous operation. This not only reduces waiting time, but also avoids waste of resources, significantly improves the response efficiency of the vehicle and the flexibility of path planning, and transmits the optimal driving path to the vehicle control system. Among them, deep learning is a branch of machine learning, which uses multi-layer neural networks for data learning and pattern recognition for automatic feature extraction. Monte Carlo Tree Search (MCTS) is a heuristic search algorithm for decision-making processes, which is used for complex problems with huge and difficult-to-exhaust state spaces. By introducing bird's-eye view technology and deep learning algorithms, this application can convert multimodal data into a bird's-eye view and generate a more comprehensive and high-precision 3D environment model of the mine. By introducing deep learning algorithms and Monte Carlo tree search algorithms, using simulated random strategies to explore possible decision paths, and using statistical methods to evaluate the potential of each node, the search tree is gradually constructed and optimized, so as to provide unmanned vehicles with multi-agent interactive prediction capabilities, and can simulate and predict multi-party interactive behaviors in complex traffic scenes, so that vehicles can be more proactive in dealing with dynamic environments, and for the three-dimensional motion characteristics of mine cars (such as uphill and downhill, load changes), the planning system of the present invention enhances the adaptability to multi-dimensional dynamics through modular training. It can be seen that the dual flywheel system of the present application integrates the perception system and the planning system, so that the system can work efficiently and collaboratively in a dynamic and complex environment, and realizes the deep coupling of high-precision perception and flexible planning, and introduces a rule-based enhancement mechanism (deep learning algorithm and Monte Carlo tree search algorithm) in the decision-making and planning links, realizing the architectural innovation of combining feature extraction of end-to-end models with rule-based safety mechanisms, thereby effectively solving the problem of insufficient robustness and security in the end-to-end model.

[0028] It should be noted that the generation of environmental perception data and the optimal driving path can also be achieved by using other methods such as perception algorithms, Bayesian networks or Markov decision processes. The specific settings can be made according to actual conditions, and this application does not impose any limitations on this.

[0029] In this embodiment, the vehicle control system receives the operation scheduling instructions from the fleet management platform, controls the vehicle to travel along the optimal driving path, and performs operations such as acceleration, steering and braking.

[0030] In the embodiment of the present application, the perception system is combined with the planning system: the high-precision environmental perception data generated by the perception system is transmitted to the planning system in real time through the information interaction interface; the planning system calculates the path planning and decision-making strategy based on this information, and sends the command to the vehicle control system. It can be seen that through the deep coordination of perception and planning, as well as the integration of fleet management information and safety standards, the dual flywheel system of the present invention effectively improves the automation level and operation efficiency of mine operations, and unmanned mine vehicles can operate more accurately according to the optimal path, reducing unnecessary waiting time and path conflicts, thereby maximizing the operating productivity of the mine and promoting the development of smart mines; the planning system is then combined with the fleet management platform: the fleet management system provides the planning system with information such as vehicle location, task priority and safety standards, to ensure that the planning system is on the path The scheduling data can be updated in real time during planning and multi-agent collaborative decision-making. It can be seen that the collaborative decision-making of multiple vehicles is realized through the integration of the planning system and the fleet management system. After receiving the scheduling instructions of the fleet management system and the real-time status data of other vehicles, the planning system can optimize the operation path and operation sequence of each mine car. In this way, the driving paths between vehicles do not interfere with each other, and the overall operation efficiency of the mine vehicles is significantly improved. In addition, the planning system can dynamically adjust the driving and operation paths of the vehicles according to the real-time position of the vehicles and the traffic conditions of the mines, to ensure the efficient and collaborative operation of the entire fleet; finally, the fleet management platform is combined with the vehicle control system: the scheduling instructions of the fleet management system are sent to the vehicle control system through the control instruction interface, so as to realize the synchronization of the mine cars and the mine operation tasks, and significantly improve the production efficiency and operation safety of the mines.

[0031] For example, when working in a mine, the perception system of vehicle A detects obstacles ahead in real time and calculates the shortest path to neighboring vehicles. After receiving the job scheduling instructions from the fleet management platform, the planning system takes into account the task priority and vehicle battery status and selects a safe and efficient path. When vehicle A approaches the charging pile, the system automatically determines the timing and task priority of charging, optimizes work efficiency and reduces vehicle downtime. Through the above steps, the intelligence level of vehicle perception, planning and collaborative operation is significantly improved, ensuring the operating efficiency, safety and productivity of mine vehicles.

[0032] In one embodiment, the multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module, and a visual sensor module; The laser radar sensing module detects the surrounding environment by emitting a laser beam; The millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; The visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

[0033] In the embodiment of the present application, the multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module and a visual sensor module. The perception system uses sensors such as the laser radar sensing module, the millimeter wave radar sensing module and the visual sensor module to accurately obtain multi-dimensional environmental data around the vehicle in real time, including road conditions, obstacles, traffic signs and other vehicles, so as to integrate the multi-dimensional environmental data into a 3D environmental model of the mine, thereby detecting and identifying dynamic and static objects in real time. That is, the laser radar sensing module detects the surrounding environment by emitting laser beams; the millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; and the visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

[0034] In one embodiment, the planning system includes a task decision module and a safety monitoring module; The task decision module is used to adjust the driving path and operation sequence of each mining vehicle in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information; The safety monitoring module is used to monitor the driving path of each mining vehicle and the potential risks in the operation tasks in real time based on the mine safety standard information.

[0035] In an embodiment of the present application, the planning system includes a task decision module and a safety monitoring module. The task decision module is used to adjust the driving path and operation sequence of each mine car in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information fed back by the fleet management platform, thereby realizing multi-vehicle collaborative operation. The safety monitoring module is used to monitor the driving path and potential risks in the operation tasks of each mine car in real time based on the ISO-17757 mine safety standard information, and automatically adjust the driving strategy when encountering emergencies. By complying with mine safety standards such as ISO-17757, the system monitors the vehicle path and task execution in real time, can promptly identify potential risks and make proactive adjustments. This monitoring mechanism based on safety standards ensures the compliance operation of the vehicle and reduces the safety accident rate in mine operations. In one embodiment, the intelligent mine system further includes an environment adaptation system, and the environment adaptation system includes an adaptation module and an adjustment module; The adaptation module is used to adapt to the driving requirements of the vehicle in different physical environments; The adjustment module is used to adjust the travel speed and travel mode of each mine car in real time according to the environmental status information of the mine.

[0036] In the embodiment of the present application, the intelligent mine system also includes an environmental adaptation system, wherein the environmental adaptation system includes an adaptation module and an adjustment module. Through the sensor network, the terrain slope, gravel distribution and other information of the mine are collected in real time. The adaptation module is used to dynamically adapt to the driving needs of the vehicle in different physical environments. When encountering steep slopes or special terrains, the adjustment module is used to adjust the driving speed and driving mode of each mine car in real time according to the environmental status information of the mine to ensure the stability and safety of the operation. The environmental adaptation system of the present application can monitor the terrain changes in the mine in real time, such as slope and gravel distribution, and adjust the vehicle driving mode accordingly, so as to ensure the stability and safety of the vehicle in complex terrain. Through innovations that adapt to the physical environment, the system can maintain stable operation under different terrains and improve the operating ability of the mine car in variable terrains.

[0037] In one embodiment, the intelligent mine system further comprises an artificial closed-loop system, wherein the artificial closed-loop system comprises an artificial instruction module and a closed-loop feedback module; The manual instruction module is used to input manual instructions through a manual interface and transmit the manual instructions to the vehicle control system; The closed-loop feedback module is used to monitor and feedback the execution status of the manual instruction in real time.

[0038] In the embodiment of the present application, the intelligent mine system also includes an artificial closed-loop system, wherein the artificial closed-loop system includes an artificial instruction module and a closed-loop feedback module, the artificial instruction module is used to input artificial instructions through an artificial interface, and transmit the artificial instructions to the vehicle control system, so that the system can respond quickly and feedback the adjustment to the control system, ensuring that the mine car has flexibility in special tasks or emergencies, and the closed-loop feedback module is used to monitor and feedback the execution status of the artificial instructions in real time, ensuring the accurate execution and effectiveness of the instructions during the vehicle driving process. That is, when the operator is allowed to input artificial instructions to the intelligent mine system in special tasks or emergencies, the system responds quickly and adjusts the vehicle path in emergency or non-standard scenarios. If complex terrain such as steep slopes is detected, the system automatically adapts and adjusts, thereby ensuring the accurate execution of instructions, greatly improving the flexibility and reliability in mine operations, reducing the impact of emergencies on operations, and further improving safety and flexibility.

[0039] In one embodiment, a vehicle control system includes an energy management module; The energy management module is used to dynamically adjust the power output of each mining vehicle according to the battery status and route requirements.

[0040] In an embodiment of the present application, the vehicle control system includes an energy management module, which is used to dynamically adjust the power output of the vehicle according to the battery status and path requirements of each mine car to achieve efficient energy management. It can be seen that the present application optimizes power output and energy consumption by combining the battery status and operation requirements of the vehicle, and reduces unnecessary braking, acceleration and waiting time through efficient perception and planning. The energy consumption of the mine car during operation is significantly reduced, which prolongs the battery life and helps the mine achieve energy conservation, emission reduction and sustainable development goals.

[0041] In the embodiment of the present application, the intelligent mine system includes a fleet management platform, a dual flywheel system and a vehicle control system. The task scheduling module in the fleet management platform allocates the work tasks corresponding to each vehicle based on the acquired vehicle management information to obtain the task scheduling data, and transmits the vehicle management information and the task scheduling data to the planning system in the dual flywheel system, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle work task priority information and mine safety standard information, and the perception system in the dual flywheel system uses the bird's-eye view technology and the deep learning algorithm to fuse the acquired multi-dimensional environmental data of the mine, establish a 3D environment model of the mine, output the environmental perception data, and transmit the environmental perception data to the planning system. A planning system is provided so that the planning system can predict the planning path of each mine car based on vehicle management information, task scheduling data and environmental perception data by using deep learning algorithms and Monte Carlo tree search algorithms, generate the optimal driving path, and then transmit the optimal driving path to the vehicle control system, and transmit the job scheduling instructions provided by the data sharing module in the fleet management platform to the vehicle control system, so that the vehicle control system can control the vehicle according to the optimal driving path based on the job scheduling instructions to complete the job task efficiently and safely in a dynamic environment. It can be seen that this application solves the limitations of existing unmanned mine cars in complex mining environments by combining perception, planning and collaborative management, thereby improving the safety, efficiency and productivity of mine operations.

[0042] Figure 3 is a structural schematic diagram of a dual flywheel system according to an exemplary embodiment of the present application. Figure 3 As shown, the dual flywheel system includes a perception system and a planning system; A perception system, comprising a multimodal sensor and a data processing module, wherein the multimodal sensor is used to obtain multidimensional environmental data of the mine, and the data processing module is used to fuse the multidimensional environmental data using a bird's-eye view technology and a deep learning algorithm, establish a 3D environmental model of the mine, output environmental perception data, and transmit the environmental perception data to the planning system; The planning system includes a path planning module, which is used to predict the planning path of each mine car based on the vehicle management information, task scheduling data and the environmental perception data fed back by the fleet management platform, and generate an optimal driving path by using a deep learning algorithm and a Monte Carlo tree search algorithm, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information.

[0043] In the embodiment of the present application, the dual flywheel system includes a perception system and a planning system, wherein the perception system includes a multimodal sensor and a data processing module, the multimodal sensor is used to obtain multidimensional environmental data of the mine, the data processing module is used to pre-filter, denoise and correct the multidimensional environmental data to improve the quality and accuracy of the data, and then the multidimensional environmental data is fused through bird's-eye view technology and deep learning algorithm to form a more complete and reliable mine 3D environmental model to provide real-time detection and recognition of dynamic and static objects, thereby outputting environmental perception data, that is, extracting features from each modal data, and designing a multimodal feature fusion structure to effectively fuse features of different modalities together to construct a mine 3D environmental model, including convolutional layers, fully connected layers, etc. for extracting environmental features, and the output layer is used for classification, and the features after the fusion of multimodal data are classified into environmental conditions, wherein the bird's-eye view technology is a visualization technology that presents spatial information through a bird's-eye view, which is often used in fields such as autonomous driving, drones, and map services; deep learning is a branch of machine learning, which uses multi-layer neural networks for data learning and pattern recognition for automatic feature extraction. Through the above steps, the vehicle can effectively identify slopes, obstacles and other dynamic entities in complex mines, improving the accuracy of environmental perception. This reduces operation interruptions and safety hazards caused by environmental misjudgment, significantly improves the safety of mine operations, and transmits environmental perception data to the planning system through the information interaction interface to facilitate decision-making and path planning.

[0044] In an embodiment of the present application, the planning system includes a path planning module, wherein the path planning module is used to predict the planning path of each mine car based on vehicle management information, task scheduling data and environmental perception data, using a deep learning algorithm and a Monte Carlo tree search algorithm, generate the optimal driving path, and quickly respond to changes in task priority, vehicle status and other subject behaviors to ensure that the unmanned vehicle can maintain continuous operation. This not only reduces waiting time, but also avoids waste of resources, significantly improves the response efficiency of the vehicle and the flexibility of path planning, and transmits the optimal driving path to the vehicle control system. Among them, deep learning is a branch of machine learning, which uses multi-layer neural networks for data learning and pattern recognition for automatic feature extraction. Monte Carlo Tree Search (MCTS for short) is a heuristic search algorithm for decision-making processes, which is used for complex problems with huge and difficult-to-exhaust state spaces. By introducing bird's-eye view technology and deep learning algorithms, this application can convert multimodal data into a bird's-eye view to generate a more comprehensive and high-precision 3D environment model of the mine. By introducing deep learning algorithms and Monte Carlo tree search algorithms, using simulated random strategies to explore possible decision paths, and using statistical methods to evaluate the potential of each node, the search tree is gradually constructed and optimized, so as to provide unmanned vehicles with multi-agent interactive prediction capabilities, and can simulate and predict multi-party interactive behaviors in complex traffic scenes, so that vehicles can be more proactive in dealing with dynamic environments, and for the three-dimensional motion characteristics of mine cars (such as uphill and downhill, load changes), the planning system of the present invention enhances the adaptability to multi-dimensional dynamics through modular training. It can be seen that the dual flywheel system of the present application integrates the perception system and the planning system, so that the system can work efficiently and collaboratively in a dynamic and complex environment, and realizes the deep coupling of high-precision perception and flexible planning, and introduces a rule-based enhancement mechanism (deep learning algorithm and Monte Carlo tree search algorithm) in the decision-making and planning links, realizing the architectural innovation of combining feature extraction of end-to-end models with rule-based safety mechanisms, thereby effectively solving the problem of insufficient robustness and security in the end-to-end model.

[0045] It should be noted that the generation of environmental perception data and the optimal driving path can also be achieved by replacing other methods such as perception algorithms, Bayesian networks or Markov decision processes. The specific method can be selected to generate data and paths according to actual conditions and needs. This application does not impose any limitations on this.

[0046] In the embodiment of the present application, a dual flywheel system is formed by combining the perception system with the planning system, so that the dual flywheel system can improve the intelligence level of the unmanned mining car, so that the mine has efficient, safe and green operation capabilities, and promotes the process of intelligent and automated mines. With the substantial improvement of operating efficiency and safety, the overall competitiveness of the mine is enhanced, providing a solid technical foundation for the future development of intelligent mining.

[0047] In the embodiment of the present application, the dual flywheel system includes a perception system and a planning system. The perception system is used to use bird's-eye view technology and deep learning algorithms to integrate the multi-dimensional environmental data of the mine, establish a 3D environmental model of the mine, output environmental perception data, and transmit the environmental perception data to the planning system. The planning system is used to predict the planning path of each mine car based on the vehicle management information, task scheduling data and environmental perception data fed back by the fleet management platform, and use deep learning algorithms and Monte Carlo tree search algorithms to generate the optimal driving path, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information. It can be seen that the present application can accurately generate the optimal driving path by combining the collaborative architecture of the perception system and the planning system, as well as the integration of fleet management information and safety standards, so that the vehicle can operate more accurately according to the optimal path, reduce unnecessary waiting time and path conflicts, not only improve the operating productivity of the mine, but also lay a solid technical foundation for the development of future smart mines. In addition, the dual flywheel system of the present invention has significant advantages in target optimization and modeling complexity. By introducing rule-based enhancement mechanisms (deep learning algorithms and Monte Carlo tree search algorithms) in the decision-making and planning links, it effectively solves the problems of insufficient robustness and security in the end-to-end model.

[0048] Figure 4 is a flow chart of a path planning control method according to an exemplary embodiment of the present application. The path planning control method can be applied to Figure 1 or Figure 2 Intelligent mine systems in Figure 4 As shown, the path planning control method is implemented through the following steps. S401: Based on the acquired vehicle management information, the operation tasks corresponding to each vehicle are assigned to obtain task scheduling data, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information; S402: Sending the task scheduling data and the vehicle management information to the dual flywheel system, so that the dual flywheel system uses a deep learning algorithm and a Monte Carlo tree search algorithm to predict the planning path of each mining car based on the vehicle management information, the task scheduling data and the environmental perception data, generates an optimal driving path, and transmits the optimal driving path to the vehicle control system; S403: Sending a work scheduling instruction to the vehicle control system, so that the vehicle control system responds to the work scheduling instruction and controls the vehicle to travel along the optimal driving path. It should be noted that the implementation method of the present method embodiment is the same as the implementation method of the above-mentioned intelligent mine system embodiment, and will not be repeated here. In the embodiment of the present application, the operation tasks corresponding to each vehicle are assigned based on the acquired vehicle management information to obtain the task scheduling data, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information, and the task scheduling data and vehicle management information are sent to the dual flywheel system, so that the dual flywheel system uses the deep learning algorithm and the Monte Carlo tree search algorithm to plan the path prediction for each mine car based on the vehicle management information, task scheduling data and environmental perception data, generate the optimal driving path, and transmit the optimal driving path to the vehicle control system, and issue the operation scheduling instruction to the vehicle control system, so that the vehicle control system responds to the operation scheduling instruction and controls the vehicle to travel according to the optimal driving path. It can be seen that the present application integrates the fleet management information and safety standards, and then utilizes the deep coordination of perception and planning in the dual flywheel system, so as to accurately generate the optimal driving path, so that the vehicle can operate more accurately according to the optimal path, reduce unnecessary waiting time and path conflicts, thereby improving the operation yield of the mine and promoting the development of intelligent mines. In addition, the dual flywheel system of the present invention has significant advantages in target optimization and modeling complexity. By introducing rule-based enhancement mechanisms (deep learning algorithms and Monte Carlo tree search algorithms) in the decision-making and planning links, it effectively solves the problems of insufficient robustness and security in the end-to-end model.

[0049] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A dual flywheel system, characterized in that: The dual flywheel system includes a perception system and a planning system; The perception system includes a multimodal sensor and a data processing module. The multimodal sensor is used to obtain multi-dimensional environmental data of the mine. The data processing module is used to fuse the multi-dimensional environmental data using bird's-eye view technology and deep learning algorithm to establish a 3D environmental model of the mine to output environmental perception data, and transmit the environmental perception data to the planning system. The planning system includes a path planning module, which is used to predict the planning path of each mine car based on the vehicle management information, task scheduling data and the environmental perception data fed back by the fleet management platform, and generate an optimal driving path using a deep learning algorithm and a Monte Carlo tree search algorithm, wherein the vehicle management information includes vehicle posture and health information, vehicle position status information, vehicle operation task priority information and mine safety standard information.

2. The dual flywheel system according to claim 1, characterized in that: The multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module and a visual sensor module; The laser radar sensing module detects the surrounding environment by emitting a laser beam; The millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; The visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

3. The dual flywheel system according to claim 1, characterized in that: The planning system includes a task decision module and a safety monitoring module; The task decision module is used to adjust the driving path and operation sequence of each mining vehicle in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information; The safety monitoring module is used to monitor the driving path of each mining vehicle and the potential risks in the operation tasks in real time based on the mine safety standard information.

4. An intelligent mine system, characterized in that: The intelligent mine system includes a fleet management platform, a dual flywheel system and a vehicle control system, wherein the dual flywheel system includes a perception system and a planning system; The fleet management platform includes a task scheduling module and a data sharing module. The task scheduling module is used to assign the operation tasks corresponding to each vehicle based on the acquired vehicle management information to obtain the task scheduling data, and transmit the vehicle management information and the task scheduling data to the planning system in the dual flywheel system. The data sharing module is used to provide the operation scheduling instructions and transmit the operation scheduling instructions to the vehicle control system, wherein the vehicle management information includes the vehicle posture and health information, the vehicle position status information, the vehicle operation task priority information and the mine safety standard information; The perception system includes a multimodal sensor and a data processing module. The multimodal sensor is used to obtain multi-dimensional environmental data of the mine. The data processing module is used to fuse the multi-dimensional environmental data using bird's-eye view technology and deep learning algorithm to establish a 3D environmental model of the mine to output environmental perception data, and transmit the environmental perception data to the planning system. The planning system comprises a path planning module, wherein the path planning module is used to predict the planning path of each mining vehicle based on the vehicle management information, the task scheduling data and the environmental perception data by using a deep learning algorithm and a Monte Carlo tree search algorithm, generate an optimal driving path, and transmit the optimal driving path to the vehicle control system; The vehicle control system is used to control the vehicle to travel along the optimal driving path based on the operation scheduling instruction.

5. The intelligent mine system according to claim 4, characterized in that: The multimodal sensor includes a laser radar sensing module, a millimeter wave radar sensing module and a visual sensor module; The laser radar sensing module detects the surrounding environment by emitting a laser beam; The millimeter wave radar sensing module detects the surrounding environment by emitting millimeter wave bands; The visual sensor module uses optical elements and imaging technology to detect and obtain image information of the surrounding environment.

6. The intelligent mine system according to claim 4, characterized in that: The planning system includes a task decision module and a safety monitoring module; The task decision module is used to adjust the driving path and operation sequence of each mining vehicle in real time based on the vehicle posture and health information, vehicle position status information and vehicle operation task priority information; The safety monitoring module is used to monitor the driving path of each mining vehicle and the potential risks in the operation tasks in real time based on the mine safety standard information.

7. The intelligent mine system according to claim 4, characterized in that: The intelligent mine system further includes an environment adaptation system, which includes an adaptation module and an adjustment module; The adaptation module is used to adapt to the driving requirements of the vehicle in different physical environments; The adjustment module is used to adjust the travel speed and travel mode of each mine car in real time according to the environmental status information of the mine.

8. The intelligent mine system according to claim 4, characterized in that: The intelligent mine system also includes an artificial closed-loop system, which includes a manual instruction module and a closed-loop feedback module; The manual instruction module is used to input manual instructions through a manual interface and transmit the manual instructions to the vehicle control system; The closed-loop feedback module is used to monitor and feedback the execution status of the manual instruction in real time.

9. The intelligent mine system according to claim 4, characterized in that: The vehicle control system includes an energy management module; The energy management module is used to dynamically adjust the power output of each mining vehicle according to the battery status and route requirements.

10. A path planning control method, characterized in that: Based on the intelligent mine system according to any one of claims 4 to 9, the method comprises: Based on the acquired vehicle management information, the operation tasks corresponding to each vehicle are assigned to obtain task scheduling data, wherein the vehicle management information includes vehicle posture and health information, vehicle location status information, vehicle operation task priority information and mine safety standard information; Sending the task scheduling data and the vehicle management information to the dual flywheel system, so that the dual flywheel system uses a deep learning algorithm and a Monte Carlo tree search algorithm to predict the planning path of each mining car based on the vehicle management information, the task scheduling data and the environmental perception data, generates an optimal driving path, and transmits the optimal driving path to the vehicle control system; A work scheduling instruction is issued to the vehicle control system so that the vehicle control system responds to the work scheduling instruction and controls the vehicle to travel along the optimal driving path.

Images