Digital twin-driven optimization decision-making method for fully mechanized mining equipment and coal seam coupling system

Abstract

This invention relates to a digital twin-driven optimization decision-making method for a coupled system of fully mechanized mining equipment and coal seam. First, it constructs multiple analytical decision-making models based on a combination of complementary data and knowledge. Current multi-source heterogeneous sensing information is input into these models to derive multiple optimal solutions for the next cycle's fully mechanized mining equipment posture under different models. Then, these multiple optimal solutions are input into an optimization decision space. A unique absolute optimal solution for the next cycle's fully mechanized mining equipment posture is obtained through a combination of a self-evaluation algorithm at the bottom layer of the optimization decision space and manual visualization judgment during virtual execution in the virtual simulation space. This method can derive the absolute optimal solution for the fully mechanized mining equipment posture through optimization and decision-making, and dynamically repair and optimize the decision model, laying the foundation for subsequent precise reverse control.

Description

Technical Field

[0001] This invention relates to the field of intelligent coal mining technology, specifically to an optimization decision-making method for a fully mechanized mining equipment and coal seam coupling system driven by digital twins. Background Technology

[0002] Accelerating the intelligent development of coal mines and promoting the transformation and upgrading of the coal industry has become a consensus for high-quality development in the coal sector. Advanced information technologies such as virtual reality, Internet+, big data, cloud computing, and 5G are increasingly integrated with the coal industry, ushering in a new era of intelligent coal mining. Among these, the deep application of digital twin technology in intelligent coal mining is developing rapidly. Establishing a digital twin model of the mine production system enables behavioral simulation, operational status visualization analysis, and fault prediction for mining production processes such as fully mechanized mining, tunneling, ventilation, auxiliary transportation, and power supply.

[0003] In a mining production system, the fully mechanized longwall face is the most basic production unit and the primary production site. Its production process is complex, its equipment is large-scale, and its operating environment is harsh. The fully mechanized longwall face consists of a group of fully mechanized mining equipment and a coal seam. These two elements are coupled together to construct a "multi-constrained heterogeneous equipment group" operating system, characterized by high dynamism, incompleteness, and dynamic randomness. Therefore, research on the spatiotemporal kinematic relationship between the fully mechanized mining equipment and the coal seam, as well as the operational constraints and laws governing the equipment's advancement process, is challenging.

[0004] Currently, research on digital twin technology in fully mechanized mining faces mainly focuses on the virtual reconstruction and real-time monitoring of the coupled operation of fully mechanized mining equipment and coal seams. The goal is to construct highly reliable digital twin models of the constituent elements of the fully mechanized mining system, establish a synchronous mapping channel between the virtual and real worlds, and ultimately create a digital working face with dynamic feedback to achieve planning, simulation, and dynamic monitoring of actual coal mining conditions.

[0005] In the prior art, patent document CN113887111A discloses a virtual integrated testing method for geology, coal seam, and equipment in fully mechanized mining faces, providing a method for constructing a digital twin of a three-dimensional geological model of the coal seam; patent document CN109783962A discloses a simulation method for collaborative propulsion of fully mechanized mining equipment based on a virtual reality physics engine; and patent document CN111140231B discloses a virtual planning method for the roof and floor paths of coal seams based on the spatiotemporal kinematics of fully mechanized mining equipment, conducting research on the virtual reconstruction of fully mechanized mining faces, realizing the collaborative propulsion of the "three machines" of the fully mechanized mining face and dynamic coupling with the workspace in a virtual environment, providing a solution for virtual planning of fully mechanized mining faces; and patent document CN113746936A discloses a VR and AR distributed collaborative intelligent monitoring system for fully mechanized mining faces, using the core technologies of VR and AR in digital twins to form a full-time and spatiotemporal three-dimensional visualization monitoring technology architecture that takes into account both the local and overall aspects of the fully mechanized mining face and integrates virtual and real elements.

[0006] The virtual reconstruction and real-time monitoring of fully mechanized mining equipment coupled with the coal seam, using a "virtual simulation of reality" approach, has enabled offline simulation, one-way mapping, and dynamic visualization of the fully mechanized mining face. However, to achieve a higher level of digital twin and realize timely analysis and precise control, it is necessary to utilize strategies, algorithms, and knowledge to achieve intelligent decision-making and optimization of the fully mechanized mining face.

[0007] In the prior art, patent document CN114251092A discloses a real-time decision-making and control method for fully mechanized mining equipment based on lidar. The method uses lidar to collect point cloud images of hydraulic support side guard plates and scraper conveyors. The point cloud images are subjected to geometric mean filtering for noise reduction and the point cloud images are solved. The solution results are compared with a behavior database. After comparison, the behavior decision module determines whether the side guard plates are safely retracted and whether the straightness of the scraper conveyor meets the requirements.

[0008] In the above scheme, the behavior decisions of the fully mechanized mining equipment are made by comparing the point cloud calculation results with the behavior database. This scheme relies on the reasonableness of the behavior database construction, and the decision results are only used to solve the interference problem between the cutting section of the coal mining machine and the side guard plate of the hydraulic support, as well as the straightness problem of the scraper conveyor. It does not perform global decision-making under the coupled state of the fully mechanized mining equipment and the coal seam.

[0009] In the prior art, patent document CN113379909A discloses a big data analysis and decision-making method and system for intelligent mining of transparent working faces. By continuously correcting and updating the transparent geological model through the coal seam exposed by the cutting eye during the mining process and the geological data newly generated during the production process, the precise control decision information of the fully mechanized mining equipment is obtained. The planning and cutting model is corrected in real time using big data, thereby realizing intelligent and precise mining of coal working faces.

[0010] The above scheme mainly relies on inertial navigation technology and radar positioning technology. Based on the working condition monitoring data of the fully mechanized mining equipment, big data is used to make real-time corrections to the cutting model. However, the analysis and decision-making process does not make full use of the advantages of digital twin technology, and it is impossible to evaluate the rationality and accuracy of the current decision-making model through three-dimensional visualization.

[0011] In the prior art, patent document CN114827144A discloses a three-dimensional virtual simulation decision-making distributed system for fully mechanized coal mining faces, which includes physical entities, edge intelligent gateways, data processing servers, model training servers, and rendering simulation servers. In terms of physical architecture, it presents a three-level distributed architecture of equipment layer, data layer, and simulation layer. It uses a standardized network communication protocol for communication and provides three operating modes: time-series action simulation mode, real-time data-driven operation mode, and data-driven simulation optimization mode.

[0012] While the above scheme embodies the concept of digital twins and constructs a three-dimensional virtual simulation model of a fully mechanized coal mining face, it focuses on the study of system architecture, only providing the system hierarchy, the components of each level, and the achievable functions, without providing specific decision-making ideas and processes. Summary of the Invention

[0013] The technical problem to be solved by the present invention is to provide a digital twin-driven optimization decision-making method for a fully mechanized mining equipment and coal seam coupling system, which utilizes the advantages of digital twin technology to make global decisions under the coupled state of fully mechanized mining equipment and coal seam.

[0014] To solve the above technical problems, the technical solution adopted by this invention is: a digital twin-driven optimization decision-making method for a fully mechanized mining equipment and coal seam coupling system, comprising the following steps:

[0015] Step 1, prediction of multiple optimal solutions coupled between fully mechanized mining equipment and coal seam, including:

[0016] Step 1: Construct an analytical decision-making model that integrates multiple knowledge and data;

[0017] The coupling and collaborative operation law between fully mechanized mining equipment and coal seam was analyzed, a mathematical model was established, and a library of fully mechanized mining equipment pose information analysis and decision-making algorithms that combines data and knowledge was constructed based on a variety of data-based and knowledge-based analysis and decision-making algorithms, and encapsulated as a dynamic link library in DLL format.

[0018] Step 2: Input the multi-source heterogeneous sensing information into the analysis and decision-making model described in Step 1;

[0019] Optimize the real-time positioning, attitude determination and error correction technology of fully mechanized mining equipment, improve the accuracy of real-time sensing information of equipment, fuse actual information and virtual information from multiple sources, and input the fused multi-source heterogeneous sensing information of fully mechanized mining equipment into the analysis and decision-making algorithm library.

[0020] Step 3: Obtain the multiple optimal solutions for the pose of the fully mechanized mining equipment in the next cycle under different analysis and decision-making models;

[0021] Different analysis and decision-making models generate multiple outputs, which constitute the multiple optimal solutions for the pose of the fully mechanized mining equipment in the next cycle. The multiple optimal solutions are exported and stored.

[0022] Step two, prediction of the absolute optimal solution coupled between fully mechanized mining equipment and coal seam, including:

[0023] Step 4: Construct a self-evaluation algorithm at the bottom layer of the optimization decision space, and input the multiple optimal solutions of the next cycle fully mechanized mining equipment pose under different models into the optimization decision space;

[0024] The content and weight of the evaluation factors for the rationality of the posture of the fully mechanized mining equipment in the next cycle are determined, the evaluation index system and evaluation rules are formulated, the fuzzy comprehensive evaluation method is used to comprehensively evaluate the multiple optimal solutions of the posture of the fully mechanized mining equipment in the next cycle, the comprehensive evaluation mechanism is transformed into a self-evaluation and verification algorithm, and embedded into the bottom layer of the optimization decision space.

[0025] Step 5: By combining the optimization of the self-evaluation algorithm at the bottom layer of the decision space with the manual visualization judgment during the virtual execution process of the virtual simulation space, the unique absolute optimal solution for the posture of the fully mechanized mining equipment in the next cycle is obtained.

[0026] Multiple optimal solutions are input into the optimization decision space. Each optimal solution is evaluated by the underlying self-evaluation verification algorithm. At the same time, the three-dimensional visualization virtual execution of each optimal solution is carried out using the virtual deduction space to form a composite decision model. The unique absolute optimal solution of the pose of the fully mechanized mining equipment in the next cycle is obtained by combining the self-evaluation verification algorithm with manual visualization judgment. The self-evaluation verification algorithm is dynamically repaired and iteratively optimized.

[0027] Further, step 1 includes the following steps:

[0028] Step 101: Analyze the coupled and coordinated operation law between fully mechanized mining equipment and coal seam, and establish a mathematical model of the coupling relationship between the main parameters during the operation of fully mechanized mining equipment based on the workflow of fully mechanized mining equipment.

[0029] Step 102: Based on the theoretical basis of the coupled and coordinated operation law and mathematical model of fully mechanized mining equipment and coal seam, construct a variety of data-based analysis and decision-making algorithm models and knowledge-based analysis and decision-making algorithm models;

[0030] Step 103: Optimize and integrate various different analysis and decision-making algorithm models to form a comprehensive mining equipment pose information analysis and decision-making algorithm library that combines data and knowledge complementarity;

[0031] Step 104: The algorithm library for analyzing and deciding on the pose information of fully mechanized mining equipment, which combines data and knowledge, is packaged into a dynamic link library in DLL format for use by the digital twin system for optimizing the decision-making of the fully mechanized mining equipment and coal seam coupling system. The digital twin system for optimizing the decision-making of the fully mechanized mining equipment and coal seam coupling system is developed by Unity3d and includes a virtual monitoring space, an optimization decision-making space, and a virtual simulation space.

[0032] Furthermore, in step 103, the optimization and integration of multiple different analysis and decision-making models refers to the algorithm-level collaboration of knowledge and data, that is, selecting one or more from data-based analysis and decision-making models and knowledge-based analysis and decision-making models to collaboratively form a knowledge and data fusion analysis and decision-making model.

[0033] Furthermore, in step 104, a digital twin of the fully mechanized mining face, which exhibits the coupled and collaborative operation law of the fully mechanized mining equipment and the coal seam, is constructed in the virtual monitoring space. During the advancement of the fully mechanized mining face, some actual parameters are incorporated, and the autonomous synchronous operation of the digital twin of the fully mechanized mining face is achieved through the physics engine in Unity3d. This provides the optimization decision space with real-time dynamic updates of virtual coal seam cutting and virtual sensor sensing information. The optimization decision space is used to predict the absolute optimal solution of the coupling between the fully mechanized mining equipment and the coal seam. The virtual simulation space is used to perform virtual execution and three-dimensional visualization of the decision results output by the optimization decision space, so as to facilitate manual intervention in decision-making.

[0034] Further, step 2 includes:

[0035] Step 201: Optimize the real-time positioning, attitude determination, and error correction technology of the coal mining machine, hydraulic support, and scraper conveyor to achieve accurate judgment and perception of the working status of the fully mechanized mining equipment;

[0036] Step 202: Under the conditions of complex and changeable environment and easy interference of sensing signals, perform data preprocessing on the real-time sensing information of the equipment acquired by the actual sensors to further improve the accuracy of the real-time sensing information of the equipment.

[0037] Step 203: In the digital twin system for optimizing the decision-making of the fully mechanized mining equipment and the coal seam coupling system, multi-source heterogeneous sensing information composed of actual information and virtual information is fused to accurately obtain the current position and posture of the fully mechanized mining equipment.

[0038] Step 204: Call the DLL format dynamic link library file mentioned in step 104 to input the fused multi-source heterogeneous sensing information of the fully mechanized mining equipment into the analysis and decision-making algorithm library.

[0039] Furthermore, in step 203, the multi-source heterogeneous sensing information includes the real-time sensing information of the equipment acquired by the actual sensors in step 202, the real-time dynamic update information of virtual coal seam cutting in the virtual monitoring space, and the sensing information of virtual sensors.

[0040] Further, step 3 includes:

[0041] Step 301: Under the same input, different analysis and decision models in the analysis and decision algorithm library each produce an output, and multiple outputs constitute the multiple optimal solution of the pose of the fully mechanized mining equipment in the next cycle;

[0042] Step 302: Export and store the multiple optimal solutions output by various analytical decision models for further analysis and prediction.

[0043] Further, step 4 includes:

[0044] Step 401: Determine the evaluation factors for the rationality of the fully mechanized mining equipment position in the next cycle, including the straightness of the hydraulic support pushing and the scraper conveyor advancing, the flatness of the coal mining machine drum cutting path, the rock cutting rate and the coal retention rate, allocate the weight of each evaluation factor, and formulate the evaluation index system and evaluation rules.

[0045] Step 402: Use the fuzzy comprehensive evaluation method to comprehensively evaluate the multiple optimal solutions of the pose of the fully mechanized mining equipment in the next cycle under different models. This includes establishing the evaluation set for comprehensive evaluation, obtaining the evaluation matrix for single-factor fuzzy evaluation, determining the factor weight vector, establishing the comprehensive evaluation model, and calculating the total score of the system.

[0046] Step 403: Transform the comprehensive evaluation mechanism of multiple optimal solutions into a self-evaluation and verification algorithm through data analysis and simulation calculation;

[0047] Step 404: Embed the self-evaluation verification algorithm into the bottom layer of the optimization decision space in the fully mechanized mining equipment and coal seam coupling system optimization decision digital twin system.

[0048] Further, step 5 includes:

[0049] Step 501: Input the multiple optimal solutions output by various analysis and decision models into the optimization decision space, and evaluate each optimal solution through the self-evaluation verification algorithm at the bottom layer of the optimization decision space;

[0050] Step 502: During the evaluation of each optimal solution by the self-evaluation verification algorithm, the virtual simulation space is used to perform virtual execution of each optimal solution. The relationship between the matching and motion constraints of the three machines in the fully mechanized mining face corresponding to each optimal solution and the evolution of the working space of the three machines in the fully mechanized mining face during continuous advancement are presented in real time in three dimensions.

[0051] Step 503: Form a composite decision-making model based primarily on self-evaluation and verification of decisions in the analytical decision space, supplemented by artificial visualization intervention in the virtual deduction space;

[0052] Step 504: By combining the self-evaluation verification algorithm with manual visual judgment, the unique absolute optimal solution for the pose of the fully mechanized mining equipment in the next cycle is obtained. The self-evaluation verification algorithm is dynamically repaired and iteratively optimized to achieve the effect of decision-making without manual visual intervention.

[0053] Furthermore, in step 504, the self-evaluation verification algorithm undergoes dynamic repair and iterative optimization, which means that after obtaining the unique absolute optimal solution for the pose of the fully mechanized mining equipment in the next cycle, the consistency between the self-evaluation verification algorithm and the decision-making results of human visualization is analyzed. With human knowledge and experience as a reference, the self-evaluation verification algorithm is continuously corrected during the continuous advancement of multiple cutting cycles in the fully mechanized mining face.

[0054] Compared with existing technologies, the digital twin-driven optimization decision-making method for the coupled system of fully mechanized mining equipment and coal seam provided by this invention has the following beneficial effects:

[0055] (1) The next cycle of fully mechanized mining equipment posture decision-making process is refined into two stages: prediction of multiple optimal solutions coupled with coal seam and prediction of absolute optimal solutions coupled with coal seam. First, multiple optimal solutions are output through multiple analysis and decision-making models, and then the absolute optimal solution is further screened out through evaluation and verification. This ensures the accuracy of decision-making while effectively avoiding the blindness of solution seeking.

[0056] (2) When establishing the algorithm library for the position and posture information analysis and decision-making of fully mechanized mining equipment, a knowledge and data algorithm-level collaborative approach was adopted, which can combine data-driven inductive abstraction and knowledge-guided deductive reasoning to optimize and integrate various different analysis and decision-making algorithm models, use knowledge to guide the generation of data models, and inductively generate new knowledge from the data models to continuously improve the quality of algorithms in the algorithm library for the position and posture information analysis and decision-making of fully mechanized mining equipment.

[0057] (3) A digital twin system for optimization decision-making of fully mechanized mining equipment and coal seam coupling system was constructed, which includes a virtual monitoring space, an optimization decision-making space and a virtual simulation space. The virtual monitoring space is used to provide the optimization decision-making space with real-time dynamic update information of virtual coal seam cutting and virtual sensor sensing information. The virtual simulation space presents the evolution of the working space of the three fully mechanized mining machines in real-time three-dimensional visualization based on the decision-making results of the optimization decision-making space, so as to provide human intervention to assist decision-making. The three spaces have clear division of labor and are closely linked, making the decision-making more reasonable and efficient.

[0058] (4) When obtaining the current pose of the fully mechanized mining equipment, a multi-source heterogeneous sensing information consisting of the real-time sensing information of the equipment, the real-time dynamic update information of the virtual coal seam cutting, and the sensing information of the virtual sensor was obtained in Unity3D. The multi-source information fusion method was carried out by the extended Kalman filter algorithm. This method not only utilizes the real sensing data of the physical equipment, but also adds virtual twin data. The information of the two dimensions of the physical world and the virtual world is fused together to obtain the current pose of the fully mechanized mining equipment more accurately.

[0059] (5) A specific evaluation system was established to evaluate the decision results. That is, the fuzzy comprehensive evaluation method was used to comprehensively evaluate the multiple optimal solutions of the next cycle of fully mechanized mining equipment under different models. The comprehensive evaluation mechanism of multiple optimal solutions was transformed into a self-evaluation and verification algorithm through data analysis and simulation calculation. This algorithm was embedded into the bottom layer of the optimization decision space in the digital twin system of fully mechanized mining equipment and coal seam coupling system optimization decision. It can judge whether the decision results are reasonable in real time during the continuous operation of the fully mechanized mining equipment digital twin model. There is no need to establish an additional interface between the evaluation and verification algorithm and the digital twin model, which improves the integration of the system.

[0060] (6) In the process of obtaining the unique absolute optimal solution of the pose of the fully mechanized mining equipment in the next cycle, a combination of self-evaluation verification algorithm and manual visualization judgment was adopted. The self-evaluation verification decision was the main one, and the manual visualization intervention decision was the auxiliary one. The best pose of the fully mechanized mining equipment in the next cycle was presented in the virtual simulation space. Through human experience and intelligent auxiliary judgment, the self-evaluation verification algorithm was continuously manually intervened and optimized to avoid the cumulative error caused by the imperfection of the self-evaluation verification algorithm based on the fuzzy comprehensive evaluation method after multiple consecutive truncation cycle decisions. Attached Figure Description

[0061] The accompanying drawings, which are provided to further illustrate the invention and form part of this application, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention.

[0062] Figure 1 This invention provides a technical roadmap for an optimization decision-making method for a fully mechanized mining equipment and coal seam coupling system driven by a digital twin.

[0063] Figure 2 This is a schematic diagram of the process for predicting multiple optimal solutions by coupling fully mechanized mining equipment with the coal seam.

[0064] Figure 3 This is a diagram of the digital twin system architecture for optimizing the decision-making of the fully mechanized mining equipment and coal seam coupling system.

[0065] Figure 4 This is a schematic diagram of the process for predicting the absolute optimal solution of fully mechanized mining equipment coupled with the coal seam.

[0066] Figure 5 This is a software integration principle diagram of an optimization decision-making method for a fully mechanized mining equipment and coal seam coupling system driven by digital twins, provided by the present invention. Detailed Implementation

[0067] like Figure 1 As shown, a typical embodiment of the present invention provides a digital twin-driven optimization decision-making method for the coupled fully mechanized mining equipment and coal seam system, which includes two stages: prediction of multiple optimal solutions for the coupled fully mechanized mining equipment and coal seam, and prediction of the absolute optimal solution for the coupled fully mechanized mining equipment and coal seam.

[0068] The fully mechanized mining equipment described in this embodiment includes a coal mining machine, hydraulic supports, and a scraper conveyor.

[0069] like Figure 2 As shown, the multi-optimal solution prediction stage of fully mechanized mining equipment coupled with coal seam includes three steps: construction of an efficient analysis and decision-making model that integrates knowledge and data, input of multi-source heterogeneous sensing information into the analysis and decision-making model, and prediction of the multi-optimal solution of fully mechanized mining equipment pose under different models.

[0070] 1. The steps involved in building an efficient analytical decision-making model that integrates knowledge and data include:

[0071] Step 101: Conduct an in-depth analysis of the coupling and coordinated operation of the coal mining machine, hydraulic support, scraper conveyor, and coal seam. Based on the above-mentioned fully mechanized mining equipment workflow, establish a mathematical model of the coupling relationship between the main parameters during the operation of the fully mechanized mining equipment.

[0072] Step 102: Based on the theoretical basis of the coupled and coordinated operation law of fully mechanized mining equipment and coal seam and its mathematical model, construct a variety of data-based analysis and decision-making algorithm models and knowledge-based analysis and decision-making algorithm models.

[0073] The various data-based analysis and decision-making models mentioned in step 102 include neural networks, genetic algorithms, simulated annealing algorithms, and reinforcement learning algorithms.

[0074] The various knowledge-based analysis and decision-making models mentioned in step 102 include prior knowledge, fuzzy reasoning, decision trees, and expert systems. The prior knowledge includes logical knowledge, causal knowledge, and scientific laws extracted from the coupled and coordinated operation of fully mechanized mining equipment and coal seams.

[0075] Step 103: Optimize and integrate various different analysis and decision-making algorithm models to form a comprehensive mining equipment pose information analysis and decision-making algorithm library that combines data and knowledge complementarity.

[0076] In step 103, optimizing and integrating multiple different analysis and decision-making models specifically refers to the algorithm-level collaboration of knowledge and data. This involves selecting one or more data-based analysis and decision-making models and knowledge-based analysis and decision-making models to collaboratively construct a knowledge- and data-integrated analysis and decision-making model. Examples include neural networks and decision trees collaboratively constructing neural network trees, neural network trees and fuzzy reasoning and genetic algorithms collaboratively constructing genetic fuzzy trees, and reinforcement learning and prior knowledge collaboratively constructing hierarchical reinforcement learning.

[0077] Step 104: Encapsulate the fully mechanized mining equipment pose information analysis and decision-making algorithm library, which combines data and knowledge, into a DLL format dynamic link library for use by the fully mechanized mining equipment and coal seam coupling system optimization decision-making digital twin system.

[0078] The digital twin system for optimization decision-making of the fully mechanized mining equipment and coal seam coupling system mentioned in step 104 was developed using Unity3d, as shown in the attached figure. Figure 3 As shown, it includes a virtual monitoring space, an optimization decision-making space, and a virtual simulation space.

[0079] A digital twin of the fully mechanized mining face was constructed in the virtual monitoring space, which shows the coupled and collaborative operation law of the composite fully mechanized mining equipment and the coal seam. During the advancement of the fully mechanized mining face, some actual parameters were incorporated. The digital twin of the fully mechanized mining face was autonomously and synchronously operated through the physics engine in Unity3d, providing real-time dynamic updates of virtual coal seam cutting and virtual sensor sensing information.

[0080] The above-mentioned actual parameters include the initial position of the fully mechanized mining equipment before mining, the current traction speed of the coal mining machine, the stroke of the coal mining machine rocker arm height adjustment cylinder, and the stroke of the hydraulic support column cylinder and the pushing cylinder.

[0081] The decision space is optimized to predict the absolute optimal solution for the coupling of fully mechanized mining equipment and coal seam.

[0082] The virtual simulation space is used to perform virtual execution and three-dimensional visualization of decision results, so that human intervention in decision-making can be provided.

[0083] The aforementioned virtual simulation space can correct the virtual monitoring space through simulation results, thereby improving the accuracy of the real-time dynamic update information on virtual coal seam cutting and the information sensed by virtual sensors provided by the virtual monitoring space.

[0084] 2. The steps included in the multi-source heterogeneous sensing information input analysis and decision-making model are as follows:

[0085] Step 201: Optimize the real-time positioning, attitude determination, and error correction technology of the coal mining machine, hydraulic support, and scraper conveyor to achieve accurate judgment and perception of the working status of the fully mechanized mining equipment;

[0086] Step 202: Under the conditions of complex and changeable environment and easy interference of sensing signals, perform data preprocessing on the real-time sensing information of the equipment acquired by the actual sensors to further improve the accuracy of the real-time sensing information of the equipment.

[0087] In step 202, the real-time sensing information of the equipment acquired by the actual sensors is stored in an SQL Server database.

[0088] Specifically, the data preprocessing in step 202 includes data cleaning, data integration, and data transformation, which are performed in the SQL Server database.

[0089] The aforementioned data cleaning specifically refers to noise reduction processing of noise data generated by vibration during the operation of fully mechanized mining equipment.

[0090] The aforementioned data integration specifically refers to merging sensor data from different sources into a single data storage device. The sensor data from different sources refers to the real-time position and orientation information of the fully mechanized mining equipment acquired through various sensors mounted on it.

[0091] The aforementioned data transformation specifically refers to converting sensor data from different sources into appropriate formats and forms to meet the needs of software analysis.

[0092] Step 203: In the digital twin system for optimizing the decision-making of the fully mechanized mining equipment and the coal seam coupling system, multi-source heterogeneous sensing information composed of actual information and virtual information is fused to accurately obtain the current position and posture of the fully mechanized mining equipment.

[0093] The multi-source heterogeneous sensing information mentioned in step 203 includes the real-time sensing information of the equipment acquired by the actual sensors in step 202, the real-time dynamic update information of virtual coal seam cutting in the virtual monitoring space, and the sensing information of virtual sensors.

[0094] In step 203, the multi-source information fusion employs an extended Kalman filter algorithm. This algorithm dynamically adjusts the weights of each sensor based on optimized error estimation, achieving adaptive filtering and higher fusion accuracy. The multi-source information fusion is performed in Matlab software.

[0095] Step 204: Call the DLL format dynamic link library file mentioned in step 104 to input the fused multi-source heterogeneous sensing information of the fully mechanized mining equipment into the analysis and decision-making algorithm library.

[0096] 3. The steps included in the prediction of multiple optimal solutions for the pose of fully mechanized mining equipment under different models are as follows:

[0097] Step 301: Under the same input, different analysis and decision models in the analysis and decision algorithm library each produce an output, and multiple outputs constitute the multiple optimal solution of the fully mechanized mining equipment pose in the next cycle.

[0098] Step 302: Export and store the multiple optimal solutions output by various analytical decision models for further analysis and prediction.

[0099] As attached Figure 4 As shown, the stage of predicting the absolute optimal solution of the fully mechanized mining equipment coupled with the coal seam includes two steps: constructing a self-evaluation algorithm for optimizing the bottom layer of the decision space and solving for the unique optimal solution of the fully mechanized mining equipment pose during virtual execution.

[0100] 4. The self-evaluation algorithm construction process at the bottom layer of the optimization decision space includes the following steps:

[0101] Step 401: Determine the evaluation factors for the rationality of the fully mechanized mining equipment position in the next cycle, including the straightness of the hydraulic support pushing and the scraper conveyor advancing, the flatness of the coal mining machine drum cutting path, the rock cutting rate and the coal retention rate, allocate the weight of each evaluation factor, and formulate an evaluation index system and evaluation rules.

[0102] Step 402: Use the fuzzy comprehensive evaluation method to comprehensively evaluate the multiple optimal solutions of the pose of the fully mechanized mining equipment in the next cycle under different models. This includes establishing the evaluation set for comprehensive evaluation, obtaining the evaluation matrix for single-factor fuzzy evaluation, determining the factor weight vector, establishing the comprehensive evaluation model, and calculating the total score of the system.

[0103] Step 403: Transform the comprehensive evaluation mechanism of multiple optimal solutions into a self-evaluation and verification algorithm through data analysis and simulation calculation.

[0104] Step 404: Embed the self-evaluation verification algorithm into the bottom layer of the optimization decision space in the fully mechanized mining equipment and coal seam coupling system optimization decision digital twin system.

[0105] 5. The steps involved in finding the unique optimal solution for the pose of the fully mechanized mining equipment during virtual execution include:

[0106] Step 501: Input the multiple optimal solutions output by various analysis and decision models into the optimization decision space, and evaluate each optimal solution through the self-evaluation verification algorithm at the bottom layer of the optimization decision space.

[0107] Step 502: During the evaluation of each optimal solution by the self-evaluation verification algorithm, the virtual simulation space is used to perform virtual execution of each optimal solution. The relationship between the matching and motion constraints of the three machines in the fully mechanized mining face corresponding to each optimal solution and the evolution of the working space of the three machines in the fully mechanized mining face during continuous advancement are presented in real time in three dimensions.

[0108] Step 503: Form a composite decision-making model based primarily on self-evaluation and verification of decisions in the analytical decision space, supplemented by artificial visualization intervention decisions in the virtual deduction space.

[0109] Step 504: By combining the self-evaluation verification algorithm with manual visual judgment, the unique absolute optimal solution for the pose of the fully mechanized mining equipment in the next cycle is obtained. The self-evaluation verification algorithm is dynamically repaired and iteratively optimized to achieve the effect of decision-making without manual visual intervention.

[0110] Specifically, the dynamic repair and optimization of the composite decision model in step 504 refers to analyzing the consistency between the self-evaluation verification algorithm and the decision-making results of manual visualization after obtaining the unique absolute optimal solution of the posture of the fully mechanized mining equipment in the next cycle. With human knowledge and experience as a reference, the self-evaluation verification algorithm is continuously corrected during the continuous advancement of multiple cutting cycles in the fully mechanized mining face, so as to gradually improve its reliability.

[0111] As attached Figure 5 As shown, the software integration process of the digital twin-driven fully mechanized mining equipment and coal seam coupling system provided by this invention is as follows:

[0112] (1) The algorithm library for analyzing and making decisions on the pose information of fully mechanized mining equipment by combining data and knowledge is encapsulated as a dynamic link library in dll format;

[0113] (2) Collect actual operating data of the fully mechanized mining equipment, store it in the SQL Server database, and preprocess the sensor signals in SQL Server, including data cleaning, data integration and data transformation;

[0114] (3) Transfer the actual operation data of the equipment into Matlab, and transfer the initial pose of the fully mechanized mining equipment before the start of cutting, the current traction speed of the coal mining machine, the stroke of the coal mining machine rocker arm height adjustment cylinder, and the stroke of the hydraulic support column cylinder and the pushing cylinder into Unity3d;

[0115] (4) Input the real-time dynamic update information of virtual coal seam cutting and the sensing information of virtual sensors in Unity3d into Matlab, and read the actual operation data and virtual data of the fully mechanized mining equipment in Matlab to perform multi-source information fusion;

[0116] (5) Input the fused current cutting cycle fully mechanized mining equipment operation data into the analysis and decision-making algorithm library to predict the multiple optimal solutions for the equipment pose in the next cycle;

[0117] (6) The multi-optimal solution is returned to the fully mechanized mining equipment and the coal seam coupling system optimization decision digital twin system in Unity3d for comprehensive evaluation and verification algorithm combined with artificial visualization intervention to obtain the unique optimal solution of the fully mechanized mining equipment pose for the next cutting cycle, and the self-evaluation verification algorithm is corrected.

[0118] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A digital twin driven optimization decision method for fully mechanized mining equipment and coal seam coupling system, characterized in that, Comprising the following steps: Step one, fully mechanized mining equipment and coal seam coupling multiple optimal solution prediction, including: Step 1: Construct a variety of knowledge and data fusion analysis decision model; Analysis of fully mechanized mining equipment and coal seam coupling collaborative operation law, establish mathematical model, based on a variety of based on data and based on knowledge analysis decision algorithm to build data and knowledge complementary combination of fully mechanized mining equipment pose information analysis decision algorithm library, and encapsulated as dll format dynamic link library; Step 2: input the analysis decision model of step 1 with multi-source heterogeneous sensing information; Optimize the real-time positioning and error correction technology of fully mechanized mining equipment, improve the accuracy of real-time sensing information of equipment, and fuse the actual information and virtual information with multi-source information, input the fused multi-source heterogeneous sensing information of fully mechanized mining equipment into the analysis decision algorithm library; Step 3: obtain the next cycle of fully mechanized mining equipment pose multiple optimal solution under different analysis decision models; Different analysis decision models produce multiple outputs, which constitute the next cycle of fully mechanized mining equipment pose multiple optimal solution, and the multiple optimal solution is exported and stored; Step two, fully mechanized mining equipment and coal seam coupling absolute optimal solution prediction, including: Step 4: build a self-evaluation algorithm at the bottom of the optimization decision space, and input the next cycle of fully mechanized mining equipment pose multiple optimal solution under different models into the optimization decision space; Different analysis decision models produce multiple outputs, which constitute the next cycle of fully mechanized mining equipment pose multiple optimal solution, and the multiple optimal solution is exported and stored; Step 5: obtain the only absolute optimal solution of the next cycle of fully mechanized mining equipment pose by combining the self-evaluation algorithm at the bottom of the optimization decision space with artificial visual judgment during virtual execution in the virtual reasoning space; Input the multiple optimal solution into the optimization decision space, evaluate each optimal solution by the self-evaluation verification algorithm at the bottom, and perform three-dimensional visual virtual execution of each optimal solution by using the virtual reasoning space, form a composite decision model, and obtain the only absolute optimal solution of the next cycle of fully mechanized mining equipment pose by combining the self-evaluation verification algorithm with artificial visual judgment, and dynamically repair and iteratively optimize the self-evaluation verification algorithm.

2. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision method according to claim 1, characterized in that, The step 1 includes the following steps: Step 101: analyze the fully mechanized mining equipment and coal seam coupling collaborative operation law, establish a mathematical model of the coupling relationship between the main parameters in the operation process of the fully mechanized mining equipment based on the working process of the fully mechanized mining equipment; Step 102: based on the fully mechanized mining equipment and coal seam coupling collaborative operation law and mathematical model, construct a variety of different data-based analysis decision algorithm models and knowledge-based analysis decision algorithm models; Step 103: optimize and integrate the multiple different analysis decision algorithm models to form a fully mechanized mining equipment pose information analysis decision algorithm library complementary combined with data and knowledge; Step 104: encapsulate the fully mechanized mining equipment pose information analysis decision algorithm library complementary combined with data and knowledge into a dll format dynamic link library for calling by the fully mechanized mining equipment and coal seam coupling system optimization decision digital twin system; the fully mechanized mining equipment and coal seam coupling system optimization decision digital twin system is developed by Unity3d, including a virtual monitoring space, an optimization decision space and a virtual reasoning space.

3. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision method according to claim 2, characterized in that: In step 103, the optimization and integration of multiple different analysis decision models refer to the algorithm level cooperation of knowledge and data, that is, one or more of the data-based analysis decision model and the knowledge-based analysis decision model are selected to cooperatively form the knowledge and data fusion analysis decision model.

4. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 2 or 3, characterized in that: In step 104, the digital twin of the fully mechanized coal mining face is constructed in the virtual monitoring space to simulate the coupling and cooperative operation of the fully mechanized coal mining equipment and the coal seam, and part of the actual parameters are input during the advancing of the fully mechanized coal mining face, so that the digital twin of the fully mechanized coal mining face is autonomously and synchronously operated through the physical engine in Unity3d, and the virtual coal seam cutting real-time dynamic update information and the virtual sensor sensing information are provided for the virtual coal seam cutting real-time dynamic update information and the virtual sensor sensing information. The optimization decision space is used to complete the absolute optimal solution prediction of the coupling of the fully mechanized coal mining equipment and the coal seam, and the virtual deduction space is used to complete the virtual execution and three-dimensional visualization of the decision results output by the optimization decision space for manual intervention.

5. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 4, characterized in that, The step 2 includes: Step 201: real-time positioning and pose correction technology of the shearer, the hydraulic support and the scraper conveyor to accurately judge and perceive the working state of the fully mechanized coal mining equipment; Step 202: data preprocessing of the real-time sensing information of the equipment obtained by the actual sensor under the condition that the environment is complex and variable and the sensing signal is easily disturbed, to further improve the accuracy of the real-time sensing information of the equipment; Step 203: multi-source information fusion of the multi-source heterogeneous sensing information composed of the actual information and the virtual information in the fully mechanized coal mining equipment and coal seam coupling system optimization decision digital twin system to accurately obtain the current pose of the fully mechanized coal mining equipment; Step 204: calling the dll format dynamic link library file in step 104, and inputting the fused multi-source heterogeneous sensing information of the fully mechanized coal mining equipment into the analysis decision algorithm library.

6. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 5, characterized in that: In step 203, the multi-source heterogeneous sensing information includes the real-time sensing information of the equipment obtained by the actual sensor in step 202, the virtual coal seam cutting real-time dynamic update information and the virtual sensor sensing information in the virtual monitoring space.

7. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 6, characterized in that, The step 3 includes: Step 301: under the same input, different analysis decision models in the analysis decision algorithm library generate one output each, and multiple outputs constitute multiple optimal solutions of the next cycle of the fully mechanized coal mining equipment pose; Step 302: exporting and storing the multiple optimal solutions output by the multiple analysis decision models for further analysis and prediction.

8. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 7, characterized in that, The step 4 includes: Step 401: determining the evaluation factors of the rationality of the next cycle of the fully mechanized coal mining equipment pose, including the straightness in the advancing process of the hydraulic support and the scraper conveyor, the flatness of the shearer drum cutting path, the rock cutting rate and the coal leaving rate, performing weight distribution of each evaluation factor, and formulating the evaluation index system and the evaluation details; Step 402: adopting the fuzzy comprehensive evaluation method to comprehensively judge the multiple optimal solutions of the next cycle of the fully mechanized coal mining equipment pose under different models, including establishing the evaluation set for comprehensive evaluation, obtaining single factor fuzzy evaluation and evaluation matrix, determining the factor weight vector, establishing the comprehensive evaluation model, and calculating the total score of the system. Step 403: convert the comprehensive evaluation mechanism of multiple optimal solutions into a self-evaluation verification algorithm through data analysis and simulation calculation; Step 404: embed the self-evaluation verification algorithm into the optimization decision space bottom layer of the fully mechanized coal mining equipment and coal seam coupling system optimization decision digital twin system.

9. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 8, characterized in that, The step 5 comprises: Step 501: input multiple optimal solutions output by multiple analysis decision models into the optimization decision space, and evaluate each optimal solution through the self-evaluation verification algorithm of the optimization decision space bottom layer; Step 502: during the evaluation of each optimal solution by the self-evaluation verification algorithm, perform virtual execution of each optimal solution using the virtual deduction space, and perform real-time three-dimensional visualization of the fully mechanized working face three-machine matching and motion constraint relationship corresponding to each optimal solution and the evolution of the working space of the fully mechanized three machines during the continuous advance of the fully mechanized working face; Step 503: form a composite decision model based on the self-evaluation verification decision of the analysis decision space as the main part and the artificial visualized intervention decision of the virtual deduction space as the auxiliary part; Step 504: obtain the absolute optimal solution of the next cycle of the unique position of the fully mechanized mining equipment by combining the self-evaluation verification algorithm and artificial visualized judgment, and dynamically repair and iteratively optimize the self-evaluation verification algorithm, finally achieving the effect of no artificial visualized intervention decision.

10. The digital twin driven fully mechanized coal mining equipment and coal seam coupling system optimization decision-making method according to claim 9, characterized in that: In step 504, the dynamic repair and iterative optimization of the self-evaluation verification algorithm means that after obtaining the absolute optimal solution of the next cycle of the unique position of the fully mechanized mining equipment, the consistency of the self-evaluation verification algorithm and the artificial visualized judgment decision result is analyzed, and the self-evaluation verification algorithm is continuously corrected in the process of continuous advance of the fully mechanized working face in multiple cutting cycles, with the knowledge and experience of people as the reference.

Images