Closed-loop cutting planning method driven by fusion of VR navigation and coal rock identification online planning

Abstract

The invention discloses a closed-loop cutting planning method based on fusion driving of VR navigation and coal rock recognition online planning, and the method comprises the steps: planning a cutting path of a coal mining machine in advance macroscopically by introducing a VR navigation dynamic pre-planning technology; and meanwhile, a real-time coal rock identification technology is combined, and the characteristics of hardness, structure, position and the like of the coal rock are identified through an online sensor and an algorithm, so that accurate data of a micro-level are obtained. According to the fusion of the two, coal rock information sensed in real time is combined with a virtual pre-planning result through a closed-loop feedback control mechanism, and the cutting strategy of the coal mining machine is dynamically adjusted. The core of the method is that through deep combination of a virtual layer and a physical layer, real-time operation of physical equipment is guided through virtual pre-planning, virtual planning is optimized in turn through physical feedback, closed-loop control is achieved, and then cutting planning is dynamically adjusted.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent coal mining, and in particular to a closed-loop cutting planning method driven by the fusion of VR navigation dynamic pre-planning and coal-rock identification online planning. Background Art

[0002] With the continuous advancement of coal mining technology, the intelligence level of fully mechanized mining equipment (such as fully mechanized mining equipment and coal mining machines) has been continuously improved. However, the existing cutting planning methods still face some challenges.

[0003] First, the coal mining environment is complex and changeable, including factors such as the distribution and thickness of coal seams and the hardness of rocks, which can significantly affect the cutting effect. In the existing technology, many devices rely on fixed cutting schemes, which often cannot adapt to the changes in coal and rock conditions during actual operations, resulting in low cutting efficiency and waste of resources.

[0004] In addition, existing coal and rock identification technologies mostly rely on fixed sensor data and manual inspection. These technologies have a slow response speed in real-time dynamic coal mine environments and are difficult to accurately capture changes in coal and rock properties. The real-time state of the coal seam has a significant impact on the cutting process, but traditional systems often cannot adjust the cutting path and parameters in a timely manner, which may lead to problems such as equipment overload, reduced cutting quality or waste of resources, thereby affecting the safety and efficiency of operations.

[0005] The introduction of virtual reality (VR) technology has provided new ideas for mining. By building a three-dimensional virtual model of the mine, preliminary cutting planning can be carried out in a simulated environment. However, the current VR pre-planning system is usually limited to static models and cannot reflect the actual situation in the mine in real time. This leads to a gap between the pre-planning results and the changes in actual operations, reducing the actual application effect of the system.

[0006] In order to improve the intelligence level of fully mechanized mining equipment in complex mine environments, there is an urgent need for a method that can combine VR navigation technology with a real-time coal and rock identification system to form a dynamically adjusted cutting planning scheme. This method should have the ability to dynamically adjust, respond to changes in coal seams and rocks in real time, optimize the cutting path, and ensure efficient operation of the equipment and maximize resource utilization. By achieving a seamless combination of dynamic pre-planning and online adjustment, the overall efficiency and safety of fully mechanized mining operations in coal mines can be significantly improved.

[0007] The invention patent with publication number CN117744604A provides a method and system for comprehensive mining planning and cutting of coal mines based on the Internet of Things and data visualization. The upper computer edits the process file of the coal machine and sends the process file to the lower computer. The lower computer parses the received process file and controls the coal machine to perform the first cut test run; the upper computer collects the operation data of the first cut test run, calculates the operation data through the planning and cutting method algorithm, and generates the planning and cutting method data; the planning and cutting method data is edited according to the actual situation; the upper computer sends the edited data to the lower computer, and the lower computer parses the edited data and controls the coal machine according to the analysis results. Beneficial effects of the present invention: It realizes the intelligent planning and optimization of the coal machine work, realizes the closed-loop control of the coal machine, improves the automation degree of the coal machine operation, realizes remote monitoring through the virtual interface, realizes the collection, analysis and utilization of data, improves the efficiency and output of the coal machine operation, and reduces safety hazards.

[0008] The invention patent with publication number CN116398134A provides a coal mining machine cutting planning method and equipment based on patrol visual perception. The method combines the spatial mapping relationship between the coordinate system of the mobile platform of the mining field and the geological coordinate system to perform digital construction of the mining field space and automatic detection and identification of the coal mining machine; performs perception analysis on the movement state of the coal mining machine, obtains the movement state information of the coal mining machine, and performs coal mining machine capture and tracking; performs state perception on the target objects of the coal mining machine components in the cutting space, and determines the optimal safe cutting curve based on the state perception data. Through the present invention, it is possible to complete a real-time cutting planning method by visually intelligently perceiving the equipment and environmental status of the cutting space, with strong data information timeliness and engineering implementation continuity; real-time visual interaction is performed on the associated equipment in the cutting space, and the surrounding environmental status perception results, so that the staff can perform cutting auxiliary intervention in real time and obtain more accurate cutting planning method data.

[0009] The invention patent with publication number CN113722979A provides a virtual-reality interactive system for coal mining machines based on digital twins and its construction method, which realizes the analysis of the real-time working status data of coal mining machines and the evaluation and prediction of the health status of equipment, and realizes the advanced simulation and optimization of the cutting action of coal mining machines based on twin models and iterative algorithms. The real-time status data collected during the working process of coal mining machines and the real-time prediction simulation results of twin models are used to perform real-time dynamic feedback control on coal mining machines, so as to realize efficient and autonomous cutting of coal mining machines. The real-time data of the working process of coal mining machines can be used to drive the dynamic simulation of twin models in real time. The twin models can be used to perceive the state signals that are not easy to directly measure from the physical entities of coal mining machines, which is beneficial to the measured data of coal mining machines and the perception data of twin models. The real-time health status of coal mining machines is evaluated and predicted through algorithms, and a coal mining machine health status evaluation result database is established to manage the health status of coal mining machines throughout their life cycle and perform predictive maintenance.

[0010] The above patent documents respectively realize the cutting planning of coal mining machines through the Internet of Things and data visualization, inspection visual perception and digital twin technology. These methods have achieved good results in improving the intelligence and efficiency of coal mining operations. However, they still have some shortcomings. First, although the cutting planning method based on the Internet of Things and data visualization can realize remote data monitoring and optimization operations, it has poor adaptability to real-time changing environments and lacks dynamic adjustment capabilities; although the inspection visual perception method realizes real-time perception of the coal mining machine and environmental status through visual technology, the visual system is easily disturbed by factors such as light and coal dust in complex coal mining environments, and the perception accuracy is limited; and although digital twin technology has highly accurate prediction and simulation capabilities, it relies heavily on real-time data, and the model update lags, which will affect the accuracy of decision-making. Overall, these methods still have problems such as poor dynamic adaptability, perception limitations and model feedback lag. Summary of the invention

[0011] The technical problem to be solved by the present invention is to provide a closed-loop cutting planning method driven by the fusion of VR navigation and coal and rock identification online planning, which combines virtual reality technology for dynamic environmental pre-planning, and obtains accurate coal and rock feature information through real-time coal and rock identification technology to realize dynamic cutting planning of the closed-loop control system, thereby improving the adaptability and real-time performance of the system.

[0012] In order to solve the above technical problems, the technical solution adopted by the present invention is: a closed-loop cutting planning method driven by the integration of VR navigation and coal and rock identification online planning, including the following modules:

[0013] (1) VR navigation pre-planning module;

[0014] First, historical data of coal mines are obtained and preprocessed, and a Transformer model is constructed. Real-time data or new input data is input into the Transformer model, and the Transformer model is deduced based on these inputs to predict future coal and rock conditions and changes in the top and bottom morphology of coal seams.

[0015] (2) Parameter construction and posture solution module;

[0016] The morphological parameters and physical parameter models of the coal seam are constructed through the Unity 3D platform, and the coal mining machine posture and cutting trajectory are solved; based on the above parameters, the coal mining machine posture is solved by solving the movement changes of the roof and floor to obtain the real-time posture of the coal mining machine; then, the coal mining machine's cutting trajectory is further calculated in combination with the posture data of the coal mining machine;

[0017] (3) Coal and rock identification module;

[0018] Combine ultrasonic sensors and laser induced breakdown spectroscopy (LIBS) technology to identify coal seams and rock strata in real time and obtain coal and rock characteristic parameters; after completing coal and rock identification, build a detailed model of the coal seam roof in Unity 3D to achieve reconstruction of the coal seam roof; based on the reconstructed roof, perform cutting simulation of the virtual coal mining machine in Unity 3D to correct the coal mining machine posture;

[0019] (4) Roller adjustment and dynamic speed regulation module;

[0020] The motion trajectory of the drum is optimized through the coal and rock feature data provided by the coal and rock recognition module and the posture data provided by the posture solution module;

[0021] (5) Path execution and feedback module;

[0022] According to the prediction and optimization results jointly generated by the VR navigation pre-planning module, parameter construction and posture solution module, coal rock identification module and drum adjustment and dynamic speed regulation module, the actual operation path of the coal mining machine is executed, and the operation data and operation feedback information of the coal mining machine are transmitted back to the system;

[0023] (6) Online prediction and decision-making module;

[0024] Based on the coal and rock data obtained from the coal and rock identification module, the coal and rock status is updated in real time through data processing and the prediction ability of the Transformer model, dynamic prediction is performed, and the data is fed back to the working path planning of the coal mining machine. In addition, according to the real-time coal and rock data, the parameters of the coal mining machine are dynamically adjusted to optimize the mining strategy.

[0025] Furthermore, the steps of constructing a VR navigation pre-planning module include:

[0026] Step 101, historical data collection and preprocessing: firstly, obtain the historical data of the coal mine, record the data in time series, then perform data preprocessing, and divide the data into a training set, a validation set and a test set in time sequence;

[0027] Step 102, data segmentation and feature extraction: after data preprocessing, the coal mine historical data is divided into blocks according to the mining location interval, and features are extracted for each block to construct a feature matrix, which is arranged in chronological order;

[0028] Step 103: Transformer model construction: Use the deep learning framework TensorFlow to build the Transformer model;

[0029] Step 104, model training and verification: In the Transformer model, the training set and the verification set are loaded to train the model. After the training is completed, a comprehensive evaluation is performed on the test set. The trained and verified model is used for coal rock status and coal seam roof and floor prediction.

[0030] Step 105, model verification and feedback: Apply the model to field operation data independent of the training process, evaluate whether the model runs stably during actual operation, analyze the robustness of the model under extreme conditions, and feed back deviations to the system to improve the model's algorithm or add new input features.

[0031] Furthermore, the steps of constructing the parameter construction and posture solving module include:

[0032] Step 201, morphological parameter construction: using the Unity 3D platform, simulating the actual geometric structure of the coal mining machine, and digitally reconstructing the top and bottom morphologies of the coal seam according to the morphological changes of the top and bottom morphologies of the coal seam;

[0033] Step 202, physical property parameter construction: obtain historical data of the coal mine including force, hardness, density and chemical element composition, and establish a physical property parameter library in Unity 3D based on the above historical data of the coal mine; integrate the physical property parameters with the geometric model constructed in step 201, and add a C# script to the coal mining machine to realize its operation and control; use the Transform component of Unity3D to dynamically change the X, Y, Z coordinates, as well as the pitch angle (RX), yaw angle (RY) and roll angle (RZ) of the coal mining machine model to accurately control the position and orientation of the virtual coal mining machine in three-dimensional space, so as to simulate its actual operation in the coal mine operating environment;

[0034] Step 203, coal mining machine posture calculation;

[0035] Step 204: Calculate cutting trajectory.

[0036] Furthermore, in step 202, during the construction process, virtual sensors are introduced to simulate the functions of actual sensors, collect data and feed back to the physical parameter model to obtain and monitor the status of coal seams and rock formations in real time. These virtual sensors automatically update data according to changes in physical parameters and provide real-time feedback.

[0037] Furthermore, step 203, the step of solving the coal mining machine posture comprises:

[0038] To construct a virtual scene of the coal mining face in the coal mining machine working face, it is necessary to first establish an absolute coordinate system O with the initial position of the coal mining machine as the origin. n -X n Y n Z n , where Xn The Y axis points to the direction of travel of the coal mining machine. n Vertically upward, Z n The axis points to the direction of the working face advancement. The coal mining machine posture solution is to calculate its position coordinates and travel angle in the virtual workspace by analyzing the axis encoder data to accurately describe the dynamic trajectory of the coal mining machine;

[0039] According to the shaft encoder data of the coal mining machine at the i-th position and the i+1-th position, the position information of the coal mining machine is calculated, and the specific position coordinates are calculated by formula (2-1):

[0040]

[0041] Among them, X i With Y i They represent the horizontal and vertical coordinates of the coal mining machine at the i-th position, s i is the shaft encoder reading at the i-th position of the coal mining machine, θ i is the pitch angle of the coal mining machine at the i-th position;

[0042] In order to accurately describe the rotational posture of the coal mining machine, quaternion is used to represent the posture of the coal mining machine in space. Suppose the posture quaternion of the coal mining machine at position i is q i , which is of the form:

[0043]

[0044] Among them, w is the real part of the quaternion, and x, y, and z are the imaginary components, which are used to represent the current rotation state of the coal mining machine;

[0045] The rotation increment of the coal mining machine between position i and i+1 is Δq, and the increments of pitch angle, yaw angle and roll angle (Δθ pitch , Δθ yaw , Δθ roll ) are converted into corresponding quaternion increments:

[0046] Δq=q pitch ·q yaw ·q roll

[0047] in:

[0048]

[0049] Through the quaternion update formula, the increment is applied to the original posture to obtain the posture quaternion q of the next posture i+1 i+1 :

[0050] q i+1 =q i ·Δq

[0051] The position information and attitude parameters obtained through posture solution provide the necessary initial posture reference for the roller adjustment and dynamic speed regulation modules, and also provide real-time posture data for the path execution and feedback modules.

[0052] Furthermore, step 204, the step of calculating the cutting trajectory includes:

[0053] According to the mining height information and the coal mining machine structure (such as Figure 4 The rocker arm inclination angle is calculated as shown in formula (2-2).

[0054]

[0055] in and are the cutting heights of the front and rear drums when the coal mining machine is at the i-th position; h is the distance from the rocker arm pin to the bottom plate of the scraper conveyor; D is the drum diameter; L1 is the length of the rocker arm;

[0056] Then, based on the corrected trajectory of the coal mining machine, the cutting trajectory is solved using the calculated rocker arm inclination angle, and the coordinates of the upper drum cutting point and the lower drum cutting point are calculated according to equations (2-3) and (2-4):

[0057]

[0058] Where L is the distance from the front rocker arm connecting pin to the rear rocker arm connecting pin, θ i is the pitch angle of the coal mining machine at the i-th position.

[0059] Furthermore, it is characterized in that: the steps of constructing the coal-rock identification module include:

[0060] Step 301, equipment initialization and calibration: before the coal-rock identification module starts, first initialize and calibrate the ultrasonic sensor and LIBS sensor;

[0061] Step 302, ultrasonic sensor detection: The ultrasonic sensor detects the coal seam and the rock layer. The ultrasonic sensor determines the physical properties of the coal and the rock by emitting high-frequency sound waves and receiving echo signals reflected from the coal and the rock.

[0062] Step 303, LIBS detection: After the ultrasonic sensor completes the preliminary coal rock density judgment, the system starts the LIBS sensor to perform a more accurate chemical composition analysis of the coal rock;

[0063] Step 304, coal seam roof reconstruction: after completing coal rock identification, the system will reconstruct a fine model of the coal seam roof through voxelization and Marching Cubes algorithm in Unity 3D;

[0064] Step 305, coal mining machine posture correction: Based on the reconstructed roof obtained in step 304, the cutting simulation of the virtual coal mining machine is carried out in Unity3d, and the front drum of the virtual coal mining machine is fitted to the coal seam roof to obtain a new drum cutting control strategy, which can correct the coal mining machine posture.

[0065] Furthermore, the steps of constructing the drum adjustment and dynamic speed regulation module include:

[0066] Step 401, coal rock analysis and preliminary judgment: the system receives coal rock feature data and parameter construction from the coal rock recognition module and posture data from the posture solution module through the MQTT protocol, first analyzes the coal rock features, calculates the physical properties such as hardness and density of the current cutting area, and serves as a basis for preliminary judgment of the coal rock structure; at the same time, the system judges the shape of the top and bottom plates of the coal seam based on the posture data, identifies whether there is a height difference or tilt change, and preliminarily determines whether the height and angle of the drum need to be adjusted, so as to further optimize the drum motion trajectory in subsequent steps;

[0067] Step 402, adjusting the drum height: after the coal-rock analysis and preliminary judgment are completed, the system calculates the ideal cutting height based on the characteristic information of the height difference between the top and bottom plates of the coal seam, the hardness and density of the coal and rock, and uses it as the target height for the drum adjustment; the drum height is adjusted by changing the extension of the hydraulic cylinder of the coal mining machine to meet the needs of different cutting layers and coal and rock characteristics;

[0068] Step 403, adjusting the speed of the coal mining machine: the system uses the MQTT protocol to receive sensor data and obtain real-time drum load and speed information, and dynamically adjusts the traveling speed of the coal mining machine by analyzing these data;

[0069] Step 404, feedback and optimization: Real-time feedback of the coal mining machine's operating data is used to adapt to the changing characteristics of the coal and rock layers; the optimization module compares the actual state with the target state, and further optimizes the speed and cutting depth through the reinforcement learning algorithm.

[0070] Furthermore, the steps of constructing a path execution and feedback module include:

[0071] Step 501, path execution: according to the optimized cutting strategy and travel path, the system sends instructions to the coal mining machine, and the coal mining machine starts actual operation according to the instructions. During the operation, the system monitors the key data of the coal mining machine in real time through sensors;

[0072] Step 502, feedback and data analysis: During the operation of the coal mining machine, the system will continuously collect and provide real-time feedback of key data from the sensors. The system will compare these feedback data with the optimized instructions, detect any deviations or anomalies in the operation, and immediately update the operation status;

[0073] Step 503, memory cutting and path reproduction: The system establishes a "demonstration knife" as a benchmark in the demonstration operation. When actually performing memory cutting, the system calls the data recorded by the demonstration knife, compares the parameters of the current operation and the demonstration knife in real time, minimizes the deviation, and reproduces the motion trajectory and operation mode of the demonstration knife. The reference data is transmitted to the actuator through the MQTT protocol. The system can dynamically adjust the height and angle of the roller and the propulsion speed to ensure consistency with the demonstration knife.

[0074] Furthermore, the steps of building an online prediction and decision module include:

[0075] Step 601, data processing and status update: After the system obtains real-time coal and rock data, it first performs data preprocessing, such as noise filtering and outlier processing; the processed data is input into the trained Transformer model in step 104, and the coal mining status is updated to ensure that the model prediction can reflect the latest coal and rock conditions;

[0076] Step 602, Transformer model dynamic prediction: After data processing and status update, the system calls the Transformer model for real-time prediction. By continuously inputting real-time sensor data and prediction results into the Transformer model, the system makes a detailed prediction of future coal and rock changes. Through continuous iterative prediction, the system can gradually optimize the path and cutting strategy;

[0077] Step 603, cutting strategy optimization: Based on the dynamic prediction results provided by the Transformer model, the cutting strategy of the coal mining machine is optimized by combining real-time data and prediction information. According to the changing trend of the coal-rock structure, the cutting depth, cutting path and travel speed of the coal mining machine are adjusted to ensure that the equipment always remains in the best working state. The optimal cutting parameters are determined by analyzing key factors including the hardness and thickness of the coal seam and the coal-rock boundary.

[0078] Step 604, dynamic adjustment and feedback: After the cutting strategy is optimized, the system dynamically adjusts the operating parameters of the coal mining machine according to real-time feedback, continuously monitors equipment operation and coal and rock changes through sensor data, and adjusts the cutting depth, path and speed in real time to ensure that the coal mining machine operates in the best condition. The feedback loop is used to analyze the difference between the actual effect and the optimization strategy and make adjustments quickly.

[0079] The present invention introduces virtual reality (VR) navigation dynamic pre-planning technology to plan the cutting path of the coal mining machine in advance at the macro level; at the same time, combined with real-time coal and rock identification technology, the hardness, structure, position and other characteristics of coal and rock are identified through online sensors and algorithms to obtain accurate data at the micro level. The fusion of the two combines the real-time perceived coal and rock information with the virtual pre-planning results through a closed-loop feedback control mechanism to dynamically adjust the cutting strategy of the coal mining machine. The core of the method of the present invention is to use virtual pre-planning to guide the real-time operation of physical equipment through the deep integration of the virtual and physical layers, and the physical feedback in turn optimizes the virtual planning, thereby realizing closed-loop control and dynamically adjusting the cutting plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 This is the overall framework diagram of the closed-loop cutting planning method of the present invention;

[0081] Figure 2 This is a flow chart of coal rock identification of the present invention;

[0082] Figure 3 This is a navigation map flow chart of the present invention;

[0083] Figure 4 It is the structural diagram of the coal mining machine of the present invention. DETAILED DESCRIPTION

[0084] The present invention provides a closed-loop cutting planning method driven by the fusion of VR navigation and coal-rock identification online planning, which comprises the following six modules: (1) VR navigation pre-planning module; (2) parameter construction and posture solution module; (3) coal-rock identification module; (4) drum adjustment and dynamic speed regulation module; (5) path execution and feedback module; and (6) online prediction and decision-making module.

[0085] Based on the above six modules, the overall concept of the present invention is: the VR navigation pre-planning module provides the future state of coal and rock and the morphological changes of the top and bottom plates of the coal seam for the parameter construction and posture solution module; the parameter construction and posture solution module provides the posture parameters and cutting trajectory parameters for the roller adjustment and dynamic speed regulation module, and also provides the physical property parameters for the coal and rock identification module through the virtual sensor; the coal and rock identification module detects the distribution of coal and rock based on ultrasonic and LIBS sensors, corrects the coal mining machine posture, and feeds back to the roller adjustment and dynamic speed regulation module; the roller adjustment and dynamic speed regulation module receives the posture and trajectory parameters, and optimizes the roller work in real time in combination with the feedback information of the coal and rock identification module. The first four modules jointly generate the pre-planning plan and pass it to the path execution and feedback module, which returns the feedback data of the equipment operation to each module to optimize the entire system; the online prediction and decision module combines the roller adjustment and dynamic speed regulation module, the coal and rock identification module and the path execution and feedback module, and performs online prediction and dynamic decision adjustment on the feedback data to achieve working state and path optimization.

[0086] In order to make those skilled in the art better understand the present invention, the present invention is further clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the features in the embodiments and examples of this application can be combined with each other without conflict.

[0087] 1. VR navigation pre-planning module

[0088] VR navigation pre-planning is an advance planning method based on virtual reality technology. It integrates historical data and real-time data, and uses the Transformer model to pre-simulate the changes in coal rock status and coal seam roof and floor morphology, providing dynamic navigation maps and prediction data support for subsequent modules.

[0089] Step 101, historical data collection and preprocessing: First, obtain the historical data of the coal mine, including the hardness, density, structure and other characteristics of the coal seam, record the data in time series, then perform data denoising, normalization, and divide the data into training set, validation set and test set in time order, where the training set is used to train the model, the validation set is used to verify the performance of the model, and the test set is used for the final test. Finally, write the data into an XML file, read the data through Unity 3D code, and be able to read specific types of data for subsequent model processing and pre-planning analysis.

[0090] Step 102, data segmentation and feature extraction: After data preprocessing, the historical data of the coal mine is divided into blocks according to the mining location interval, with each block being 10 meters, so as to capture the local feature changes of the coal seam section by section and enhance the coherence of the time series data. For each block, feature extraction is performed in terms of physical properties (such as hardness, density, coal seam thickness), chemical composition (such as sulfur content, carbon content), and structural characteristics (such as coal rock bedding, thickness of interlayer gangue), etc. In the process of feature extraction, the accuracy of sensor data such as hardness and density is optimized through signal denoising, and then the physical property data is standardized to construct a feature matrix so as to uniformly analyze the characteristics of each coal seam. Subsequently, the feature selection algorithm is used to screen out the features that are important for prediction and reduce the interference of redundant information. Finally, these feature matrices are arranged in chronological order to provide clear and time-consistent input features for the subsequent prediction of coal rock state and coal seam roof and floor morphology.

[0091] Step 103, Transformer model construction: Use the deep learning framework TensorFlow to build a Transformer model to achieve accurate prediction of coal rock status and coal rock roof and floor.

[0092] First, by importing TensorFlow and Keras libraries, basic support is provided for model building. When building the initial structure, the core parameters of the model are defined, including the embedding dimension, the number of attention heads, and the number of hidden layer units of the feedforward network. The input layer of the model uses the Embedding layer to embed the original data and convert it into a high-dimensional feature representation, which makes it easier for the model to capture complex sequence information. Then, the multi-head self-attention mechanism is introduced in the encoder part. This mechanism enables the model to focus on the features at different positions in the input sequence through multi-head parallel processing. Each layer of the encoder consists of a multi-head self-attention layer, a feedforward neural network (Feed-Forward Network), and a residual connection. In each layer of the encoder, the self-attention operation is first used to make the model pay attention to the dependencies between features; then the output is passed through the feedforward neural network to further extract and abstract important features. In order to maintain the stability and efficiency of the model, layer normalization is introduced after each layer operation, and residual connections are used to ensure gradient stability and accelerate model convergence. The final layer of the model is the output layer, which is used to generate prediction results.

[0093] When compiling the model, the Adam optimizer and MSE loss function are used to continuously optimize the model. The Adam optimizer is responsible for updating the model weights during the model training process. By continuously adjusting the weights, the optimizer gradually makes the model approach the optimal solution. It determines how the model adjusts parameters in each training iteration. The MSE loss function is responsible for measuring the difference between the model prediction result and the true value. The smaller the loss function, the higher the prediction accuracy of the model. After the model is built, save the model weights and structure for subsequent training, verification and evaluation.

[0094] Step 104, model training and verification: First, by loading the collected and preprocessed coal and rock data, including training sets and verification sets, ensure that the input data is aligned with the predicted label, and import the model in batches according to the specified batch size for efficient training. The core of model training is iterative optimization through forward propagation and back propagation. After the training starts, the input data is passed to the model, and the input data will be processed in sequence through the embedding layer, the multi-head self-attention layer, the encoder's self-attention module, and the feedforward network to extract important features from the sequence data layer by layer, and finally generate the predicted value through the output layer. Then, the error between the model prediction value and the true label is measured by the MSE mean square error loss function, and the error is propagated back layer by layer through the back propagation algorithm.

[0095] The model is optimized by the Adam optimizer, which adjusts the weights of each layer to gradually approach the minimum value of the loss function to improve the prediction accuracy of the model. At the end of each training cycle, the validation set data is input into the model to evaluate the performance of the model on unseen data, and the error of the validation set is used to detect whether the model is overfitting. If the validation error continues to decrease, it means that the model has good generalization ability. If the validation error of the model starts to rise, that is, overfitting occurs, the EarlyStopping callback function can be used to implement the early stopping strategy and stop training to prevent the model from continuing to overfit.

[0096] At the same time, the learning rate scheduler is used to dynamically adjust the learning rate to ensure the smooth convergence of the model at different training stages. After the training is completed, a comprehensive evaluation is conducted on the test set, and indicators such as MSE and MAE are comprehensively considered to confirm the prediction accuracy and stability of the model. The model that has completed the training verification will be used for the prediction of coal rock status and coal seam roof and floor. At this time, real-time data or new input data will be input into the model. The model will be deduced based on these inputs to predict the future coal rock status and coal seam roof and floor morphological changes, thereby providing a basis for subsequent parameter construction and posture solution modules.

[0097] Step 105, model verification and feedback: After model training and verification, model verification and feedback are mainly used to evaluate the performance of the model in the actual application environment and provide optimization suggestions.

[0098] In this step, the model is applied to field operation data independent of the training process, and its prediction accuracy and adaptability to coal and rock properties are observed through field testing. The focus of the verification is to evaluate whether the model operates stably during actual operations and analyze the robustness of the model under extreme conditions.

[0099] In addition, deviation data from actual operations is collected and fed back to the system to improve the model algorithm or add new input features to better match on-site needs. Ultimately, the results of verification and feedback are used to confirm the production feasibility of the model, providing more reliable data support for closed-loop path execution and feedback, thus forming a complete dynamic optimization closed-loop system.

[0100] 2. Parameter construction and pose calculation module

[0101] The parameter construction and posture solution module is responsible for constructing the morphological parameters and physical parameter models of the coal seam, and solving the coal mining machine posture and cutting trajectory. First, the morphological parameters (such as the height and structure of the roof and floor) and physical parameters (including hardness, density, chemical element distribution, etc.) of the coal seam are generated through the Unity 3D platform to accurately reflect the physical characteristics of the coal seam. Then, based on these parameters, the coal mining machine posture solution is performed by solving the movement changes of the roof and floor to obtain the real-time posture of the coal mining machine. Finally, combined with the posture data, the cutting trajectory of the coal mining machine is further calculated to support subsequent path planning and dynamic adjustment.

[0102] Step 201, morphological parameter construction: This part mainly involves accurate modeling of the geometric morphology of the coal mining machine and the roof and floor of the coal seam. Through the Unity 3D platform, the actual geometric structure of the coal mining machine is simulated, and the roof and floor morphology of the coal seam is digitally reconstructed according to the changes in the morphology of the coal seam roof and floor, including information such as the height, undulating characteristics of the roof, and the structure of the floor.

[0103] The system uses the collected real-time coal-rock data and the future coal-rock state predicted by the Transformer model and the changes in the morphology of the coal seam roof and floor as basic data input. In order to ensure the accuracy and consistency of the data, these basic data will undergo data smoothing, noise reduction, interpolation and other processing steps to eliminate the noise in the collection and the inconsistency between the data. Based on these data, the coal mining machine and the three-dimensional navigation map containing the geometry of the coal mine working surface and the distribution of coal and rock, that is, the coal seam roof and floor, are modeled in UG, and the required pins are repaired. The model is then saved in STL format and imported into 3D Max, and the model is exported in Fbx format in 3D Max, and the Fbx format model is imported into Unity 3d. The relevant parameters of the model are preliminarily set in the virtual scene, and the parent-child hierarchical relationship is established to ensure the coordinated linkage of different objects in the virtual scene.

[0104] Step 202, physical property parameter construction: First, obtain historical data of coal mines, including stress, hardness, density and chemical element composition, and perform smoothing, noise reduction and interpolation processing on these data to ensure the accuracy and consistency of the data. Next, create a new project in Unity 3D and establish a physical property parameter library. Each physical property parameter will be managed as an independent script or component. In this process, C# scripts are used to define the properties of each parameter, including value, unit and range of variation, to facilitate dynamic adjustment and precise control in simulation.

[0105] During the construction process, virtual sensors are introduced to obtain and monitor the status of coal seams and rock formations in real time. Virtual sensors can simulate the functions of actual sensors, collect data and feed back to the physical parameter model, thereby enhancing the accuracy and real-time performance of the model. These virtual sensors can automatically update data based on changes in physical parameters and provide real-time feedback.

[0106] In Unity 3D, a dark black high-gloss material is used for the coal seam to reflect its high carbon content and smooth characteristics. A light gray rough material is used for the rock layer to show its hard and porous characteristics, which can realistically simulate the physical structure of the rock. The dynamic deformation effect is used to show the deformation of the coal seam under stress, so as to intuitively reflect the response of the coal seam during the mining process.

[0107] In addition, an interactive function is added to the physical parameter model, so that users can adjust the parameters in real time by clicking or sliding, and observe the impact of these changes on the coal seam characteristics and simulation results. After that, the physical parameters are integrated with the geometric model constructed in step 201 (such as the coal mining machine and the coal seam roof and floor), and a C# script is added to the coal mining machine to realize its operation and control. By using the Transform component of Unity3D, the X, Y, and Z coordinates of the coal mining machine model, as well as the pitch angle (RX), yaw angle (RY), and roll angle (RZ) are dynamically changed to accurately control the position and orientation of the virtual coal mining machine in three-dimensional space, thereby simulating its actual operation in the coal mine operating environment.

[0108] Step 203, coal mining machine posture solution: To construct a virtual scene of the coal mining working face in the coal mining machine working face, it is first necessary to establish an absolute coordinate system with the initial position of the coal mining machine as the origin. n -X n Y n Z n Among them, X n The Y axis points to the direction of travel of the coal mining machine. n Vertically upward, Z n The axis points to the advancing direction of the working face. The coal mining machine's posture solution is to calculate its position coordinates and travel angle in the virtual workspace by analyzing the axis encoder data to accurately describe the dynamic trajectory of the coal mining machine.

[0109] According to the shaft encoder data of the coal mining machine at the i-th position and the i+1-th position, the position information of the coal mining machine can be inferred. The specific position coordinates are calculated by formula (2-1):

[0110]

[0111] Where X i With Y i They represent the horizontal and vertical coordinates of the coal mining machine at the i-th position, s iis the shaft encoder reading at the i-th position of the coal mining machine, θ i is the pitch angle of the coal mining machine at the i-th position.

[0112] In order to accurately describe the rotational posture of the coal mining machine, quaternion is used to represent the posture of the coal mining machine in space. Suppose the quaternion of the coal mining machine at position i is q i , which is of the form:

[0113]

[0114] Among them, w is the real part of the quaternion, and x, y, and z are the imaginary components, which are used to indicate the current rotation state of the coal mining machine.

[0115] The rotation increment of the coal mining machine between position i and i+1 is Δq, and the increments of pitch angle, yaw angle and roll angle (Δθ pitch , Δθ yaw , Δθ roll ) are converted into corresponding quaternion increments:

[0116] Δq=q pitch ·q yaw ·q roll

[0117] in:

[0118]

[0119]

[0120] Through the quaternion update formula, the increment is applied to the original posture to obtain the posture quaternion q of the next posture i+1 i+1 :

[0121] q i+1 =q i ·Δq

[0122] The position information and attitude parameters obtained through posture solution provide the necessary initial posture reference for the roller adjustment and dynamic speed regulation modules, and also provide real-time posture data for the path execution and feedback modules.

[0123] Step 204, calculation of cutting trajectory: the shearer moves along the scraper conveyor, and the drum cuts the coal wall to form the coal seam roof and floor. The shearer cutting trajectory reflects the change trend of the coal seam floor. Figure 4 The rocker arm inclination angle is solved as shown in formula (2-2).

[0124]

[0125] in and are the cutting heights of the front and rear drums when the coal mining machine is at the i-th position; h is the distance from the rocker arm pin to the bottom plate of the scraper conveyor. Since the coal seam is relatively flat, the influence of the rotation angle of the support shoe relative to the machine body is ignored, so this value is a constant; D is the drum diameter; L1 is the length of the rocker arm.

[0126] Then, based on the corrected trajectory of the coal mining machine, the cutting trajectory is solved using the calculated rocker arm inclination angle, and the coordinates of the upper drum cutting point and the lower drum cutting point are calculated according to equations (2-3) and (2-4):

[0127]

[0128] Where L is the distance from the front rocker arm connecting pin to the rear rocker arm connecting pin, θ i is the pitch angle of the coal mining machine at the i-th position. These coordinate data provide accurate trajectory reference for the drum adjustment and dynamic speed regulation module, and also provide position information support for the path execution and feedback module.

[0129] 3. Coal and rock identification module

[0130] The coal-rock identification module is responsible for real-time identification of coal seams and rock strata through sensor technology, helping the coal mining machine to accurately determine the geological conditions of the working surface. In this module, ultrasonic sensors are used to detect the physical properties of coal and rock, such as hardness and density, to help distinguish between coal seams and rock strata; while laser-induced breakdown spectroscopy (LIBS) technology identifies the chemical composition of coal and rock by analyzing the spectral information on the surface of coal and rock. The two technologies work together to enable the system to conduct comprehensive analysis from the physical and chemical levels to ensure the accuracy of identification.

[0131] Step 301, equipment initialization and calibration: Before the coal-rock identification module starts, the ultrasonic sensor and LIBS sensor are first initialized and calibrated to ensure that the equipment can accurately detect the physical and chemical properties of coal and rock formations. The calibration of the ultrasonic sensor mainly involves the adjustment of wave velocity, emission frequency and sensitivity to ensure that the propagation characteristics in different media can be accurately measured. At the same time, the LIBS sensor needs to set the laser power and spectral detection range to ensure that it can accurately identify the elemental composition of coal and rock formations. During the calibration process, the system will read the known characteristics of the reference sample and adjust the sensor parameters by comparing the deviation between the test results and the standard values ​​until the expected accuracy is achieved.

[0132] Step 302, ultrasonic sensor detection: After the equipment calibration is completed, the ultrasonic sensor begins to detect coal seams and rock formations. Ultrasonic sensors determine their physical properties by emitting high-frequency sound waves and receiving echo signals reflected from different media (such as coal and rock). Specifically, the propagation speed of ultrasonic waves in materials of different densities and hardnesses is different, and the time delay, attenuation and waveform changes of echoes can reveal information such as the density, thickness and internal structure of the material. The system will collect these signals in real time and improve the accuracy of the data through processing steps such as data filtering, noise reduction and signal enhancement. Then, the system evaluates the density based on the measured sound wave propagation speed and density model, and compares it with the system's built-in coal and rock characteristics database to determine whether the density is lower than the preset density cutoff point. If the measured density is low, the system will preliminarily determine that the current detection area is a coal seam; if the density is high, it will be determined as a rock formation. This preliminary judgment result provides a basic basis for the subsequent LIBS detection.

[0133] Step 303, LIBS detection: After the ultrasonic sensor completes the preliminary judgment of the coal rock density, the system starts the laser induced breakdown spectroscopy (LIBS) sensor to perform a more accurate chemical composition analysis of the coal rock. LIBS irradiates the surface of the coal rock with high-energy laser pulses to excite plasma, and then analyzes the spectrum emitted by the plasma through a spectrometer to extract elemental composition information. For areas initially judged to be coal seams, if the LIBS test shows a high carbon content, the area will be finally determined to be a coal seam; and for areas initially judged to be rock formations, if low carbon content and high content of typical rock formation elements such as silicon or aluminum are detected, the area is confirmed to be a rock formation.

[0134] For test results that are inconsistent with the initial judgment, such as the initial judgment of coal seams with low carbon content and high rock element content, or the initial judgment of rock layers with high carbon content and low rock element content, the system will mark these results as abnormal data and store them in the coal and rock characteristics database. These abnormal data will be analyzed by machine learning algorithms in the future to optimize the coal and rock identification model, ensure the detection accuracy and robustness of the identification process, and gradually improve the system's adaptability to complex changes in coal and rock characteristics.

[0135] Step 304, coal seam roof reconstruction: After completing the coal rock identification, the system will construct a detailed model of the coal seam roof through voxelization and Marching Cubes algorithm in Unity 3D.

[0136] The coal-rock data is voxelized to enhance the spatial resolution of the model, and the voxel grid is triangulated in combination with the Marching Cubes algorithm to generate a smooth and natural grid structure to ensure that the coal seam surface highly restores the true geological characteristics. On this basis, the height map and Perlin Noise function are further applied to introduce detailed bumps and depressions to the coal seam surface to simulate the irregularity and thickness differences of the rock layer. The system will also reconstruct the rock layer part in the grid based on the rock layer thickness information determined in the preliminary analysis, appropriately reduce the height of the rock layer part, so as to achieve the separation of the coal seam and the rock layer, and finally obtain a complete coal-rock boundary.

[0137] During the modeling process, the system will also use the Unity 3D physics engine to enhance the physical interaction between the coal seam roof and the fully mechanized mining equipment. Specifically, a Rigidbody component is added to the fully mechanized mining equipment model to make it affected by physical factors such as gravity and inertia, and a Mesh Collider component is added to the equipment and coal seam to simulate the weight and motion feedback of the equipment on the coal seam roof.

[0138] Step 305, coal mining machine posture correction: Based on the reconstructed roof obtained in step 304, the cutting simulation of the virtual coal mining machine is carried out in Unity3d, and the front drum of the virtual coal mining machine is fitted to the coal seam roof to obtain a new drum cutting control strategy, which can correct the coal mining machine posture.

[0139] 4. Roller adjustment and dynamic speed regulation module

[0140] The drum adjustment and dynamic speed control module is mainly responsible for real-time adjustment of the drum angle and height of the coal mining machine, and dynamically adjusts the speed of the coal mining machine to adapt to changes in the coal and rock layers. This module optimizes the movement trajectory of the drum through the coal and rock feature data provided by the coal and rock recognition module and the posture parameters provided by the posture solution module to achieve more accurate cutting control. At the same time, the system combines the changes in the state of the roof and floor of the coal seam and the physical parameters to generate a drum adjustment strategy and implement dynamic speed control.

[0141] Step 401, coal rock analysis and preliminary judgment: The system receives coal rock feature data and parameter construction from the coal rock identification module and posture data from the posture solution module through the MQTT protocol, including hardness, density and real-time morphological parameters of the coal seam roof and floor. The system first analyzes these coal rock features and calculates the physical properties such as hardness and density of the current cutting area as a basis for preliminary judgment of the coal rock structure. At the same time, the system judges the morphology of the coal seam roof and floor based on the posture data, identifies whether there is a height difference or tilt change, and preliminarily determines whether the height and angle of the roller need to be adjusted, so as to further optimize the roller motion trajectory in subsequent steps.

[0142] Step 402, roller height adjustment: After the coal-rock analysis and preliminary judgment are completed, the system calculates the ideal cutting height based on the height difference between the top and bottom plates of the coal seam, the hardness of the coal rock, the density and other characteristic information, and uses it as the target height of the roller adjustment. The adjustment of the roller height is achieved by changing the extension of the hydraulic cylinder of the coal mining machine to adapt to the needs of different cutting layers and the characteristics of coal and rock. In the specific implementation, the system compares the target height with the real-time roller height, receives the roller height data obtained by the posture solution at a frequency of 10Hz per second through the MQTT protocol, ensures the continuity and real-time feedback of the roller height, and transmits it to the PID controller. The PID controller calculates and feedbacks the error value between the roller and the target height in real time during the adjustment process. The control process consists of three parts: proportional (P), integral (I) and differential (D): proportional control adjusts the roller position according to the instantaneous error size to quickly approach the target height; integral control is used to eliminate long-term deviations to ensure that the roller is stable near the target height; differential control is smoothed according to the error change rate to avoid excessive or insufficient displacement. Under this control framework, the PID controller makes feedback adjustments based on the error size and sends the control signal to the hydraulic cylinder actuator through the MQTT protocol, dynamically adjusting the extension of the hydraulic cylinder to accurately adjust the roller height.

[0143] Step 403, speed adjustment of the coal mining machine: The system uses the MQTT protocol to receive sensor data at a frequency of 10Hz and obtain real-time drum load and speed information, and dynamically adjusts the travel speed of the coal mining machine by analyzing these data. To this end, the system adopts an adaptive speed control method to determine whether to accelerate or decelerate according to the real-time changes in the drum load and physical parameters such as coal rock hardness. Specifically, when the load increases or encounters a coal rock layer with higher hardness, the speed is appropriately reduced to avoid equipment overload; otherwise, the speed is increased to improve work efficiency. The control signal is then transmitted to the coal mining machine hydraulic drive system through the MQTT protocol to achieve fine speed control to ensure the stability and cutting efficiency of the system.

[0144] Step 404, feedback and optimization: This step dynamically optimizes the control parameters by real-time feedback of the coal mining machine's operating data, such as speed, drum height and angle, to adapt to the changing characteristics of the coal and rock layers. The optimization module compares the actual state with the target state, and further optimizes the speed and cutting depth through the reinforcement learning algorithm to improve the coal mining efficiency and equipment stability. After combining the drum adjustment, speed control and coal and rock identification analysis of the first three modules, the system generates an optimized pre-planning plan and transmits it to the path execution and feedback module for real-time path control and decision-making.

[0145] 5. Path execution and feedback module

[0146] The path execution and feedback module is mainly responsible for executing the actual operation path of the coal mining machine based on the previous prediction and optimization results, and at the same time transmitting the operation data and operation feedback information of the coal mining machine back to the system for further optimization and adjustment.

[0147] Step 501, path execution: According to the optimized cutting strategy and travel path, the system sends instructions to the coal mining machine, and the coal mining machine starts the actual operation according to the instructions. During the execution process, the coal mining machine will strictly follow the path, cutting depth and speed parameters optimized by the system to operate. At the same time, during the operation process, the system monitors the key data of the coal mining machine in real time through sensors, such as travel speed, position, cutting depth, equipment load, etc.

[0148] Step 502, feedback and data analysis: During the operation of the coal mining machine, the system will continuously collect and provide real-time feedback of key data from sensors, such as travel speed, equipment position, cutting depth, etc. The system will compare these feedback data with the optimized instructions, detect any deviations or anomalies in the operation, and immediately update the operation status to ensure that the system can quickly adjust the path, speed or cutting depth.

[0149] Step 503, memory cutting and path reproduction: The system will establish a "demonstration knife" as a benchmark in the demonstration operation. The role of the demonstration knife is to record the key cutting parameters of the coal mining machine when it passes through the middle slot of each scraper conveyor, and store these parameters in XML format. These key parameters include the height and angle of the drum, the propulsion speed and position of the coal mining machine, etc.

[0150] When actually performing memory cutting, the system will call the data recorded by the demonstration knife, compare the parameters of the current operation with the demonstration knife in real time, minimize the deviation, and reproduce the movement trajectory and operation mode of the demonstration knife. These reference data are transmitted to the actuator through the MQTT protocol, and the system can dynamically adjust the height and angle of the roller and the propulsion speed to ensure consistency with the demonstration knife.

[0151] 6. Online prediction and decision-making module

[0152] The online prediction and decision-making module is responsible for ensuring that the operation path of the coal mining machine can adapt to the complex and changing coal and rock environment through real-time monitoring and dynamic adjustment. Based on the accurate coal and rock data obtained from the coal and rock identification module, this module updates the coal and rock status in real time through data processing and the prediction ability of the Transformer model, performs dynamic prediction, and feeds back to the working path planning of the coal mining machine. The module will also dynamically adjust the working path, speed, cutting depth and other parameters of the coal mining machine according to the real-time coal and rock data to optimize the mining strategy.

[0153] This module is closely related to the VR navigation pre-planning of the coal mining machine: VR navigation pre-planning sets the preliminary path and operation plan for the coal mining machine, while the online prediction and decision-making module continuously calibrates and optimizes the path based on real-time feedback in actual operation, thereby achieving more accurate and efficient coal mining operations.

[0154] Step 601, data processing and status update: After the system obtains real-time coal and rock data, it first performs data preprocessing, such as noise filtering and outlier processing. Then, the processed data is input into the trained Transformer model in step 104, and the coal mining status is updated to ensure that the model prediction can reflect the latest coal and rock conditions.

[0155] Step 602, Transformer model dynamic prediction: After data processing and status update, the system will call the Transformer model for real-time prediction. Unlike the Transformer model in the VR navigation pre-planning stage, which focuses on global path planning, the Transformer model here is based on real-time data in the actual operation process (such as coal and rock conditions, equipment operating parameters, etc.) for dynamic updates and short-term predictions. By continuously inputting real-time sensor data and prediction results into the Transformer model, the system can make detailed predictions on future coal and rock changes. In particular, during the operation process, the Transformer model will predict the changing trend of coal and rock structures, possible obstacles, and the working status of the coal mining machine. Through continuous iterative predictions, the system can gradually optimize the path and cutting strategy.

[0156] Step 603, cutting strategy optimization: Based on the dynamic prediction results provided by the Transformer model, the system will optimize the cutting strategy of the coal mining machine by combining real-time data and prediction information. The main goal of the optimization is to adjust the cutting depth, cutting path and travel speed of the coal mining machine according to the changing trend of the coal rock structure to ensure that the equipment always maintains the best working condition. The system will analyze key factors such as the hardness, thickness, and coal-rock boundary of the coal seam to determine the best cutting parameters, thereby improving coal mining efficiency and reducing equipment wear.

[0157] Step 604, dynamic adjustment and feedback: After the cutting strategy is optimized, the system dynamically adjusts the operating parameters of the coal mining machine based on real-time feedback. Through sensor data, the operation of the equipment and the changes in coal and rock are continuously monitored. The system adjusts the cutting depth, path and speed in real time to ensure that the coal mining machine operates in the best condition. The feedback loop is used to analyze the difference between the actual effect and the optimization strategy and make adjustments quickly.

Claims

1. A closed-loop cutting planning method driven by VR navigation and coal-rock identification online planning fusion, characterized in that: Includes the following modules: (1) VR navigation pre-planning module; First, historical data of coal mines are obtained and preprocessed, and a Transformer model is constructed. Real-time data or new input data is input into the Transformer model, and the Transformer model is deduced based on these inputs to predict future coal and rock conditions and changes in the top and bottom morphology of coal seams. (2) Parameter construction and posture solution module; The morphological parameters and physical parameter models of the coal seam are constructed through the Unity 3D platform, and the coal mining machine posture and cutting trajectory are solved; based on the above parameters, the coal mining machine posture is solved by solving the movement changes of the roof and floor to obtain the real-time posture of the coal mining machine; then, the coal mining machine's cutting trajectory is further calculated in combination with the posture data of the coal mining machine; (3) Coal and rock identification module; Combine ultrasonic sensors and laser induced breakdown spectroscopy (LIBS) technology to identify coal seams and rock strata in real time and obtain coal and rock characteristic parameters; after completing coal and rock identification, build a detailed model of the coal seam roof in Unity 3D to achieve reconstruction of the coal seam roof; based on the reconstructed roof, perform cutting simulation of the virtual coal mining machine in Unity 3D to correct the coal mining machine posture; (4) Roller adjustment and dynamic speed regulation module; The motion trajectory of the drum is optimized through the coal and rock feature data provided by the coal and rock recognition module and the posture data provided by the posture solution module; (5) Path execution and feedback module; According to the prediction and optimization results jointly generated by the VR navigation pre-planning module, parameter construction and posture solution module, coal rock identification module and drum adjustment and dynamic speed regulation module, the actual operation path of the coal mining machine is executed, and the operation data and operation feedback information of the coal mining machine are transmitted back to the system; (6) Online prediction and decision-making module; Based on the coal and rock data obtained from the coal and rock identification module, the coal and rock status is updated in real time through data processing and the prediction ability of the Transformer model, dynamic prediction is performed, and the data is fed back to the working path planning of the coal mining machine. In addition, according to the real-time coal and rock data, the parameters of the coal mining machine are dynamically adjusted to optimize the mining strategy.

2. The method according to claim 1, characterized in that: The steps to build a VR navigation pre-planning module include: Step 101, historical data collection and preprocessing: firstly, obtain the historical data of the coal mine, record the data in time series, then perform data preprocessing, and divide the data into a training set, a validation set and a test set in time sequence; Step 102, data segmentation and feature extraction: after data preprocessing, the coal mine historical data is divided into blocks according to the mining location interval, and features are extracted for each block to construct a feature matrix, which is arranged in chronological order; Step 103: Transformer model construction: Use the deep learning framework TensorFlow to build the Transformer model; Step 104, model training and verification: In the Transformer model, the training set and the verification set are loaded to train the model. After the training is completed, a comprehensive evaluation is performed on the test set. The trained and verified model is used for coal rock status and coal seam roof and floor prediction. Step 105, model verification and feedback: Apply the model to field operation data independent of the training process, evaluate whether the model runs stably during actual operation, analyze the robustness of the model under extreme conditions, and feed back deviations to the system to improve the model's algorithm or add new input features.

3. The method according to claim 1 or 2, characterized in that: The steps to construct the parameter construction and pose solving module include: Step 201, morphological parameter construction: using the Unity 3D platform, simulating the actual geometric structure of the coal mining machine, and digitally reconstructing the top and bottom morphologies of the coal seam according to the morphological changes of the top and bottom morphologies of the coal seam; Step 202, physical property parameter construction: obtain historical data of the coal mine including force, hardness, density and chemical element composition, and establish a physical property parameter library in Unity 3D based on the above historical data of the coal mine; integrate the physical property parameters with the geometric model constructed in step 201, and add a C# script to the coal mining machine to realize its operation and control; use the Transform component of Unity3D to dynamically change the X, Y, Z coordinates, as well as the pitch angle (RX), yaw angle (RY) and roll angle (RZ) of the coal mining machine model to accurately control the position and orientation of the virtual coal mining machine in three-dimensional space, so as to simulate its actual operation in the coal mine operating environment; Step 203, coal mining machine posture calculation; Step 204: Calculate cutting trajectory.

4. The method according to claim 3, characterized in that: In step 202, during the construction process, virtual sensors are introduced to simulate the functions of actual sensors, collect data and feed back to the physical parameter model to obtain and monitor the status of coal seams and rock formations in real time. These virtual sensors automatically update data according to changes in physical parameters and provide real-time feedback.

5. The method according to claim 3 or 4, characterized in that: Step 203, the step of coal mining machine posture solution includes: To construct a virtual scene of the coal mining face in the coal mining machine working face, it is necessary to first establish an absolute coordinate system O with the initial position of the coal mining machine as the origin. n -X n Y n Z n , where X n The Y axis points to the direction of travel of the coal mining machine. n Vertically upward, Z n The axis points to the direction of the working face advancement. The coal mining machine posture solution is to calculate its position coordinates and travel angle in the virtual workspace by analyzing the axis encoder data to accurately describe the dynamic trajectory of the coal mining machine; According to the shaft encoder data of the coal mining machine at the i-th position and the i+1-th position, the position information of the coal mining machine is calculated, and the specific position coordinates are calculated by formula (2-1): Among them, X i With Y i They represent the horizontal and vertical coordinates of the coal mining machine at the i-th position, s i is the shaft encoder reading at the i-th position of the coal mining machine, θ i is the pitch angle of the coal mining machine at the i-th position; In order to accurately describe the rotational posture of the coal mining machine, quaternion is used to represent the posture of the coal mining machine in space. Suppose the posture quaternion of the coal mining machine at position i is q i , which is of the form: Among them, w is the real part of the quaternion, and x, y, and z are the imaginary components, which are used to represent the current rotation state of the coal mining machine; The rotation increment of the coal mining machine between position i and i+1 is Δq, and the increments of pitch angle, yaw angle and roll angle (Δθ pitch , Δθ yaw , Δθ roll ) are converted into corresponding quaternion increments: Δq=q pitch ·q yaw ·q roll in: Through the quaternion update formula, the increment is applied to the original posture to obtain the posture quaternion q of the next posture i+1 i+1 : q i+1 =q i ·Δq The position information and attitude parameters obtained through posture solution provide the necessary initial posture reference for the roller adjustment and dynamic speed regulation modules, and also provide real-time posture data for the path execution and feedback modules.

6. The method according to claim 5, characterized in that: Step 204, the step of calculating the cutting trajectory includes: According to the mining height information and the coal mining machine structure (as shown in Figure 4), the rocker arm inclination angle is calculated as shown in formula (2-2): in and are the cutting heights of the front and rear drums when the coal mining machine is at the i-th position; h is the distance from the rocker arm pin to the bottom plate of the scraper conveyor; D is the drum diameter; L1 is the length of the rocker arm; Then, based on the corrected trajectory of the coal mining machine, the cutting trajectory is solved using the calculated rocker arm inclination angle, and the coordinates of the upper drum cutting point and the lower drum cutting point are calculated according to equations (2-3) and (2-4): Where L is the distance from the front rocker arm connecting pin to the rear rocker arm connecting pin, θ i is the pitch angle of the coal mining machine at the i-th position.

7. The method according to claim 6, characterized in that: The steps of building the coal-rock identification module include: Step 301, equipment initialization and calibration: before the coal-rock identification module starts, first initialize and calibrate the ultrasonic sensor and LIBS sensor; Step 302, ultrasonic sensor detection: The ultrasonic sensor detects the coal seam and the rock layer. The ultrasonic sensor determines the physical properties of the coal and the rock by emitting high-frequency sound waves and receiving echo signals reflected from the coal and the rock. Step 303, LIBS detection: After the ultrasonic sensor completes the preliminary coal rock density judgment, the system starts the LIBS sensor to perform a more accurate chemical composition analysis of the coal rock; Step 304, coal seam roof reconstruction: after completing coal rock identification, the system will reconstruct a fine model of the coal seam roof through voxelization and Marching Cubes algorithm in Unity 3D; Step 305, coal mining machine posture correction: Based on the reconstructed roof obtained in step 304, the cutting simulation of the virtual coal mining machine is carried out in Unity3d, and the front drum of the virtual coal mining machine is fitted to the coal seam roof to obtain a new drum cutting control strategy, which can correct the coal mining machine posture.

8. The method according to claim 7, characterized in that: The steps to build the drum adjustment and dynamic speed regulation module include: Step 401, coal rock analysis and preliminary judgment: the system receives coal rock feature data and parameter construction from the coal rock recognition module and posture data from the posture solution module through the MQTT protocol, first analyzes the coal rock features, calculates the physical properties such as hardness and density of the current cutting area, and serves as a basis for preliminary judgment of the coal rock structure; at the same time, the system judges the shape of the top and bottom plates of the coal seam based on the posture data, identifies whether there is a height difference or tilt change, and preliminarily determines whether the height and angle of the drum need to be adjusted, so as to further optimize the drum motion trajectory in subsequent steps; Step 402, adjusting the drum height: after the coal-rock analysis and preliminary judgment are completed, the system calculates the ideal cutting height based on the characteristic information of the height difference between the top and bottom plates of the coal seam, the hardness and density of the coal and rock, and uses it as the target height for the drum adjustment; the drum height is adjusted by changing the extension of the hydraulic cylinder of the coal mining machine to meet the needs of different cutting layers and coal and rock characteristics; Step 403, adjusting the speed of the coal mining machine: the system uses the MQTT protocol to receive sensor data and obtain real-time drum load and speed information, and dynamically adjusts the traveling speed of the coal mining machine by analyzing these data; Step 404, feedback and optimization: Real-time feedback of the coal mining machine's operating data is used to adapt to the changing characteristics of the coal and rock layers; the optimization module compares the actual state with the target state, and further optimizes the speed and cutting depth through the reinforcement learning algorithm.

9. The method according to claim 8, characterized in that: The steps to build a path execution and feedback module include: Step 501, path execution: according to the optimized cutting strategy and travel path, the system sends instructions to the coal mining machine, and the coal mining machine starts actual operation according to the instructions. During the operation, the system monitors the key data of the coal mining machine in real time through sensors; Step 502, feedback and data analysis: During the operation of the coal mining machine, the system will continuously collect and provide real-time feedback of key data from the sensors. The system will compare these feedback data with the optimized instructions, detect any deviations or anomalies in the operation, and immediately update the operation status; Step 503, memory cutting and path reproduction: The system establishes a "demonstration knife" as a benchmark in the demonstration operation. When actually performing memory cutting, the system calls the data recorded by the demonstration knife, compares the parameters of the current operation and the demonstration knife in real time, minimizes the deviation, and reproduces the motion trajectory and operation mode of the demonstration knife. The reference data is transmitted to the actuator through the MQTT protocol. The system can dynamically adjust the height and angle of the roller and the propulsion speed to ensure consistency with the demonstration knife.

10. The method according to claim 9, characterized in that: The steps to build an online prediction and decision module include: Step 601, data processing and status update: After the system obtains real-time coal and rock data, it first performs data preprocessing, such as noise filtering and outlier processing; the processed data is input into the trained Transformer model in step 104, and the coal mining status is updated to ensure that the model prediction can reflect the latest coal and rock conditions; Step 602, Transformer model dynamic prediction: After data processing and status update, the system calls the Transformer model for real-time prediction. By continuously inputting real-time sensor data and prediction results into the Transformer model, the system makes a detailed prediction of future coal and rock changes. Through continuous iterative prediction, the system can gradually optimize the path and cutting strategy; Step 603, cutting strategy optimization: Based on the dynamic prediction results provided by the Transformer model, the cutting strategy of the coal mining machine is optimized by combining real-time data and prediction information. According to the changing trend of the coal-rock structure, the cutting depth, cutting path and travel speed of the coal mining machine are adjusted to ensure that the equipment always remains in the best working state. The optimal cutting parameters are determined by analyzing key factors including the hardness and thickness of the coal seam and the coal-rock boundary. Step 604, dynamic adjustment and feedback: After the cutting strategy is optimized, the system dynamically adjusts the operating parameters of the coal mining machine according to real-time feedback, continuously monitors equipment operation and coal and rock changes through sensor data, and adjusts the cutting depth, path and speed in real time to ensure that the coal mining machine operates in the best condition. The feedback loop is used to analyze the difference between the actual effect and the optimization strategy and make adjustments quickly.

Images