A hydraulic support digital twin multi-level communication method based on heterogeneous data fusion

Abstract

The application provides a hydraulic support digital twin multi-level communication method based on heterogeneous data fusion, relates to the technical field of automatic monitoring of underground coal mine equipment and digital twin, and the method comprises the following steps: acquiring real-time operation condition data of a hydraulic support in a physical layer; processing the real-time operation condition data by using a core coordination layer to generate simulation instruction data for driving simulation; transmitting the simulation instruction data to a simulation calculation layer in an asynchronous mode through a message middleware to drive the simulation calculation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result data set; receiving the multi-dimensional simulation result data set returned by the simulation calculation layer through the message middleware, and performing comprehensive early warning evaluation and control instruction feedback based on the multi-dimensional simulation result data set.

Description

Technical Field

[0001] This invention relates to the field of automated monitoring and digital twin technology for underground equipment in coal mines, and in particular to a multi-level communication method for digital twins of hydraulic supports based on heterogeneous data fusion. Background Technology

[0002] Hydraulic supports, as key support equipment in fully mechanized coal mining faces, require precise monitoring and performance optimization to ensure mining safety and efficiency. In actual operation, it is necessary to collect structural dynamic data such as the angle of the four-link support, the three-dimensional position of the top beam base, and the stress on components. Simultaneously, it is necessary to monitor hydraulic system operating parameters such as the working pressure of the left and right columns, and obtain structural mechanical property data such as the stress distribution of flexible components. These data originate from various models and systems, including physical sensors, dynamic simulations, hydraulic simulations, and finite element analysis, exhibiting characteristics such as multi-source, heterogeneous, high-dimensionality, and strong coupling.

[0003] Traditional communication methods struggle to meet the real-time, reliability, and synchronization requirements of digital twin systems for efficient data acquisition, fusion, and distribution. This results in isolated data, fragmented models, and significant delays in virtual-to-physical interaction, hindering precise monitoring and intelligent decision-making for hydraulic supports. Existing monitoring systems often focus only on local structures (such as roof beams) or primarily on structured data storage and transmission, lacking in-depth fusion and collaborative analysis of multi-dimensional working condition data. They also fail to address the construction of high-fidelity virtual models, dynamic simulation, or real-time interaction with physical entities, making it difficult to comprehensively and accurately reflect the overall working status and performance of hydraulic supports under complex conditions. Summary of the Invention

[0004] To address the aforementioned technical problems, according to a first aspect of the embodiments of this application, a multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion is provided. The method includes: A multi-level communication method for digital twins of hydraulic supports based on heterogeneous data fusion is provided. This method includes: The system acquires real-time operating condition data of the hydraulic support at the physical layer; processes the real-time operating condition data using the core coordination layer to generate simulation command data for driving the simulation; asynchronously transmits the simulation command data to the simulation calculation layer via a message middleware to drive the simulation calculation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset; receives the multi-dimensional simulation result dataset returned by the simulation calculation layer via the message middleware, and performs comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset.

[0005] This solution enables real-time acquisition, efficient transmission, and deep integration of multi-source data for the entire hydraulic support system. It also drives a high-fidelity digital twin model for dynamic simulation and intelligent decision-making, comprehensively improving the accuracy of support condition perception, real-time response, and system reliability.

[0006] In one embodiment, acquiring real-time operating condition data of the hydraulic support at the physical layer includes: collecting initial state parameters, pose data, and pressure signals through a sensor network deployed on the hydraulic support; and transmitting the encapsulated and verified pose data and pressure signals to the core coordination layer based on a preset one-way serial communication protocol.

[0007] This solution establishes a reliable, low-latency uplink channel for real-world operational data, providing a real data foundation for the digital twin system.

[0008] In one implementation, the core coordination layer processes the real-time operating condition data to generate simulation command data for driving the simulation, including: matching the three-dimensional hydraulic support pose data with the actual collected pressure data using a physical behavior model to establish a pose-pressure mapping relationship; generating control signals for the control actuator using a hydraulic control model based on the deviation between the target pressure expectation value generated by the pose-pressure mapping relationship and the measured pressure feedback signal; and generating dynamic selection codes for the flexible component analysis mode using an operation decision model based on the real-time status and process knowledge base.

[0009] This solution enables intelligent central coordination between the physical layer and the simulation computing layer, facilitating data aggregation, model collaboration, real-time decision-making, and instruction distribution.

[0010] In one implementation, generating a dynamic selection code for a flexible component analysis mode using an operational decision model based on a real-time status and process knowledge base includes: generating a mode selection array containing multiple component identifier bits, wherein each component identifier bit is used to indicate whether the corresponding component is simulated using a rigid body model or a flexible body model; and upon receiving a warning assessment result for a risk area, dynamically updating the identifier bits of the corresponding high-risk components in the mode selection array, generating an updated dynamic selection code, and sending it to the simulation calculation layer.

[0011] This solution enables the intelligent selection of key components that require analysis using high-precision flexible body models, thereby optimizing the allocation of simulation resources and accurately focusing on risk areas.

[0012] In one implementation, driving the simulation calculation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset includes: driving the kinematic model of the support to solve and extract basic motion state perception data; driving the hydraulic control model of the support to solve and extract hydraulic system state perception data; and dynamically scheduling the rigid-flexible coupling model of the support to solve and extract structural stiffness and deformation perception data based on the received component selection encoding.

[0013] This solution enables the internal model to perform high-precision co-simulation and produces a structured, multi-dimensional performance dataset.

[0014] In one implementation, based on the received component selection code, dynamically scheduling the rigid-flexible coupling model of the support to solve and extract structural stiffness and deformation sensing data includes: according to the component selection code, dynamically calling the rigid body model or importing the flexible body model generated by the flexible neutral file converted from the finite element model through a parameter-driven conditional instantiation mechanism; automatically reconstructing the mechanical connection relationship between the instantiated flexible body model and adjacent components through node mapping technology, thereby realizing dynamic and seamless switching of the calculation and analysis mode.

[0015] This solution enables the system framework to dynamically switch between various analysis modes, such as fully rigid, partially flexible, or fully flexible, based on instructions, thereby achieving an adaptive balance between computational accuracy and efficiency.

[0016] In one implementation, the multi-dimensional simulation result dataset is generated based on a unified integration and mapping mechanism of multi-source data. After the simulation instruction data is asynchronously transmitted to the simulation computing layer through a message middleware, the method further includes: real-time aggregation of the data streams of various sub-models generated by multi-domain collaborative simulation; and fusion of the sub-model data streams with historical state sequences to construct a unified multi-model state vector to characterize the mapping state of entity objects in the physical layer in the digital space.

[0017] This solution achieves a comprehensive, consistent, and traceable state mapping of physical entities in the digital space, providing a solid and unified data foundation for upper-level early warning and diagnosis.

[0018] In one implementation, receiving the multi-dimensional simulation result dataset returned by the simulation calculation layer through the message middleware includes receiving a thematic data packet published in a progressive logic. The thematic data packet includes the following themes: a basic motion state perception theme, containing kinematic and basic mechanical data; a hydraulic system state perception theme, containing hydraulic drive system operating condition data; a structural stiffness and deformation perception theme, containing stress distribution of flexible components and structural deformation displacement information; and a comprehensive early warning and risk assessment theme, containing diagnostic and decision information generated based on multi-dimensional correlation analysis performed on multi-model state vectors.

[0019] This solution enables reliable data flow organization from high-fidelity simulation data generation and hierarchical aggregation to intelligent application distribution.

[0020] In one implementation, the comprehensive early warning assessment and control instruction feedback based on the multi-dimensional simulation result dataset includes: converting the multi-dimensional simulation result dataset into dynamic time series curves, hydraulic system monitoring curves, and multi-dimensional structural state visualization charts for intuitive interactive display; based on the comprehensive early warning and risk assessment results, when a specified risk indicator is predicted to exceed a preset safety threshold, a flexible component replacement instruction is generated to trigger the simulation calculation layer to perform a secondary fine simulation pre-run of the specific component.

[0021] This solution enables the digital twin to respond to changes in physical state in real time, greatly improving the system's real-time response and forward-looking decision-making closed loop.

[0022] In one implementation, converting the multi-dimensional simulation result dataset into dynamic time series curves, hydraulic system monitoring curves, and multi-dimensional structural state visualization charts for intuitive interactive display includes: generating real-time stress cloud maps based on structural flexibility analysis data to drive the three-dimensional scene; quantifying safety margins by combining stiffness state indicators and structural health indices; and simultaneously displaying a global risk level panel, fault mode diagnosis cards, and maintenance operation suggestion lists on the central control dashboard.

[0023] This solution provides operators with a well-structured, multi-dimensional diagnostic support and visual interactive platform.

[0024] In a second aspect, embodiments of this specification provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method as described in any of the first aspects.

[0025] Thirdly, embodiments of this specification provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any of the first aspects. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0027] Figure 1 This is a flowchart illustrating a multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion, according to Embodiment 1 of this application. Figure 2 This is a flowchart illustrating the multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion according to Embodiment 2 of this application. Figure 3 This is a schematic diagram of a multi-level communication architecture for a hydraulic support digital twin based on heterogeneous data fusion, as described in an embodiment of this application. Figure 4 This is a schematic diagram of the simulation calculation layer model involved in the embodiments of this application; Figure 5 This is a schematic diagram of the information flow diagram of hierarchical processing and integrated display of hydraulic support data involved in the embodiments of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided for the implementation of this specification. Detailed Implementation

[0028] Unless otherwise defined, the technical or scientific terms used in the embodiments of this specification shall have the ordinary meaning understood by one skilled in the art to which this specification pertains. The terms "first," "second," and similar terms used in the embodiments of this specification do not indicate any order, quantity, or importance, but are merely used to avoid confusion of constituent elements.

[0029] Unless the context otherwise requires, throughout this specification, "a plurality of" means "at least two," and "including" is interpreted as open-ended or encompassing, that is, "including, but not limited to." In the description of this specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this specification. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example.

[0030] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0031] As mentioned in the background technology, hydraulic supports are key support equipment in fully mechanized coal mining faces, and accurate monitoring and performance optimization of their operating status are crucial for ensuring mining safety and efficiency. In actual operation, it is necessary to collect structural dynamic data such as the angle of the four-link support, the three-dimensional position of the top beam base, and the stress on components. Simultaneously, it is necessary to monitor hydraulic system operating parameters such as the working pressure of the left and right columns, and obtain structural mechanical property data such as the stress distribution of flexible components. The aforementioned data originates from different models and systems, including physical sensors, dynamic simulations, hydraulic simulations, and finite element analysis, exhibiting characteristics such as multi-source, heterogeneous, high-dimensional, and strongly coupled nature.

[0032] Traditional communication methods struggle to meet the real-time, reliability, and synchronization requirements of digital twin systems for efficient data acquisition, fusion, and distribution. This results in isolated data, fragmented models, and significant delays in virtual-to-physical interaction, hindering precise monitoring and intelligent decision-making for hydraulic supports. Existing monitoring systems often focus only on local structures (such as roof beams) or primarily on structured data storage and transmission, lacking in-depth fusion and collaborative analysis of multi-dimensional working condition data. They also fail to address the construction of high-fidelity virtual models, dynamic simulation, or real-time interaction with physical entities, making it difficult to comprehensively and accurately reflect the overall working status and performance of hydraulic supports under complex conditions.

[0033] Based on the above inventive concept, the following is an exemplary description of the multi-level communication method for digital twins of hydraulic supports based on heterogeneous data fusion provided in the embodiments of this specification.

[0034] like Figure 1 As shown, as a first embodiment of this application, a multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion is provided. It includes: S101, acquire real-time operating condition data of the hydraulic support in the physical layer.

[0035] In practical implementation, the real-time operating data of the hydraulic supports in the physical layer generally refers to the multi-source heterogeneous sensor data generated by the physical entity. Acquiring this type of data forms the input basis for connecting the digital twin system with the real environment.

[0036] S102, the core coordination layer processes the real-time operating condition data to generate simulation instruction data for driving the simulation.

[0037] In practice, the core coordination layer analyzes and models the input operating data, outputting dynamic selection codes that drive the underlying simulation. These dynamic selection codes represent parameterized configuration instructions that control the model granularity and computational complexity. They define the use of rigid and flexible bodies by setting corresponding Boolean values ​​and control the switching mechanism between local and fully flexible modes.

[0038] S103, the simulation instruction data is asynchronously transmitted to the simulation calculation layer through the message middleware to drive the simulation calculation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset.

[0039] Specifically, the message middleware refers to the communication bus responsible for asynchronous, decoupled data distribution. The multi-domain collaborative simulation refers to a cross-disciplinary, cross-dimensional joint simulation calculation process executed in a high-performance computing environment, encompassing the cross-fusion of kinematics, fluid mechanics, and structural mechanics. Furthermore, this process utilizes flexible neutral files, which are standardized middleware formats containing component dynamic parameter systems, including mass matrices, stiffness matrices, and mode shapes, serving as a crucial data bridge connecting finite element analysis software and multibody dynamics software. When generating the dataset, the system extracts multi-model state vectors, which represent structured digital mirror data after data fusion processing. These vectors possess the ability to cascade historical state sequences to ensure the traceability and consistency of data for situation assessment and early warning diagnosis.

[0040] S104, receive the multi-dimensional simulation result dataset returned by the simulation calculation layer through the message middleware, and execute comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset.

[0041] In this process, a closed-loop data flow is achieved through a communication bus, thereby completing early warning feedback and situational awareness. As another preferred embodiment of this application, the above-mentioned multi-level communication method for hydraulic support digital twins based on heterogeneous data fusion will be further elaborated with richer technical details.

[0042] like Figure 2 As shown, a specific implementation method for multi-level communication of a hydraulic support digital twin based on heterogeneous data fusion is provided. It includes: S201, acquire real-time operating data of the hydraulic support in the physical layer.

[0043] Specifically, the real-time operating condition data can be implemented as test bench posture and pressure data, initial posture, and column raising and lowering signals.

[0044] In this step, the system acquires three-dimensional spatial attitude coordinates using high-precision stroke and tilt sensors deployed on the base, columns, top beam, and shield beam. Simultaneously, pressure sensors installed on the columns continuously extract working pressure data at a sampling rate of hundreds of milliseconds (a higher-frequency fiber optic sensing solution can also be used as an alternative in extreme high-pressure environments). This signal acquisition action through the core sensor network achieves a precise microscopic technical effect by acquiring a dataset characterizing the actual combined actions of the hydraulic support group, eliminating blind spots in the underlying data. Furthermore, the system employs a lightweight frame structure oriented towards the working conditions as the communication protocol, establishing a unidirectional data link through the System.IO.Ports.SerialPort serial communication module to encapsulate and verify the signals.

[0045] S202, the core coordination layer processes the real-time operating condition data to generate simulation instruction data for driving the simulation.

[0046] Specifically, the dynamic selection coding can be implemented as a five-element coding system (top beam, column, shield beam, four-link, base).

[0047] like Figure 3 As shown, after receiving the real column pressure data transmitted via serial port, the physical behavior model within the core coordination layer maps and matches it with the parsed 3D hydraulic support pose data in a high-precision rendering engine such as Unity3D. Furthermore, the hydraulic control model integrates a multi-channel PID controller, continuously acquiring feedback signals and comparing them with the expected value of the optimal target pressure under the current operating condition. It dynamically calculates and generates precise command signals for the electro-hydraulic proportional valves required to control the actuators (column or pushing jack). This PID algorithm can also be replaced by an adaptive fuzzy control algorithm to handle more nonlinear fluid damping fluctuations. Subsequently, the operation decision model rapidly decomposes the task based on the process specification knowledge base, outputting a five-element code containing five component identifier bits, in the form of (0,1,0,0,0), thus initially setting the column as a flexible body model. This intelligent selection mode mechanism effectively filters out invalid computational redundancy and precisely focuses the system's computing resources on the microscopic technical effects of high-pressure stress distribution areas.

[0048] S203, the simulation instruction data is asynchronously transmitted to the simulation calculation layer through the message middleware to drive the simulation calculation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset.

[0049] like Figure 4As shown, specifically, the message middleware can be implemented as an MQTTX server. Through this lightweight protocol's publish-subscribe mechanism (or alternatively, Kafka as a solution for ultra-large-scale clusters), the system stably decouples control signals from the quinary encoding and sends them to high-performance computing environments such as MWORKS.Sysplorer. Specifically, the multi-domain collaborative simulation can be implemented as simulations of multibody dynamics models, hydraulic system models, and rigid-flexible coupling models. During the collaborative simulation execution phase, the built-in solver dynamically schedules the received encoding. Specifically, the flexible neutral file can be implemented as a .mnf binary file. When the flag bit is 1, a standard .mnf format file, pre-processed by ANSYS software and containing complete dynamic parameters such as the mass stiffness matrix, is imported through a dedicated interface to establish a high-precision flexible body component. Utilizing node mapping technology, the system automatically reconstructs the mechanical connection relationships between the newly instantiated model and adjacent rigid components in the system (such as between a top beam and a column) within microseconds. This action achieves topological integrity preservation when adaptively switching between fully rigid, locally flexible, or fully flexible modes, avoiding the microscopic technical problems of penetration distortion and computational divergence caused by mesh reconstruction in traditional co-simulation.

[0050] S204 aggregates data streams from various sub-models in real time and merges them with historical state sequences to construct a unified multi-model state vector.

[0051] Specifically, the multi-model state vector can be implemented as multi-model state vector data. After the simulation layer outputs the data, the communication bus mechanism captures massive amounts of streaming data in real time, such as the linkage angles and constraint reactions from multibody dynamics, the cylinder strokes and pressures from the hydraulic system, and the stress distribution and structural deformation displacements of flexible components from rigid-flexible coupling. The system splices and fuses the multidimensional parameters of the current period with the associated historical state sequences, encapsulating them into a structured unified dataset. This integrated mapping operation brings about the micro-technical effect of timestamp synchronization of heterogeneous data across disciplines, ensuring that there is no time phase misalignment in the upper-level diagnostic analysis.

[0052] S205, receive the multi-dimensional simulation result dataset returned by the simulation calculation layer through the message middleware, and execute comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset.

[0053] Combination Figure 5It is known that the core coordination layer captures the returned data by listening to four logically progressive topic packets. The platform then renders and outputs the otherwise monotonous data stream: real-time plotting of time-series curves with the displacement and support height as the Y-axis, and hydraulic monitoring curves displaying imbalance and leakage coefficients; simultaneously, it uses JSON format standard data packets to drive the deformation of the underlying 3D mesh, intuitively rendering real-time stress cloud maps and structural health index dials with deformation magnification coefficients. In an optional implementation, when the system's situation assessment algorithm, based on fused historical data, detects a sharp increase in column pressure gradient approaching a dangerous threshold, the decision-making mechanism quickly triggers an action: updating the five-element code to switch the high-risk component identifier to 1, generating a flexible component replacement instruction and issuing it to the simulation layer to perform a secondary fine-grained failure mode pre-simulation of extreme conditions. This pre-simulation decision-making closed loop gives the system the ability to intervene proactively before substantial damage to the physical structure occurs, ultimately achieving the micro-technical effect of transforming from reactive, reactive maintenance to data-driven predictive maintenance.

[0054] In one specific implementation, the construction and operation process of the core coordination layer is as follows: First, a high-precision 3D model of the hydraulic support and scraper conveyor is built in UG software. Kinematic and dynamic analysis is performed on each kinematic pair. The model is converted into FBX format and imported into the Unity3D engine. A strict parent-child hierarchical structure is constructed based on the actual assembly relationship to accurately describe the motion constraints between components. Mesh colliders are created for the components in contact between the top beam and the coal seam, and physical material parameters such as the friction coefficient are set to provide a foundation for accurate physical interference detection between the top beam and the coal wall and subsequent command-pose feedback verification. Subsequently, rigid body components and mesh collision bodies were added to the hydraulic support to accurately define physical parameters such as mass distribution and friction coefficient. A kinematic model of the support based on a finite state machine was developed using C# scripts to achieve precise control of actions such as column lifting and lowering. The working cycle logic of "column lowering-support moving-column raising-pushing" was designed based on the finite state machine architecture to establish a behavioral rule system consistent with the real equipment. Using real column pressure data transmitted in real time from the physical layer through the serial port, the pose of the three-dimensional hydraulic support obtained in real time was correlated and matched with the actual pressure values ​​of multiple columns collected synchronously to complete the pose-pressure mapping. The hydraulic control model, acting as a generator of control commands, integrates a multi-channel PID controller. It continuously receives measured pressure feedback signals from the physical layer and compares them in real time with the target pressure expectation value representing the optimal state of the current operating condition provided by the pose-pressure mapping module. It calculates the pressure deviation of each control loop and then dynamically generates control signals corresponding to each hydraulic actuator. Simultaneously, the model performs in-depth analysis of the pressure-command correlation in parallel, monitoring and recording multi-variable time series data such as control signals, valve position feedback, actual system pressure, and pressure tracking error in real time. Through online identification algorithms, it dynamically analyzes key dynamic characteristics such as the gain and response time of the command-pressure transmission relationship.The operational decision-making model embeds a process knowledge base and decision rules for the standard operation cycle of the fully mechanized mining face. When the system needs to execute a certain process command, the decision engine integrates real-time hydraulic support pose information from the physical layer, the current system pressure state, and references the control dynamic characteristics revealed by pressure-command correlation analysis. It performs rapid logical judgment and task decomposition, parsing high-order process commands in real time and outputting them as a series of specific basic action commands with timing and parameter constraints. Furthermore, the model also embeds intelligent selection logic for flexible component analysis modes, responsible for intelligently selecting key components in the hydraulic support that require high-precision flexible body models for analysis. This optimizes simulation resource allocation and focuses on risk areas. The decision results are output as a set of five-element codes in the form of (top beam, column, shield beam, four-link, ...). (Base), where "0" indicates that the component adopts a rigid body model and "1" indicates that it adopts a flexible body model; during the initial mode selection, according to the preset of the process knowledge base, the main pressure-bearing and easily deformable columns are initially set as flexible bodies, that is, the output code is (0,1,0,0,0); the model continuously receives comprehensive early warning and risk assessment results from the simulation calculation layer, analyzes the early warning information, identifies risky components, and dynamically updates the five-element code, switching the corresponding component bit from "0" to "1". The updated code will be issued to the simulation calculation layer in real time as a new flexible component selection instruction, driving it to adjust the model configuration, thereby completing the closed-loop optimization of the analysis mode.

[0055] In another specific implementation, the detailed construction and dynamic switching mechanism of the simulation calculation layer is as follows: When constructing the kinematic model of the support, the three-dimensional model of the hydraulic support is first geometrically repaired and simplified in UG software. Non-critical features such as small-scale fillets, chamfers, and bolt holes that do not affect the overall stiffness are suppressed, and the geometric topology is optimized. The material mechanical parameters, including basic properties such as elastic modulus, Poisson's ratio, and material density, are accurately defined for the simplified geometric model. The load conditions, boundary conditions, and required output variables are clarified according to the simulation objectives. Based on the processed geometric model, the kinematic pair constraints between the main components such as the top beam, column, and base are accurately defined in the MWORKS.Sysplorer platform according to the actual mechanical structure of the hydraulic support using CAD tools. A complete system topology relationship is constructed, and this rigid model is used to simulate the overall kinematic behavior of the support. The model calculates and outputs the link angles of each hinge point in the four-bar linkage of the hydraulic support in real time, accurately calculates the three-dimensional position coordinates of the top beam and base in the global coordinate system, and synchronously extracts the constraint reaction forces at each kinematic pair and the force data of key components such as the top beam and base, which together constitute the basic motion state perception dataset. When constructing the hydraulic control model for the support structure, the built-in hydraulic model library of the MWORDS.Sysplorer platform was used to select standard physical components such as proportional valves, hydraulic cylinders, and pipelines. These components were connected according to the actual system schematic diagram of the hydraulic support, and a complete closed-loop hydraulic system simulation model including the pump station, multi-way valve group, column, and pushing jack hydraulic circuits was integrated to construct the model. To achieve precise control of the cylinder displacement, a multi-channel PID controller was integrated into the model for the column and pushing jack. The controller receives control signals from the core coordination layer and simultaneously obtains the actual cylinder displacement calculated internally by the model. As feedback, the PID algorithm calculates the deviation between the target value and the actual value in real time and dynamically outputs the control quantity. This control quantity directly drives the valve core opening of the proportional valve in the hydraulic model, thereby adjusting the flow rate into the oil cylinder, forming a closed-loop control system with displacement as the controlled variable. The output flow rate of the hydraulic pump, the real-time flow rate of each main oil circuit and control oil circuit, as well as the working pressure of the rodless chamber and rod chamber of the push jack, and the real-time working pressure of the lower chamber of the left and right columns are obtained through simulation calculations. The real-time extension and retraction displacement of the columns and the push jack are also output. The above data together constitute the hydraulic system state perception dataset.When constructing the rigid-flexible coupling model of the support structure, for components requiring flexible analysis (top beam, column, shield beam, four-bar linkage, base), preprocessing is performed in the ANSYS environment: detailed connection point definitions are defined for the imported component geometry; remote points with behavior set to "deformable body" are created at the assembly interfaces between components, and their degrees of freedom are precisely constrained according to the actual connection type; mesh generation is performed, and local refinement is implemented for complex areas to ensure element quality; flexible-specific settings are performed through the APDL command flow, and modal extraction methods such as Block are defined. The Lanczos method, order (up to the first 20), and material nonlinearity options are used. After the solution is completed, a standard modal neutral file (.mnf) is generated through a dedicated interface. This binary file contains a complete system of dynamic parameters, including the component's mass matrix, stiffness matrix, mode shapes, nodal coordinates, and participation factors. The rigid system model established in the support kinematics model is imported, and the generated .mnf file is imported using the flexible body model library function to create a high-precision flexible body component. This flexible body component is used to connect the corresponding rigid components in the system one by one. The node mapping technique ensures that the mechanical connection relationship between the flexible body and adjacent components is accurately established. During simulation, the platform's built-in solver selects the encoding based on the received real-time flexible components and dynamically schedules the analysis models of each component through a parameter-driven conditional instantiation mechanism. When the encoding is updated during the simulation, the platform selects the flexible body based on the new parameter values ​​through dynamic model reconstruction technology. The system destroys the original rigid body model and instantiates a new flexible body model based on the corresponding .mnf file. It automatically reconstructs the connection relationships between the newly instantiated model and adjacent components in the system using node mapping technology. The system automatically identifies the key node set used for connections on the flexible body model through preset interface identifiers and precisely reapplies the hinge constraints defined in the original rigid model to the corresponding key node set of the new flexible body model. Simultaneously, it ensures that the energy and momentum of the system do not undergo abrupt changes after connection reconstruction by solving the Lagrange equation. This allows the same system framework to seamlessly and dynamically switch between various analysis modes, such as fully rigid, partially flexible, or fully flexible, based on coded instructions, achieving an adaptive balance between computational accuracy and efficiency. Through model simulation, it extracts stress distribution data, structural deformation displacement data, and modal coordinate information of flexible components in real time, uniformly encapsulating them into standardized data packets in JSON format, collectively forming a structural stiffness and deformation sensing dataset.

[0056] As another preferred embodiment of this application, in order to further describe the complete technical solution of the present invention and expand the application depth of the digital twin model in the actual fully mechanized mining face, the specific underlying calculation logic and specific application scenarios are now extended and explained to provide sufficient modification support for the response to possible examination opinions.

[0057] Furthermore, in constructing the physical behavior model and 3D visualization scene, the system's underlying implementation logic includes building high-precision 3D models of the hydraulic support and scraper conveyor in 3D modeling software such as UG, performing geometric repair and feature simplification. Specifically, the system converts the optimized model into FBX format and imports it into the Unity3D engine, creating a Mesh collider for the contact components between the roof beam and the coal seam and setting physical material parameters such as the friction coefficient. Alternatively, in an alternative approach, other rendering platforms with underlying physical collision detection engines, such as Unreal Engine, can be used for construction. This action results in the ability to achieve accurate physical interference detection between the roof beam and the coal wall, and provides a high-precision visualization verification foundation for subsequent command-pose feedback, thus achieving a microscopic technical effect. Based on this, the system uses C# scripts to develop a finite state machine-based kinematic model of the support structure. Its specific parameters are strictly set to cover the state transition thresholds that cover the working cycle logic of "lowering the column - moving the support - raising the column - pushing the slide". In other implementation scenarios, the traditional hard-coded finite state machine can be replaced by introducing a hidden Markov model to predict the probabilistic state transitions. This brings about the micro-technical effect of establishing a behavior rule system that is highly consistent with the real equipment and achieving millisecond-level precise control of core actions.

[0058] Furthermore, in the refined processing stage of driving the simulation calculation layer to perform multi-domain collaborative simulation, for the state calculation of the hydraulic system, the model specifically integrates the extraction of fine fluid dynamic parameters such as the working pressure of the rodless and rod-type chambers of the pushing jack, the real-time working pressure of the lower chambers of the left and right columns, and the output flow rate of the hydraulic pump. Alternatively, higher-dimensional flow field streamline distribution monitoring is used as an alternative. This brings about a comprehensive microscopic technical effect in characterizing the nonlinear dynamic response characteristics of the hydraulic drive system. For the analysis of flexible components, the system performs flexible-specific settings in the finite element environment such as ANSYS through APDL command flow, defines the modal extraction method and order, and generates a binary modal neutral file (.mnf) containing the mass matrix and stiffness matrix. Through the platform's built-in dynamic model reconstruction technology and node mapping technology, the system automatically reconstructs the mechanical connection relationship between the instantiated flexible body model and adjacent components based on the new parameter values. This brings about a seamless dynamic switching between fully rigid, locally flexible, or fully flexible modes within the same framework, achieving a microscopic technical effect of adaptive balance between computational accuracy and solver computing power consumption.

[0059] In one optional implementation, the method provided in this application can be specifically applied to scenarios involving real-time monitoring and safety assessment of the entire support cycle. Specifically, during the actual coal mining process in a fully mechanized longwall face, the physical layer collects real-time position and pressure data of the support group at each stage of the cycle, and uploads this data in real-time to the core coordination layer and the simulation calculation layer via multi-level communication. The simulation layer drives multi-model collaborative simulation based on the received real-time data, synchronously calculating the motion state of the supports, the response of the hydraulic system, and the stress and deformation of key components, and communicates all the perceived data and comprehensive assessment results back to the visualization interface of the core coordination layer in real-time. Through this complete closed loop, operators can intuitively monitor and conduct immediate safety assessments of the entire support cycle, bringing about a macro-level technical effect that transforms underground coal mine safety management from traditional post-event result inspection to high-fidelity pre-event and in-event process visualization.

[0060] Preferably, the method provided in this application can also be applied to virtual debugging and control system optimization verification scenarios for support processes. Specifically, during the development of new support control system algorithms or the adjustment of existing processes, the operation decision model generates an action sequence based on the new process plan, and the hydraulic control model generates corresponding control signals. These instructions drive a high-fidelity model in the simulation calculation layer to perform collaborative simulation through a message middleware. The system response generated by the simulation is fed back to the core coordination layer in real time and compared with the expected behavior. Through the aforementioned virtual debugging actions, the rationality of the control logic can be fully verified, the smoothness of the actions can be evaluated, and possible hydraulic shocks and structural vibrations can be predicted before investing in expensive physical testing equipment. This results in the technical effect of efficiently and cost-effectively optimizing control strategies and tuning process parameters.

[0061] Furthermore, in response to extreme and sudden safety threats such as roof pressure in coal mines, the method provided in this application can be implemented in scenarios involving impact load early warning and simulation-based pre-emptive decision-making. Specifically, the simulation calculation layer continuously runs high-fidelity dynamics and structural simulations, predicting the system state change trend in the near future based on real-time fused data and historical sequences. When the simulation model predicts that specific key parameters (such as a steep increase in column pressure gradient or abnormal acceleration of the roof beam) are about to approach a preset danger threshold, the system immediately issues a high-level warning through the communication layer. At the same time, based on the judgment of the operation decision model, the core coordination layer automatically initiates a secondary fine simulation command, requiring the simulation calculation layer to switch specific high-risk components in stress concentration areas to flexible body models, and to perform high-precision pre-emptive simulations of their failure modes under extreme working conditions. The pre-emptive simulation results are transmitted back to the dashboard in real time through the communication layer. This action provides key technical support for the core coordination layer's emergency plan selection and manual intervention decision-making, significantly improving the proactive disaster prevention and mitigation capabilities of the fully mechanized mining face.

[0062] In one exemplary embodiment of this specification, an electronic device is also provided, such as Figure 6 As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute a multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion, the method including: Obtain real-time operating condition data of the hydraulic supports in the physical layer; The core coordination layer processes the real-time operating condition data to generate simulation instruction data for driving the simulation. The simulation instruction data is asynchronously transmitted to the simulation computing layer through a message middleware to drive the simulation computing layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset. The message middleware receives the multi-dimensional simulation result dataset returned by the simulation calculation layer, and performs comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset.

[0063] Alternatively, another hydraulic support digital twin multi-level communication method based on heterogeneous data fusion can be implemented, which includes: listening to the preset instruction topic of the dedicated communication center through the second receiving client, and receiving the target control instruction sent by the first terminal; The target control command is parsed and converted into a precise physical signal according to a preset mapping relationship; The precise physical signal is injected into the multidisciplinary coupling model to drive the multidisciplinary coupling model to perform dynamic calculations of a preset simulation step size and obtain real-time simulation state data. The real-time simulation status data is verified based on a preset conditional release mechanism, and when the release conditions are met, the encapsulated real-time simulation status data is released to the preset status topic of the dedicated communication hub through the second sending client.

[0064] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0065] In addition to the methods, apparatus, and devices described above, the multi-level communication method for hydraulic support digital twins based on heterogeneous data fusion provided in the embodiments of this specification can also be a computer program product, which includes computer program instructions that, when executed by a processor, cause the processor to perform the steps in the multi-level communication method for hydraulic support digital twins based on heterogeneous data fusion according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0066] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this specification. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages.

[0067] Furthermore, embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor of the steps in the heterogeneous data fusion-based digital twin multi-level communication method for hydraulic supports according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0068] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this specification can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0069] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0070] The embodiments described above are merely illustrative of several implementation methods outlined in this specification. While the descriptions are specific and detailed, they should not be construed as limiting the scope of the solutions provided in this specification. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this specification, and these all fall within the scope of protection of this specification. Therefore, the scope of protection for this patent should be determined by the appended claims.

Claims

1. A multi-level communication method for a hydraulic support digital twin based on heterogeneous data fusion, characterized in that, include: Obtain real-time operating condition data of the hydraulic supports in the physical layer; The core coordination layer processes the real-time operating condition data to generate simulation instruction data for driving the simulation. The simulation instruction data is asynchronously transmitted to the simulation computing layer through a message middleware to drive the simulation computing layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset. The message middleware receives the multi-dimensional simulation result dataset returned by the simulation calculation layer, and performs comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset.

2. The method according to claim 1, characterized in that, The acquisition of real-time operating condition data of the hydraulic support at the physical layer includes: Initial state parameters, pose data, and pressure signals are collected through a sensor network deployed on the hydraulic support. Based on a preset unidirectional serial communication protocol, the encapsulated and verified pose data and pressure signal are transmitted to the core coordination layer.

3. The method according to claim 1, characterized in that, The process of using the core coordination layer to process the real-time operating condition data and generate simulation instruction data for driving the simulation includes: A physical behavior model is used to match the pose data of the three-dimensional hydraulic support with the actual collected pressure data to establish a pose-pressure mapping relationship. Based on the deviation between the target pressure expectation value generated by the posture-pressure mapping relationship and the measured pressure feedback signal, the control signal for the control actuator is generated using the hydraulic control model. Based on the real-time status and process knowledge base, dynamic selection codes for flexible component analysis modes are generated using an operational decision model.

4. The method according to claim 3, characterized in that, The dynamic selection coding of flexible component analysis modes based on real-time status and process knowledge base, using an operation decision model, includes: Generate a mode selection array containing multiple component identifier bits, wherein each component identifier bit is used to indicate whether the corresponding component is simulated and analyzed using a rigid body model or a flexible body model; After receiving the early warning assessment results for the risk area, the identifier bits of the corresponding high-risk components in the mode selection array are dynamically updated, and the updated dynamic selection code is generated and sent to the simulation calculation layer.

5. The method according to claim 1, characterized in that, The process of driving the simulation computation layer to perform multi-domain collaborative simulation and generate a multi-dimensional simulation result dataset includes: The kinematic model of the drive support is solved and the basic motion state sensing data is extracted; The hydraulic control model of the drive bracket is solved and the hydraulic system state perception data is extracted. Based on the received component selection encoding, the rigid-flexible coupling model of the dynamic scheduling support is solved and the structural stiffness and deformation sensing data are extracted.

6. The method according to claim 5, characterized in that, The process of dynamically scheduling the rigid-flexible coupling model of the support structure based on the received component selection encoding and extracting structural stiffness and deformation sensing data includes: Based on the component selection encoding, the rigid body model is dynamically invoked or the flexible body model generated by importing the flexible neutral file converted from the finite element model is dynamically invoked through the parameter-driven conditional instantiation mechanism. By automatically reconstructing the mechanical connection relationship between the instantiated flexible body model and adjacent components through node mapping technology, dynamic and seamless switching of computational analysis modes is achieved.

7. The method according to claim 1, characterized in that, The multi-dimensional simulation result dataset is generated based on a unified integration and mapping mechanism for multi-source data. After the simulation instruction data is asynchronously transmitted to the simulation calculation layer via message middleware, the method further includes: Real-time aggregation of data streams from various sub-models generated by multi-domain collaborative simulation; The sub-model data stream is fused with the historical state sequence to construct a unified multi-model state vector, which represents the mapping state of entity objects in the physical layer in the digital space.

8. The method according to claim 1, characterized in that, Receiving the multi-dimensional simulation result dataset returned by the simulation calculation layer through the message middleware includes receiving a thematic data packet published in a progressive logic, the thematic data packet including the following topics: The basic motion state perception topic includes kinematic and basic mechanical data; The topic of hydraulic system status awareness includes operating condition data of hydraulic drive systems; The topic of structural stiffness and deformation perception includes stress distribution and structural deformation displacement information of flexible components; The comprehensive early warning and risk assessment theme includes diagnostic and decision-making information generated by performing multi-dimensional correlation analysis based on multi-model state vectors.

9. The method according to claim 1, characterized in that, The process of executing comprehensive early warning assessment and control command feedback based on the multi-dimensional simulation result dataset includes: The multi-dimensional simulation result dataset is converted into dynamic time series curves, hydraulic system monitoring curves, and multi-dimensional structural state visualization charts for intuitive interactive display. Based on the comprehensive early warning and risk assessment results, when a specified risk indicator is predicted to exceed a preset safety threshold, a flexible component replacement instruction is generated to trigger the simulation calculation layer to perform a secondary fine simulation pre-run of the specific component.

10. The method according to claim 9, characterized in that, The process of converting the multi-dimensional simulation result dataset into dynamic time series curves, hydraulic system monitoring curves, and multi-dimensional structural state visualization charts for intuitive interactive display includes: Real-time stress cloud map generation based on structural flexibility analysis data-driven 3D scene; A quantitative safety margin is displayed by combining stiffness state indicators with structural health indices. The central control dashboard simultaneously displays the overall risk level panel, fault mode diagnostic cards, and a list of maintenance operation suggestions.

Images