Digital twinning method and system for cutting operation production process of tunneling equipment

Abstract

The invention provides a digital twinning method and system for a cutting operation production process of tunneling equipment. The method comprises the following steps: acquiring data such as a cutting roller track, a coal wall position, a cutting boundary and a cutting state based on multi-source sensing of a cutting arm and each moving part of the cutting arm; based on the multi-axis geometrical relationship, the control mechanism and the spatial position of the cutting arm, constructing a digital three-dimensional working scene of the cutting arm, the roadway and the cutting section of the cutting arm; the operation process, the motion trail and the boundary points of the cutting roller in the coal wall section are visually cut in the digital twinning process; on the basis of multi-source data perception, simulation and prediction, cutting roller boundary points are accurately obtained; on the basis of off-line simulation, different working environment data are applied in advance to obtain optimal control parameters; on the basis of online simulation, each control parameter is adjusted in real time to adapt to an uncertain working environment; the service life prediction of the cutting arm and key parts thereof is realized through multi-source data; by means of digital twinning, coal wall section cutting process visualization, cutting operation simulation, cutting part service life prediction and health diagnosis are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tunneling intelligent cutting digital control, and particularly relates to a tunneling equipment cutting operation production process digital twin method and system. BACKGROUND

[0002] Although a digital twin has been established for the tunneling working face, it is more for displaying the action state and action data of each component of the tunneling equipment. For the cutting operation of the tunneling equipment, only the acquisition and remote control of the cutting trajectory of the cutting arm and the operation data associated with cutting have been realized.

[0003] The existing digital twin of the tunneling operation fails to truly reflect the action mechanism simulation of the tunneling equipment, the three-dimensional operation process of the equipment, the relative position of the cutting drum to the coal wall, the cutting-in process of the cutting drum to the coal wall, the change of the coal wall section, the health condition of the equipment, and the like. In addition, in the cutting operation process of the tunneling equipment, the cutting arm is subjected to force, the cutting teeth of the cutting drum are worn, and especially when strong impact load is encountered, the reliability and service life of the cutting arm and the cutting drum will be affected to a certain extent. How to simulate the control parameters in the digital twin to realize optimal economic cutting and control the cutting speed is not involved in the prior art. SUMMARY

[0004] The present application provides a tunneling equipment cutting operation production process digital twin method and system to at least partially solve one of the technical problems in the related art. The technical solution of the present disclosure is as follows: In a first aspect, the present application provides a tunneling equipment cutting operation production process digital twin method. The tunneling equipment includes a machine body, a cutting arm, and a cutting drum connected to the front end of the cutting arm. The cutting arm is located at the front end of the machine body. The method includes the following steps: Acquiring multi-source sensor data collected by a plurality of types of sensors deployed on the tunneling equipment, the plurality of types including device perception type, environment perception type, audio perception type, and video perception type; Based on the multi-source sensor data, acquiring cutting arm state information, cutting drum state information, machine body pose information, relative position information of the cutting arm relative to the roadway space, and cutting arm boundary line information; Based on the cutting arm state information, cutting drum state information, and machine body pose information, controlling the cutting arm three-dimensional digital model to execute the same running state as the cutting arm in the real scene in the digital twin three-dimensional scene; and based on the relative position information of the cutting arm relative to the roadway space and the specific boundary, visualizing the front coal roadway section and the side coal wall in the digital twin three-dimensional scene; the roadway three-dimensional digital model includes a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model; acquire the latest simulation-derived optimal cutting control parameters, based on the latest simulation-derived optimal cutting control parameters, based on the multi-source sensor data and the cutting arm boundary line information, acquire the current cutting control parameters; and control the cutting arm operation based on the current cutting control parameters; wherein the latest simulation-derived optimal cutting control parameters are obtained by simulation through a cutting operation digital simulation control model; Based on the multi-source sensor data, identify the fault information of the tunneling equipment cutting and the health evaluation results of multiple components. In a second aspect, the present application provides a tunneling equipment cutting operation production process digital twin system, the tunneling equipment comprising a machine body, a cutting arm and a cutting drum connected to the front end thereof, the cutting arm being located at the front end of the machine body; the system comprises: A data acquisition module is configured to acquire multi-source sensor data, which is collected by multiple types of sensors deployed on the tunneling equipment, including device perception type, environment perception type, audio perception type and video perception type. A data processing module is configured to acquire cutting arm state information, cutting drum state information, machine body pose information, cutting arm relative position information relative to the tunnel space and cutting arm boundary line information based on the multi-source sensor data. A front-end visualization module is configured to control the cutting arm three-dimensional digital model to execute consistent running state with the cutting arm in the real scene in the digital twin three-dimensional scene based on the cutting arm state information, cutting drum state information and machine body pose information; and visualize the front coal roadway section and side coal wall in the digital twin three-dimensional scene based on the relative position information of the cutting arm relative to the tunnel space and the specific boundary, the cutting arm relative position information relative to the tunnel space and the tunnel three-dimensional digital model; the tunnel three-dimensional digital model comprises a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model. An equipment control module is configured to acquire the latest simulation-derived optimal cutting control parameters, based on the latest simulation-derived optimal cutting control parameters, based on the multi-source sensor data and the cutting arm boundary line information, acquire the current cutting control parameters; and control the cutting arm operation based on the current cutting control parameters; wherein the latest simulation-derived optimal cutting control parameters are obtained by simulation through a cutting operation digital simulation control model. A life prediction module is configured to identify the fault information of the tunneling equipment cutting and the health evaluation results of multiple components based on the multi-source sensor data.

[0005] The tunneling equipment cutting operation production process digital twinning method provided by the application can realize the visualization process of the tunneling equipment in the cutting coal arm operation process, utilize the position relationship between the two coal arms, the front coal roadway section, the tunneling equipment and the coal arm, present the whole dynamic process of the cutting drum cutting into the coal arm in the digital twinning interface, the consistency of the cutting trajectory and the cutting section, and can completely show the coal breaking process of the cutting drum; the optimal cutting control parameter can be obtained through simulation and online adjustment to improve the cutting efficiency and ensure the standardization of the roadway; the cutting arm can be fault recognized and warned, and the health state and service life of the cutting arm structure and key components can be predicted; the simulation, prediction and diagnosis of the cutting arm of the tunneling equipment are realized by using the digital twinning technology.

[0006] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those having ordinary skill in the art upon examination of the following or can be learned from practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0007] The above and / or additional aspects and advantages of the application will become apparent and be readily appreciated from the following description, taken in conjunction with the accompanying drawings, in which: Figure 1 A flowchart of a tunneling equipment cutting operation production process digital twinning method provided by an embodiment of the application; Figure 2 A block diagram of a tunneling equipment cutting operation production process digital twinning system; Figure 3 A flowchart of a tunneling equipment cutting operation production process digital twinning method provided by an example of the application. DETAILED DESCRIPTION

[0008] The embodiments of the application are described in detail below with reference to the accompanying drawings, examples of which are shown in the drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the drawings are exemplary and are intended to explain the application, and cannot be understood as a limitation of the application.

[0009] The tunneling equipment cutting operation production process digital twinning method and system of the embodiments of the application are described below with reference to the accompanying drawings.

[0010] Figure 1 A flowchart of a tunneling equipment cutting operation production process digital twinning method provided by an embodiment of the application. As shown in the figure, the tunneling equipment cutting operation production process digital twinning method includes the following steps: Figure 1 ​Step S101, acquiring multi-source sensor data collected by multiple types of sensors deployed on the tunneling equipment, including device perception type, environment perception type, audio perception type, and video perception type.

[0011] Acquiring multi-source sensor data of the cutting arm in the roadway space collected by multiple sensors deployed on the tunneling equipment and in the working environment of the tunneling equipment.

[0012] It should be noted that the tunneling equipment includes a machine body, a cutting arm, and a cutting drum connected to the front end of the cutting arm. The cutting arm is located at the front end of the machine body, and a sliding frame mechanism is arranged at the bottom of the cutting arm. The rear end of the sliding frame mechanism is connected to the cutting feed sliding rail at the rear end of the cutting arm, and a rotation sensor is arranged at the rear end of the sliding frame mechanism. The cutting arm is driven by a hydraulic cylinder (lifting, telescopic, and rotating cylinder) to realize the actions of lifting, telescoping, and rotating. The cutting drum is driven by a cutting motor to realize its rotating motion, and the end of the cutting drum is installed on a reducer connected to the cutting motor. The tunneling equipment is an underground device, and all sensors installed thereon for collecting monitoring signals of the cutting arm during operation are mine-used sensors, which meet the coal mine explosion-proof requirements. Multiple sensors are processed for explosion-proof and form a multi-source sensor acquisition and processing system according to intrinsic safety and explosion-proof isolation.

[0013] The multi-source sensor data of the cutting arm can include but is not limited to: hydraulic cylinder displacement, hydraulic pressure at the inlet of the hydraulic cylinder, cutting motor operating current, cutting motor power supply voltage, motor winding temperature, motor bearing temperature, reducer oil temperature, reducer lubricating oil pressure, cutting arm vibration signal, cutting motor vibration signal, audio signal, infrared signal, ultrasonic radar signal, millimeter wave radar signal, machine body posture, infrared thermal imager signal, walking track running speed, and methane concentration data.

[0014] As an example, the present embodiment acquires multi-source sensor data of the cutting arm through a multi-source perception layer, which includes: A hydraulic cylinder built-in displacement sensor is used to collect the displacement of the hydraulic cylinder, and a pressure sensor is arranged at the inlet of the hydraulic cylinder to collect the hydraulic pressure at the inlet of the hydraulic cylinder.

[0015] A current sensor and a voltage sensor are installed on the cutting motor to collect the cutting motor operating current and the cutting motor power supply voltage, respectively. The cutting load characteristics can be monitored through the cutting motor operating current.

[0016] A PT100 sensor is arranged near the cutting motor to collect the motor winding temperature and the motor bearing temperature to monitor the temperature field distribution of the motor winding and bearing.

[0017] The vibration accelerometers are arranged on the speed reducer and the cutting arm respectively to collect cutting motor vibration signals and cutting arm vibration signals, so as to monitor the mechanical vibration frequency spectrum of the cutting part through the vibration signals, and monitor the load fluctuation and the equipment health state of the cutting arm according to the obtained vibration frequency spectrum.

[0018] The microphone array or the audio pickup is arranged on the cutting arm to collect audio signals, and the audio signals can be used to identify the difference in high-frequency component attenuation of the coal rock breaking soundprint, so as to assist in verifying the position of the coal rock interface.

[0019] The infrared sensor is arranged on the cutting arm to collect infrared signals, and the infrared signals are used to detect and track the position and form of dust generated when the cutting arm cuts to the adjacent coal rock boundary, so as to identify the information of the coal rock boundary.

[0020] The millimeter wave radar and the ultrasonic sensor are arranged on the cutting arm to collect millimeter wave radar signals and ultrasonic signals respectively, and through the fused millimeter wave radar signals and ultrasonic signals, the dynamic monitoring of the distance between the tunneling equipment and the front and both sides of the coal wall can be realized; for example, the distance between the machine body and the both sides of the coal wall is obtained through the ultrasonic radar signal, and the distance between the machine body and the front of the coal wall is obtained through the millimeter wave radar signal. The IMU is arranged on the machine body to obtain the attitude of the machine body, so as to obtain the motion trajectory of the tunneling equipment according to the attitude of the machine body.

[0021] The infrared thermal imager is arranged on the cutting arm to collect infrared thermal imager signals, and the temperature field distribution of the contact area between the cutting drum and the coal wall can be obtained through the infrared thermal imager signals, so as to obtain the relative distance between the cutting drum and the front cross section of the coal roadway according to the temperature field distribution.

[0022] The laser methane remote sensor is arranged on the cutting arm to collect methane concentration data, so as to realize dynamic monitoring of the gas concentration.

[0023] The infrared thermal imager arranged on the cutting arm collects infrared radiation signals to realize the monitoring of the thermal radiation difference of the coal rock interface.

[0024] Therefore, the present step perceives the change of each moving part of the tunneling equipment, the relative position, perceives the tunneling operation environment including gas, coal rock, and roadway shape through the device layer sensor, the space sensor, and the environment sensor. The sensors are arranged on the tunneling equipment to collect and process, and the processor has the processing ability of multiple categories of signals, forms a process from mechanical change and position change to digitization, and provides a data basis for the simulation and prediction of digital twinning.

[0025] In step S102, based on the multi-source sensor data, the cutting arm state information, the cutting drum state information, the machine body pose information, the relative position information of the cutting arm relative to the roadway space, and the cutting arm boundary line information are obtained.

[0026] In some embodiments, the cutting arm state includes the lifting distance of the lifting cylinder, the telescopic distance of the telescopic cylinder, and the inclination angle of the cutting arm, which are obtained by displacement sensors and angle sensors respectively. The cutting drum state information includes the rotation speed of the cutting motor, which is obtained by a rotation speed sensor. The machine body pose information is obtained by an inertial sensor.

[0027] In some embodiments, based on the multi-source sensor data, the relative position information of the cutting arm relative to the roadway space is obtained, including: based on the millimeter wave radar signal in the multi-source sensor data, the front head reflection signal and the first distance between the machine body or the cutting drum and the two sides of the coal wall are obtained; the dust interference in the front head reflection signal is corrected through an attenuation model to obtain a corrected front head reflection signal; and based on the corrected front head reflection signal, the front coal roadway section information and the boundary of the upper and lower excavated coal seams are obtained; based on the infrared thermal imager signal in the multi-source sensor data, the temperature field distribution and abnormal smoke of the contact area between the cutting drum and the front coal roadway section are obtained; based on the temperature field distribution, the pyroelectric effect is combined to obtain the second distance between the cutting drum and the front coal roadway section; and the first distance and the second distance are taken as the relative position information of the cutting arm relative to the roadway space.

[0028] In some embodiments, based on the multi-source sensor data, the cutting arm boundary line information is obtained, including: based on the displacement sensor data of the lifting cylinder of the cutting arm at the last cutting and the device pose, the cutting boundary line position at the last cutting is obtained; when the distance between the position of the cutting drum and the cutting boundary line position at the last cutting satisfies a first distance threshold, the related data reaching the cutting boundary is obtained; based on the related data reaching the cutting boundary and a preset judgment rule, the cutting boundary line position of the current cutting is determined; and based on the cutting boundary line position of the current cutting, the cutting arm boundary line information is determined.

[0029] It can be understood that the tunneling boundary is the upper and lower boundary lines on the tunnel section when the tunneling equipment performs the cutting operation. Based on the infrared camera device, the audio device, the cutting arm vibration sensor and the cutting current sensor integrated in the tunneling equipment, and in combination with the cutting drum upward or downward cutting boundary identification model, the cutting drum control and visual boundary in the digital twin are determined. When the tunneling equipment tunnels along the top or along the bottom, the cutting drum needs to perceive the cutting boundary line (including the upper cutting boundary line and the lower cutting boundary line) between the coal and the gangue or other hard solids. The method comprises the following steps: in response to the distance between the position of the cutting drum and the cutting boundary line in the last cutting satisfying a first distance threshold, obtaining related data reaching the cutting boundary; and based on the related data reaching the cutting boundary and a cutting boundary identification model, determining the cutting boundary line in the current cutting.

[0030] As an example, when the cutting drum position is a certain distance away from the cutting boundary line in the last cutting, the cutting boundary identification model collects and processes at high speed, the infrared camera automatically adjusts the angle to align with the cutting edge of the cutting drum, the audio pickup device accurately identifies the abnormal sound when cutting coal at this position, and the video, sound, current and vibration are analyzed and processed in multiple modes.

[0031] As a possible implementation, the related data reaching the cutting boundary is obtained; based on the related data reaching the cutting boundary and the abnormal judgment rules of various data, it is identified whether there is boundary abnormal information; if there is, the cutting boundary line in the current cutting is determined according to the position of the cutting drum. As an example, multiple types of data are obtained, including video V, sound S, current I and vibration data; each type of data is classified by a cutting boundary identification model, and a screening algorithm is used to determine whether any type of data has obvious abnormalities according to the abnormal judgment rules corresponding to each type of data; if there is, the cutting boundary line is determined. In addition, each type of data stores the identified abnormal points by a data self-learning method, which serves as the basis for cutting boundary identification, such as determining the abnormal judgment rules.

[0032] According to the single data judgment, there is a problem of low accuracy, which has certain limitations. As another possible implementation, when the increment of the current satisfies a first increment threshold, it is sequentially determined whether the vibration signal, video and sound are abnormal to obtain a judgment result; and whether there is boundary abnormal information is determined according to the judgment result. This method is aimed at the case where any one of the video, sound (or acoustic emission signal), current and vibration is abnormal but the cutting boundary line cannot be clearly determined. The video, sound, current and vibration are fused and processed. First, the current is judged. The current increases, but the cutting boundary identification model is not triggered. Then, the vibration signal, video and sound are sequentially judged, superimposed and identified to determine whether there is boundary abnormal information. This example determines the cutting boundary line by fusing four types of data, which improves the accuracy of the judgment.

[0033] In some embodiments, when the cutting boundary is identified by multi-source information fusion, the D-S evidence theory is used to make an "and" decision on the reliability values of the above four types of data to determine the cutting boundary.

[0034] In some embodiments, the method for obtaining the cutting boundary in the last cutting includes: obtaining the displacement sensor data of the cutting drum lifting cylinder in the last cutting.

[0035] In one example, during the cutting process, the infrared thermal imager data (i.e. video) at the cutting drum, the vibration data of the cutting drum, the audio data when cutting coal, and the current data of the cutting motor (which can reflect the load condition of the cutting motor) are collected in real time; the vibration data and the audio data are filtered to eliminate the influence of mechanical vibration of the equipment; the current data is compensated for load fluctuation; the temperature change rate and the dust form at the cutting point are obtained according to the infrared thermal imager data at the cutting drum; the coal seam hardness and the coal rock condition are identified according to the temperature change rate and the dust form at the cutting point; because the vibration amplitude significantly increases when approaching the cutting edge, the vibration intensity is extracted according to the vibration data of the cutting drum; whether the cutting point is close to the cutting boundary is identified according to the vibration intensity (RMS value); when the cutting point reaches the coal rock interface, a sudden acoustic emission pulse is generated due to brittle fracture, therefore, the coal rock interface can be identified through the high-frequency energy peak value of the acoustic emission signal, so as to identify the cutting boundary; the cutting current fluctuation information is obtained according to the current data of the cutting motor, and whether the rock layer is cut is identified according to the cutting current fluctuation information, so as to determine the cutting boundary.

[0036] In one example, because of the hysteresis of temperature, in the initial stage (0-1 minute), the vibration intensity is extracted according to the vibration data of the cutting drum; the vibration mutation point is identified based on the vibration intensity; the current mutation point is identified based on the current data of the cutting motor; the cutting boundary is preliminarily determined according to the vibration mutation point and the current mutation point; in the stable cutting stage, the judgment based on the infrared thermal imager data at the cutting drum is given priority, and the acoustic emission signal verification is supplemented. This example realizes adaptive identification of the cutting boundary in different stages, and the vibration and current signals are preferred due to the hysteresis of temperature signals.

[0037] As an example, through comprehensive analysis and data learning by infrared, audio, vibration, current, the cutting boundary line is outlined, including: the temperature change and white smoke phenomenon during cutting coal and rock or during compression can be captured in real time by an infrared thermal imager. Experiments show that: the higher the hardness of coal and rock, the higher the cutting temperature and the faster the temperature rise rate; the temperature characteristics can be used to identify the type of coal and rock in the stable cutting stage, but the initial stage needs to be compensated by combining other sensors due to the insignificant temperature change. For example, by combining audio and vibration signals for monitoring, the vibration intensity is positively correlated with the hardness of coal and rock, and is not affected by the cutting time, which is suitable for coal and rock identification in the early stage of cutting and temperature-insensitive scenarios; the audio signals collected by the voiceprint sensor can capture abnormal high-frequency sound waves when coal and rock breaks or the equipment hits rock. In addition, the change of motor operating current of the cutting motor can reflect the load state, such as the increase of cutting resistance caused by the increase of coal and rock hardness, and the synchronous rise of current. Therefore, by combining infrared and vibration signals, the hardness and decomposition degree of coal and rock can be indirectly inverted, and the change of coal and rock characteristics (such as overload) can be distinguished. Therefore, in the early stage of cutting, vibration signals and current signals can be combined as the main guide to identify the cutting boundary when encountering rock, making up for the delay of infrared response; in the stable stage, the infrared temperature characteristics are mainly used, and the voiceprint is used to assist the change of internal structure of coal and rock, that is, to identify the cutting boundary line when encountering rock.

[0038] In some embodiments, based on the above multi-sensor data, a cutting boundary identification model based on an AI model is used to identify the cutting boundary to obtain the position of the cutting boundary line; and a warning can be given according to the position of the cutting boundary line to improve the accuracy and real-time performance of the position of the cutting boundary line. As an example, the AI model can select an industrial mechanism model, and wavelet denoising, EEMD decomposition and other processing are performed on the multi-sensor data to improve the signal-to-noise ratio; and the adaptability of the cutting boundary identification model is optimized in combination with coal mine geological parameters (such as coal and rock composition, water content).

[0039] When identifying the cutting boundary line through the above-mentioned multiple data, different data types are focused on for different scenarios, such as frequent flash temperature or pick wear scenarios, vibration signal weight is improved, and temperature misjudgment is avoided.

[0040] After obtaining the cutting boundary line, i.e., obtaining the upper boundary line and the lower boundary line, the cutting arm boundary line information is determined according to the upper boundary line and the lower boundary line; and the cutting boundary control of the cutting arm is realized according to the cutting arm boundary line information.

[0041] In one example, the distance between the cutting drum of the tunneling equipment and the front coal wall and the two side coal walls is acquired in a millimeter wave radar signal dominant manner, and is calibrated by an ultrasonic signal; the body pose data, the radar data and the body size are fused by an extended Kalman filter to avoid collision. When the ultrasonic signal is disturbed by dust, the millimeter wave radar data is automatically weighted to realize the extended Kalman filter and the optimized joint positioning; the three-dimensional point cloud data acquired by the millimeter wave radar and the body pose data can be fused to establish a conversion matrix of the relative coordinate system of the tunneling equipment to the tunnel section, so as to acquire the spatial position relationship of the tunneling equipment.

[0042] The embodiment can capture the temperature field mutation of the cutting area in real time by the infrared thermal imager, identify the thermal radiation difference of the coal seam and rock interface according to the temperature field mutation, and use it as the preliminary basis for judging the cutting boundary; then, the vibration signal and the current signal are used to identify the synchronous jump of the vibration intensity and the motor operating current when the cutting resistance jumps, and the critical point of the hardness change of coal and rock is jointly marked; the audio signal can also be used to identify the attenuation difference of the high-frequency component of the coal and rock cracking voiceprint, to assist in verifying the position of the coal seam and rock interface; the fuzzy entropy particle swarm algorithm can also be used to optimize the membership function of each sensing signal with the minimum fuzzy entropy as the target, to adapt to different working conditions; the D-S evidence theory "and" decision criterion is constructed to fuse the multi-source sensor data and output the coal cutting ratio reliability value. Based on the identified cutting boundary, the three-dimensional geological model and the real-time IMU data are integrated to pre-play the cutting path and dynamically adjust the height of the cutting drum and the excavation bottom amount; the multi-sensing fusion cutting boundary control model realizes millimeter-level accurate control of the cutting boundary through the "multi-sensing fusion perception-digital twin pre-play-dynamic membership decision-edge rapid response" closed loop, provides core algorithm support for intelligent mining, and improves the accuracy of cutting boundary identification.

[0043] In step S103, based on the cutting arm state information, the cutting drum state information and the body pose information, the cutting arm three-dimensional digital model is controlled to execute the running state consistent with the cutting arm in the real scene in the digital twin three-dimensional scene; and based on the relative position information of the cutting arm relative to the roadway space and the roadway three-dimensional digital model, the front coal roadway section and the side coal wall are visualized in the digital twin three-dimensional scene; the roadway three-dimensional digital model includes a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model.

[0044] As an example, a tunneling equipment three-dimensional digital model is constructed, which includes a machine body three-dimensional digital model and a cutting arm three-dimensional digital model, and the cutting arm three-dimensional digital model is established by combining 3D scanning and Solidwork industry model reconstruction. The fine modeling of the cutting arm is completed by using 3Dmax, and the key moving parts such as the hydraulic rod of the cutting arm are implemented with bone binding and IK inverse kinematics binding to ensure the natural simulation and precise linkage of the hydraulic drive action. When constructing the cutting arm three-dimensional digital model, the cutting arm and its cutting drum are finely modeled, and designed one-to-one with the actual object. The action angle, telescopic length, number of cutting teeth and angle of the cutting drum of the cutting arm are consistent with the actual object. The assembly relationship of the machine body with the cutting arm, the cutting arm with each oil cylinder, the cutting arm with the cutting drum, the cutting drum with the speed reducer and its motor is one-to-one mapped with the actual equipment.

[0045] In some embodiments, the tunneling equipment three-dimensional digital model is imported into the Unity engine, and the real-time sensor data is connected through the Modbus protocol interface. The real space position of the machine body is dynamically updated combined with the positioning information feedback of the field trigger. Based on the preset kinematics constraint algorithm, the running state of the cutting arm is converted into the action instruction of the cutting arm three-dimensional digital model, i.e., the cutting arm digital twin. The real restoration of the position relationship between the cutting arm and the machine body, as well as the real-time visual display of the working state and the telescopic action details of the hydraulic rod of the cutting arm are realized in the Unity scene. To realize the dynamic presentation of the real-time state of the cutting arm, the cutting arm control system periodically refreshes the running state data of the cutting arm through the Modbus protocol, and synchronously updates the position, posture and working parameters of the cutting arm digital twin in the Unity engine combined with the real-time signal of the field trigger. For 3DUI display, interactive 3D information panels are set for key components in the Unity scene, and core data such as hydraulic rod pressure and cutting speed are displayed in real time. Through the space anchoring technology of UI and model, the operator can intuitively obtain the real-time state of each part of the tunneling equipment, and realize the all-round visual presentation of the working state and action details of the tunneling equipment digital twin.

[0046] As an example, a rigid-flexible coupling dynamics model is built by ADAMS software, specifically, a three-dimensional digitized model (SolidWorks model) of the tunneling equipment is imported into the ADAMS software, and constraint conditions such as telescoping, lifting, rotating of the cutting arm, rotating of the cutting drum, attitude of the machine body, telescoping of the stabilizing shoe, etc. are added; the steel parts such as the cutting arm and the cutting drum are imported into ANSYS to generate a modal neutral file, so as to realize the flexibility of the key components; parameters of each component are set, such as displacement of the oil cylinder, weight of the cutting arm, weight of the machine body, characteristics of the bottom plate, shape of the cutting pick of the cutting drum, first feed amount, cutting coal rock coefficient, speed of the cutting drum, etc., and the dynamics modeling is completed. According to the force of the hydraulic system of the cutting arm, the lifting / rotary cylinder thrust is set according to the valve-controlled double-cylinder parameters, the expression of the coal-rock cutting resistance is obtained, and the load force model is completed.

[0047] In this embodiment, in the digital twin three-dimensional scene, the three-dimensional animation of the twin of each component of the tunneling equipment and its associated data can be truly displayed, and the three-dimensional positional relationship between the tunneling equipment and the coal roadway, and between the tunneling cutting arm and the coal roadway; the three-dimensional animation includes walking of the tunneling equipment in the roadway, cutting of the cutting arm into the coal wall, rotating of the cutting drum, cutting drum trajectory, lifting of the cutting arm, telescoping, rotating, loading and transportation of the tunneling equipment, cutting of the coal roadway section into the state, etc.

[0048] In some embodiments, based on the obtained cutting arm boundary line information, the control of the cutting boundary of the digital twin of the cutting arm is guided.

[0049] In one example, based on the cutting boundary line, in combination with the cutting boundary control model, the precise regulation and control of the coal-rock interface are realized through data collaboration and intelligent decision-making; the cutting boundary control model controls the stopping position and speed of the cutting drum according to the cutting arm boundary value, which is also part of the digital twin function.

[0050] It can be understood that when the cutting arm is breaking coal, the cutting arm has up and down movement, and during the up and down movement, it cannot exceed the position of the cutting boundary line; when the cutting arm approaches the cutting boundary line during movement, the cutting arm needs to slow down, that is, the cutting boundary determines the upper and lower boundaries of the movement trajectory of the cutting arm and the speed change trend.

[0051] Therefore, the embodiments of the present application realize the up and down control and the movement speed control of the cutting arm by acquiring multi-sensor data, time stamping and preprocessing the multi-sensor data, and identifying the cutting boundary line after preprocessing. Wherein, when the multi-sensor data is time and space aligned, a hierarchical fusion strategy is adopted.

[0052] In some embodiments, the position information of the tunneling equipment relative to the three-dimensional space of the roadway is taken as a reference point, and a virtual coal wall and a virtual front coal roadway section on both sides of the roadway are constructed based on the first distance and the second distance in combination with a Poisson surface reconstruction algorithm. As an example, ultrasonic sensor arrays are deployed on both sides of the tunneling equipment, and the distance of the coal-rock interface on the side of the roadway is measured in real time according to the obtained time difference of ultrasonic reflection; the millimeter wave radar signal has strong dust penetration capability, and the profile of the roadway is scanned through a 77GHz high-frequency signal, and a three-dimensional point cloud model of the coal wall on both sides of the excavated roadway is generated by combining the Poisson surface reconstruction algorithm. In this example, the ultrasonic signal provides near-field high-precision data, and the millimeter wave radar signal covers the far distance blind area, and the dynamic splicing of the two types of data is realized through Kalman filtering to form a continuous virtual coal wall.

[0053] The embodiment of the present application can realize the visualization process of the tunneling equipment during the cutting coal arm operation process, and the position relationship between the two coal arms, the front coal roadway section, the tunneling equipment, and the coal arm is utilized to present the entire dynamic process of the cutting drum cutting into the coal arm in the digital twin interface, and the consistency between the cutting trajectory and the cutting section can completely exhibit the coal breaking process of the cutting drum.

[0054] In step S104, the latest simulation-optimized cutting control parameter is obtained, and based on the latest simulation-optimized cutting control parameter, the multi-source sensor data, and the cutting arm boundary line information, a current cutting control parameter is obtained, and the cutting arm operation is controlled based on the current cutting control parameter.

[0055] The latest simulation-optimized cutting control parameter is obtained by simulation through the cutting operation digital simulation control model.

[0056] In some embodiments, the method for obtaining the latest simulation-optimized cutting control parameter includes: simulating through the cutting operation digital simulation control model based on different input variables to obtain multiple cutting trajectories and corresponding cutting control parameters; taking the cutting efficiency as the highest target based on the cutting operation at a constant power, the vibration intensity of the cutting drum being less than a vibration intensity threshold, and the cutting arm boundary line information satisfying a preset condition; selecting an optimal cutting trajectory from the multiple cutting trajectories; and taking the cutting control parameter corresponding to the optimal cutting trajectory as the latest simulation-optimized cutting control parameter.

[0057] In some embodiments, the method for obtaining the latest simulation-optimized cutting control parameter includes: Obtaining multi-source sensor data near multiple cutting operations; based on the multi-source sensor data of the multiple cutting operations, obtaining a change trend graph of cutting current, vibration intensity and cutting section with operation time and coal rock coefficient as horizontal coordinates, respectively; determining an optimal cutting trajectory based on the change trend graph of the cutting current, vibration intensity and cutting section; taking the cutting control parameter corresponding to the optimal cutting trajectory as the optimal cutting control parameter obtained by the latest simulation.

[0058] In the embodiment, simulation is performed based on a cutting operation digital simulation control model to obtain optimal cutting control parameters, which include left-right control speed S_horizontal(t), up-down control speed S_vertical(t) and extension control speed S_extension(t), and the mathematical expression of the cutting control parameters is as follows:

[0059] wherein I(t) is the cutting current, which is the real-time monitored motor current value; V(t) is the cutting vibration, which is the device vibration sensor data, k1...k6 are proportional coefficients determined through system identification, and ε1, ε2, ε3 are noise disturbance terms.

[0060] In some embodiments, the cutting control parameters include the up-down speed of the up-down oil cylinder, the extension speed of the extension oil cylinder, the rotation speed of the structure controlling the rotation of the cutting arm, the pose adjustment angle and adjustment speed of the machine body, the start-stop signal of the cutting motor, etc.

[0061] In one example, the cutting trajectory is formed by the joints (oil cylinders) of the cutting arm and the pose movement of the machine body, and is also the running route of the cutting drum, that is, the cutting curve generated by the composite movement of the oil cylinders. That is, the size of the cutting arm and the cutting drum and their positional relationship are determined according to the three-dimensional geometric model of the cutting arm; the relative positional relationship between the cutting drum and the cutting arm is fed back in real time according to the displacement sensor and the position sensor; the position of the cutting arm changes with the change of the machine body; the cutting trajectory of the cutting drum relative to the position of the machine body; thus, the cutting trajectory is determined by the inherent structure of the device, the relative movement of the oil cylinders and the pose change of the machine body. The movement state of each joint of the cutting arm is controlled by the electro-hydraulic control system of the mechanical arm, and the output of the electro-hydraulic control system includes the control speed of the oil cylinder from 0 to the maximum value; the input variables include: hydraulic oil cylinder driving pressure, cutting current, oil cylinder displacement, machine body inclination angle, cutting upper limit or cutting lower limit, coal rock coefficient, etc. By simulating different input variables offline, the required time of each joint oil cylinder, the required time of cutting a section, the relationship between the cutting current and the swing of the machine body, the relationship between the cutting current and the coal rock coefficient, etc. can be obtained, and the optimal cutting control parameters and cutting efficiency can be obtained according to the relationship between the obtained variables.

[0062] Acquire the recent stored manual operation parameters, find the optimal cutting control parameters and the corresponding cutting trajectory, and obtain the best cutting efficiency and the least cutting arm impact.

[0063] In one example, the sensor data of a round operation is acquired, and the data set of several operation cycles is formed according to the periodic cycle operation classification. The relevant data set of each operation cycle is statistically processed, and the data trend chart is output. The digital twin interface presents the change trend chart of each data in a work cycle and the change trend chart of the cutting trajectory with time as the reference, and the optimal cutting trajectory and the corresponding optimal cutting control parameters are selected. The data trend chart refers to the area chart of cutting current, the area chart of vibration, and the area chart of cutting section statistics (the area chart of cutting section statistics can reflect overbreak and underbreak) with operation time and coal rock coefficient as the horizontal coordinates in each operation cycle.

[0064] In one example, ultrasonic sensors and millimeter wave radar sensors are arranged on the machine body, the distance between the machine body and the coal wall of the roadway is acquired according to the ultrasonic sensors and millimeter wave radar sensors at the position, the roadway boundary line is determined according to the distance, and the trajectory of the cutting trajectory that is optimal relative to the roadway boundary.

[0065] In one example, the optimal cutting trajectory and the corresponding optimal cutting control parameters are selected with the target of high cutting efficiency (large cutting footage), infrequent overload of cutting current, vibration signal of cutting drum not exceeding the upper limit vibration index, and no overbreak or underbreak in cutting based on the cutting boundary line. The optimal cutting control parameters are determined by the control system through simulation.

[0066] In some embodiments, based on the optimal cutting control parameters obtained from the latest simulation, the multi-source sensor data and the cutting arm boundary line information, the current cutting control parameters are obtained; including: based on the multi-source sensor data and the cutting arm boundary line information, judging the geological condition mutation, overbreak or underbreak situation and equipment itself problem situation; based on the geological condition mutation, overbreak or underbreak situation and equipment itself problem situation and online adjustment strategy, adjusting the optimal cutting control parameters obtained from the latest simulation to obtain the current cutting control parameters; wherein the online adjustment strategy includes parameter adjustment strategies for different geological condition mutation, overbreak or underbreak situation and equipment itself problem situation. For example, load mutation, coal and rock hardening, overbreak or underbreak of cutting, program interruption, cutting motor and electro-hydraulic system protection, floor fluctuation and other situations, the cutting control parameters are adjusted. In specific implementation, the multi-source sensor data are fused to obtain the key parameters of the cutting arm such as swing angle change, swing speed, cutting load and motor power, the key parameters are compared with the preset threshold, the current working condition type (such as normal coal and rock working condition, hard rock working condition, equipment abnormal working condition, etc.) is identified; and the geological condition of the current cutting area (such as coal and rock layer, interlayer, etc.) is judged according to the cutting arm boundary line information, that is, whether there is overbreak / underbreak risk working condition type; the online adjustment strategy includes a preset adjustment rule library, the preset adjustment rule library includes cutting control parameter adjustment strategies corresponding to different working condition types (such as reducing the three movement speeds of the cutting arm in hard rock working condition; adjusting the swing speed to compensate for the cutting trajectory deviation in overbreak / underbreak working condition, etc.); the target parameter adjustment strategy of the corresponding working condition is obtained from the preset adjustment rule library according to the current working condition type, and the online adjustment of the cutting control parameters is realized according to the target parameter adjustment strategy; after adjustment, the key parameters such as cutting load and swing speed are continuously monitored, if the working condition type changes (such as geological condition mutation), the adjustment is re-adjusted to ensure that the control cutting control parameters are real-time matched with the actual working condition. As an example, the RBF neural network can be used to identify the cutting load signal in real time to improve the working condition judgment accuracy; for the time-varying characteristics of the swing speed control system, the GAGFPID controller is used to realize the rapid adaptive regulation and control of the swing speed. This scheme realizes the adaptive regulation and control ability of the cutting control parameters, realizes the real-time dynamic adjustment of the swing speed, power and other parameters of the cutting arm, and meets the efficient cutting demand under different geological conditions; by identifying the overbreak / underbreak and equipment abnormal working condition type, the overload damage or insufficient cutting accuracy is avoided, and the equipment service life is prolonged; therefore, based on real-time data and preset strategy, this scheme realizes the online accurate adjustment of the cutting control parameters, balances the cutting efficiency and safety, and adapts to complex underground working conditions.

[0067] It can be understood that, during the cutting operation, the cutting control parameters will be affected by changes in load and geological conditions, etc. If there is no influence, the cutting control parameters are executed according to the default optimal cutting control parameters. If there is an influence, the cutting operation automatically changes or corrects the cutting control parameters according to the changes in external conditions.

[0068] In some embodiments, based on the optimal cutting control parameters obtained by the latest simulation, based on the multi-source sensor data and the cutting arm boundary line information, the current cutting control parameters are obtained; including: based on the multi-source sensor data and the cutting arm boundary line information, combining the cutting drum control model, predicting the cutting trajectory of the next cut of the cutting drum and the corresponding cutting control parameters; comparing the cutting trajectory corresponding to the latest simulation obtained optimal cutting control parameters and the cutting trajectory corresponding to the next cut of the cutting drum, determining the cutting control parameters corresponding to the cutting trajectory with higher cutting efficiency as the current cutting control parameters. As an example, based on multi-source sensor data (such as cutting arm each cylinder speed, cutting current, vibration, etc. Information) and cutting arm boundary line information, the cutting drum control model is used to realize prediction; wherein the cutting arm boundary line information specifies the cutting range and other conditions, the multi-source sensor data reflects the current cutting actual state, the cutting drum control model comprehensively considers its kinematics, dynamics and other characteristics, and through these information and model, the cutting trajectory of the next cut of the cutting drum and the appropriate swing speed, cutting power and other cutting control parameters matched therewith are predicted. The cutting trajectory corresponding to the latest simulation obtained optimal cutting control parameters is compared with the cutting trajectory of the next cut of the cutting drum predicted above; here, the simulation obtained cutting control parameters are the parameters that can achieve good cutting effect in theory through simulation analysis of the cutting process under different working conditions; from the comparison result, the cutting control parameters corresponding to the cutting trajectory that can cut more materials in the same time and ensure the cutting quality, that is, the cutting efficiency is higher.

[0069] In some embodiments, the cutting trajectory model includes: Cutting lifting cylinder displacement Y(t) - lifting direction displacement; Cutting left and right cylinder displacement X(t) - horizontal direction displacement; Cutting telescopic cylinder displacement Z(t) - telescopic direction displacement; Cutting trajectory parameters: cutting angle θ(t), cutting line speed cutting speed V(t) and cutting depth D(t); Wherein, ; ,

[0070] Wherein: φ (t) is the cutting angle compensation, d (t) is the zero point offset of the oil cylinder, is the displacement derivative (velocity).

[0071] In some embodiments, the deviation value of the cutting motion trajectory can also be obtained, which is artificially obtained and is a deviation value of a cross section formed by artificial measurement in a work cycle (usually one day) that is not standard or has overbreak or underbreak. The deviation value can be input to the control program, and the program can correct the cutting trajectory. That is, the deviation value of the cutting trajectory can be input to the control parameters of each oil cylinder of the cutting arm and the control parameters of the machine body, and the cutting control parameters can be adjusted according to the deviation amount by using a compensation algorithm.

[0072] It can be understood that, in the embodiments, the motion relationship, position relationship and load feedback of each associated action component in the operation process of the cutting arm are obtained based on the data of each associated action component in the operation process of the cutting arm, the position of the machine body in the roadway, the position of the cutting arm and the machine body, and the position of the cutting arm and the roadway. In the automatic operation process of the cutting arm, the cutting arm executes along the trajectory set by the program according to the sensor feedback and the control amount. The tunneling equipment establishes a monitoring software and a geometric model in the cutting operation process. The geometric model can present the displacement of each actuator of the cutting arm, the output of the driving mechanism, the working state of the cutting arm, and the like, so as to realize the visualization of the cutting operation process. The display interface only reflects the working state of the cutting arm, and the simulation and online simulation are realized in the digital twin according to the data in the operation process of the cutting arm. The simulation, such as in the monitoring system, changes the coal rock coefficient, the oil cylinder control amount, the position of the machine body and the roadway, and the swing amount in the cutting process of the equipment, to determine the control amount of the machine body control and the action of each oil cylinder of the cutting arm. The online simulation optimizes each control amount of the cutting arm through the feedback data in the working process, improves the operation trajectory of the cutting arm, and improves the control accuracy.

[0073] In the embodiments, the actuator of the method is a high-performance controller having communication, analog output, switching output and PWM output functions. The controller is configured with a cutting arm control system for implementing the method. The control system can perform statistics and analysis according to the sensor type, the perception mechanism and the action rule of the tunneling equipment, and form a digital twin acquisition and analysis system for various scenes according to time and space. The system can collect and process data in a master-slave and distributed manner to ensure the real-time of sensor data processing and the authenticity of actual action. The control system has control, simulation and prediction capabilities, can classify, count and analyze the data collected by the sensor to diagnose faults of the tunneling equipment and predict the cutting operation, and can perform synchronous simulation running on the data collected by the sensor in a virtual space to correct the control data of the cutting arm digital twin. In addition, the cutting arm control system has a data storage function, which can be used for recovery and offline analysis of the cutting operation process.

[0074] In the embodiment, the perception mechanism and action law of the tunneling equipment are based on the strong coupling nonlinear dynamics of the heavy hydraulic cutting arm and are obtained by synthesizing various algorithms, for example, the adaptive control algorithm is used to solve the uncertainty problem of the parameters in the model, the sliding mode is used to suppress the disturbance algorithm, the feedforward is used to improve the dynamic response, and a composite controller is formed; for the case of frequent changes in load, the model parameter adaptive control is used, the high-precision repeated trajectory is controlled by the sliding mode iterative learning control, and the uncertainty problem of the dynamic parameters (such as mass and inertia) caused by the load change is solved; by real-time estimation of the parameters and adjustment of the control parameters of the controller, the stability of the system is ensured. Typical applications include adaptive trajectory tracking in task space, and the influence of motor parameter drift is offset by joint angular velocity reference error feedback. The perception mechanism of the tunneling equipment is a process of digitizing the cutting arm geometric model and the working environment, and is also the data that must be relied on to achieve control. The perception mechanism is a process of combining sensor principles and actual perception objects.

[0075] In one example, as shown in Figure 2 The cutting arm related multi-sensor data is acquired by the data collector, and the body pose and environment related multi-sensor data such as ultrasonic radar and millimeter wave radar are acquired by the data processing module; the cutting operation controller acquires the multi-sensor data transmitted by the data collector and the data processing module and transmits it to the cutting operation display control simulation prediction platform to realize cutting operation digital twinning, cutting operation simulation, control and other functions.

[0076] In the embodiment, the body is equipped with millimeter wave radar (front and two sides), ultrasonic array and mine intrinsic safety infrared sensor; the sensor data fusion is performed by the edge computing unit, and the fused data is returned to the tunneling equipment digital twinning platform; the application realizes the full-space digitization of the "two sides-front" of the tunneling working face by complementary advantages of multi-modal sensing, and provides centimeter-level environment perception capability for intelligent cutting control.

[0077] In some embodiments, the cutting operation digital simulation control model includes an action control model of the oil cylinder, and the action control model of the oil cylinder includes an oil cylinder speed acquisition formula, which is used to: based on the cutting arm vibration acceleration, the cutting current and the oil cylinder pressure, in combination with a trained neural network NN, the cutting load is predicted to obtain a cutting load prediction value; Based on the cutting load prediction value, the body angle and the oil cylinder displacement, in combination with a fuzzy inference Fuzzy, a proportional valve control signal is inferred to obtain a proportional valve control signal; Based on the proportional valve control signal, the valve flow gain coefficient, the oil cylinder pressure, the flow-pressure coefficient and the effective area of the oil cylinder piston, the oil cylinder speed is obtained. Specifically: The formula for determining the speed of the oil cylinder is as follows:

[0078] where V is the velocity of the cylinder, Q v is the flow rate of hydraulic oil into the cylinder; A is the effective area of the cylinder piston, which depends on the structure of the cylinder chamber; The linearized model of the flow rate of hydraulic oil Q v is expressed as follows:

[0079] where K q is the flow gain coefficient of the valve, Kc is the flow-pressure coefficient, U is the proportional valve control signal, and P is the cylinder pressure; The proportional valve control signal U is expressed as follows:

[0080] where U is generated by the fuzzy inference engine Fuzzy, L is the cutting load, θ is the body angle, and X is the cylinder displacement; The cutting load L is expressed as follows:

[0081] where L is the cutting load, which is calculated in real time by the neural network NN; Acc is the cutting arm vibration acceleration, I is the cutting current, and P is the cylinder pressure; Based on the cylinder velocity formula, the linearized model of the flow rate of hydraulic oil Q v , the proportional valve control signal U, and the cutting load L, the cylinder action control model is as follows: .

[0082] As an implementation manner, the implementation manner of the fuzzy inference engine Fuzzy includes: Determination of input and output and fuzzification: in a given scenario, the inputs are the cutting load L, the body angle θ, and the cylinder displacement X, and the output is the proportional valve control signal U; first, the fuzzy subsets describing the languages of these input and output variables are determined, such as {NB (negative big), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), and PB (positive big)}, and the domain of discourse (such as [-3, -2, -1, 0, 1, 2, 3], etc.) is set, and then the corresponding membership function (such as a triangular membership function) is selected for the fuzzy language. For example, for the cutting load L, according to the actual value range, it is mapped to the domain of discourse, and the membership degrees in each fuzzy subset (such as the fuzzy subsets corresponding to “large load”, “medium load”, and “small load”) are determined according to the membership function.

[0083] Fuzzy rule making: Fuzzy rule base is made according to practical control experience and domain knowledge. The rules are generally presented in the form of "if-then". For example, "if cutting load L is large load and body angle θ is large angle and cylinder displacement X is small displacement, then proportional valve control signal U is large value". These rules reflect the relationship between different fuzzy state combinations of input variables and fuzzy state of output variable.

[0084] Fuzzy reasoning: Based on the fuzzy results of input variables and fuzzy rule base, reasoning is carried out using reasoning methods such as Mamdani reasoning method, Larsen reasoning method, etc. Taking Mamdani reasoning method as an example, first, the membership degree of input variables is matched with the antecedent of corresponding rules to obtain the activation strength of each rule (generally the minimum value of each antecedent membership degree is taken), then the rule consequent (fuzzy set of output variable) is truncated or scaled according to the activation strength, finally the results of all rules are synthesized to obtain the fuzzy set of output variable.

[0085] Defuzzification: The fuzzy set of output variable obtained by reasoning is converted into precise value as proportional valve control signal U. Common methods include center of gravity method (calculating the horizontal coordinate value corresponding to the center of gravity of the area surrounded by the membership function curve and the horizontal axis as the precise output), maximum membership degree method (taking the element value with the maximum membership degree as the precise output, if there are multiple maximum membership degree elements, further processing such as taking the average value can be performed) and so on.

[0086] As an implementation manner, the implementation manner of neural network NN includes: Basic structure: Generally composed of input layer, hidden layer and output layer. In the context of calculating cutting load L, the input layer receives input signals such as cutting current I, cylinder pressure P, cutting arm vibration acceleration Acc; the hidden layer is composed of multiple neurons, the neurons are connected through weights, the input signals are weighted and summed and processed through activation function (such as Sigmoid function, ReLU function, Tanh function, etc.) to realize feature extraction and transformation of input data; the output layer outputs the calculated cutting load L value.

[0087] Training process: Data preparation: Collect a large amount of sample data containing cutting current I, cylinder pressure P, cutting arm vibration acceleration Acc and corresponding actual cutting load L, and divide it into training set, validation set and test set. The training set is used to train the neural network, the validation set is used to adjust the network parameters to prevent overfitting, and the test set is used to evaluate the performance of the trained network.

[0088] Forward propagation: Input the data in the training set to the neural network, the signal propagates from the input layer to the output layer through the hidden layer layer by layer, and the predicted cutting load value is obtained.

[0089] Loss calculation: Calculate the error between predicted and actual values using a loss function such as mean squared error loss function where n is the number of samples, yi is the actual cutting load value, i is the predicted cutting load value.

[0090] Backpropagation: Propagate the error from the output layer to the input layer according to the gradient of the loss function, and use gradient descent methods such as stochastic gradient descent (SGD), adaptive moment estimation (Adam), etc. optimization algorithm to update the weights and biases between neurons in the network to reduce the loss.

[0091] Iterative training: Repeat the forward propagation, loss calculation and backpropagation process, constantly adjust the weights and biases until the loss function converges to a certain degree or reaches the preset training times, etc. stopping conditions. After training, use the test set to evaluate the performance of the neural network, such as calculating the root mean square error (RMSE), mean absolute error (MAE), etc. indicators to measure the prediction accuracy. As an example, combined with MATLAB / Simulink software, develop trajectory tracking control algorithm and hydraulic system simulation algorithm for cutting operation, such as trajectory tracking control through adaptive iterative learning controller. Build a proportional valve control double cylinder model of hydraulic oil cylinder, analyze the flow-pressure characteristics, integrate PID controller to optimize response speed, and get the hydraulic system simulation model.

[0092] As an example, after getting the cutting operation digital simulation control model, perform motion performance verification, automatic cutting verification and collaborative optimization process; when performing motion performance verification, set working condition parameters, equipment parameters and working parameters, monitor key parameters, and predict and optimize control the control performance and fatigue life of the cutting arm during cutting operation. When performing automatic cutting verification, import the coal and rock layer distribution map, test the adaptive adjustment ability of the cutting path, and evaluate the indicators according to the cutting contour error and vibration amplitude mutation rate; when performing collaborative optimization process, if ANSYS stress analysis finds that the cutting arm has deformation exceeding the limit, return to SolidWorks to modify the three-dimensional model, and then re-import ADAMS to verify the dynamic performance.

[0093] In summary, the cutting operation digital simulation control model is obtained based on the structure, mechanism, control principle, tunneling process and working environment of the cutting arm.

[0094] In some embodiments, by analyzing the spatio-temporal data synthesized by the equipment perception, environment perception, space perception, and audio perception of the tunneling equipment during the cutting operation, and by fusing a given amount of control, a comprehensive knowledge graph that integrates the action mechanism of the tunneling equipment, the execution sequence of each component, and the field working conditions is established based on various data with time axis as the reference, and a sample set can be established based on the knowledge graph.

[0095] In some embodiments, the method for constructing a cutting operation digital simulation control model comprises: obtaining training sample data based on historical operation data of the cutting arm; the historical operation data comprises displacement of each actuator of the cutting arm, driving source pressure of each actuator of the cutting arm, cutting load, cutting boundary line, and operation trajectory data of one cycle of cutting during the cutting operation of the cutting arm; and obtaining a feature-level fused motion control model of the cutting arm by training a particle swarm algorithm optimized deep belief network constructed by using the training sample data. The PTPv2 protocol is used to time-align the multi-source sensor data, and the aligned multi-source sensor data is packed in a 10 ms period sliding time window to obtain sample data; a feature-level fused model is obtained by training a particle swarm algorithm optimized deep belief network constructed by using the sample data; the particle swarm algorithm optimized deep belief network can fuse vibration features, current features, temperature features, and acoustic emission features; the particle swarm algorithm optimized deep belief network introduces a long short-term memory network to process the time sequence correlation of the acoustic emission signal; as an example, the particle swarm algorithm optimized deep belief network is a BR-PSO-DBN network. Wherein, DBN stands for Deep Belief Network, which is a special architecture of deep learning model, and can extract deep features of sensor data layer by layer through multi-layer restricted Boltzmann machine (RBM) stacking; PSO (Particle Swarm Optimization) optimizes the weights, biases, and network structure (such as the number of layers and neurons) of the DBN through global optimization to avoid the gradient descent falling into local optimum; BR (Bayesian Regularization) introduces a regularization term in the loss function to suppress overfitting and enhance generalization ability.

[0096] Step S105, based on the multi-source sensor data, identifying fault information of the cutting of the tunneling equipment and health assessment results of a plurality of components.

[0097] In some embodiments, based on the multi-source sensor data, the service life of the cutting arm structure and its key components is predicted in combination with an equipment health degree assessment model.

[0098] In some embodiments, the frequency spectrum features of the cutting arm vibration signal and the cutting motor vibration signal are extracted, and based on the frequency spectrum features and a prediction judgment rule, the fault information of the tunneling equipment is obtained and a warning is given; specifically, the frequency spectrum features (such as frequency, amplitude, time-frequency feature, etc.) are extracted from the cutting arm vibration signal and the cutting motor vibration signal; the prediction judgment rule means that the frequency spectrum features are compared with a standard frequency spectrum feature library and the fault information is identified according to the comparison result; that is, different fault types often correspond to specific frequency spectrum feature patterns. For example, if the amplitude of the cutting motor vibration signal abnormally increases at a certain specific frequency, it may indicate that there is an imbalance, looseness or other fault in the motor. The extracted frequency spectrum features are compared with the pre-set standard frequency spectrum feature library (containing frequency spectrum features corresponding to various known fault types), and when the similarity exceeds a certain threshold, the corresponding fault type is judged, so that the fault information is generated according to the fault type and a warning is given. In some embodiments, the change trend of the cutting motor operating current is counted, and based on the change trend of the cutting motor operating current, the fault information of the tunneling equipment is obtained and a warning is given. In some embodiments, the temperature change trend and amplitude information are obtained from the motor winding temperature and the cutting motor vibration signal respectively, based on the temperature change trend and the amplitude information, and in combination with a motor bearing fault judgment rule, the motor bearing fault information is identified and a warning is given; specifically, the motor bearing fault judgment rule includes: Temperature change trend judgment: if the motor winding temperature rises sharply in a short time and exceeds the normal working temperature range, it may indicate that the motor bearing has a fault, which causes increased friction and in turn causes temperature rise. At this time, the amplitude information is further judged in combination.

[0099] Amplitude information judgment: according to the amplitude threshold, the motor bearing fault information is identified; when the bearing has a fault, its vibration amplitude usually increases, for example, the faults such as rolling body wear and raceway damage will cause the radial or axial amplitude of vibration to exceed the normal range; by setting different amplitude thresholds, when the amplitude exceeds the corresponding threshold, in combination with the temperature change trend, if the temperature also abnormally rises, it is judged that the motor bearing may have a fault, and a warning is given.

[0100] In some embodiments, based on the cutting motor operating current, the cutting motor vibration signal, the motor winding temperature and the motor bearing temperature, in combination with a cutting arm key component deterioration identification model, the fault information of the cutting arm is obtained and a warning is given.

[0101] It can be understood that the load characteristics can be monitored according to the change of the motor operating current. If the rotating device of the cutting arm deteriorates, the motor operating current fluctuation is intensified. When the bearing wears, the current suddenly changes, and the high-frequency component in the frequency spectrum significantly increases, reflecting the abnormal mechanical resistance of the cutting transmission system. Through the motor current signature analysis method (MCSA), the rotor air gap magnetic field change can be captured non-contact, and the winding inter-turn short circuit of the cutting motor can be identified. Through Fourier transform, the fundamental wave and the harmonic wave are separated to extract the fault characteristic. The current amplitude of the bearing damage characteristic frequency (such as BPFO / BPFI) of the transmission system increases. Based on the analysis of the vibration signals of the cutting arm and the cutting motor, the fault information of the tunneling equipment can be identified. For example, in the early stage of bearing wear, the energy ratio in the frequency band of 100-500Hz increases by 40%; when the bearing deteriorates seriously, the frequency energy above 800Hz increases sharply, and the time domain kurtosis index exceeds the threshold. The fault of the gearbox is characterized by abnormal meshing frequency sideband, and the damaged tooth position can be located by wavelet packet decomposition. When the transient impact vibration occurs in the deteriorated components (such as cracked bearings), the peak factor is greater than 5, and the early warning is triggered.

[0102] In one example, the cutting motor fault can be identified and temperature trend warning can be given through joint analysis of motor operating current and motor winding temperature. For example, when the motor winding is locally short-circuited, the temperature of the specific phase current is abnormally focused.

[0103] In one example, the motor bearing fault can be identified through joint analysis of cutting motor vibration signals and motor winding temperature. For example, when the motor bearing is out of oil, the vibration energy increases and the temperature rises, which is different from the synchronous temperature and vibration mutation of wear.

[0104] In this embodiment, the deterioration of the key components of the cutting arm can be identified through analysis of current, vibration and temperature signals, and the cutting arm can be identified and warned.

[0105] In some embodiments, based on multi-source sensor data, the working condition of the cutting operation is predicted; according to the prediction result of the working condition of the cutting operation, the health status and the life of the cutting arm structure and its key components are predicted.

[0106] It can be understood that the working condition prediction of the cutting operation is to predict the working condition of the cutting operation through the load parameters and cutting states obtained during the cutting operation, so as to infer the health status and life of the cutting arm and its key components.

[0107] In some embodiments, multi-sensor data is obtained, the multi-sensor data is time-stamped, and multiple feature data is extracted from the aligned multi-sensor data; according to the multiple feature data and the equipment health degree evaluation model, the health degree evaluation result of the equipment is obtained, and the life prediction of the tunneling equipment and its key components is realized.

[0108] The tunneling equipment cutting operation production process digital twin method of the embodiment of the application can realize the visualization process of the tunneling equipment in the cutting coal arm operation process, utilize the position relationship of the two-side coal arms, the front coal roadway section, the tunneling equipment and the tunneling equipment and the coal arm, present the whole dynamic process of the cutting drum cutting into the coal arm in the digital twin interface, the consistency of the cutting trajectory and the cutting section, and can completely show the coal breaking process of the cutting drum; the optimal cutting speed of the cutting arm can be obtained, the next-cutting trajectory of the cutting drum and the corresponding cutting control parameters thereof can be predicted; the next-cutting trajectory of the cutting drum and the corresponding cutting control parameters thereof are optimized to improve the cutting precision and ensure the standardization of the roadway; the cutting arm can be fault-identified and pre-warned, and the health state and the service life of the cutting arm structure and the key components thereof can be predicted; the digital twin technology is used to realize the simulation, prediction and diagnosis of the cutting arm of the tunneling equipment.

[0109] In order to clearly illustrate the whole process of the cutting operation process control based on the digital twin, the specific examples are used for illustration. Figure 3 The flow chart of the tunneling equipment cutting operation production process digital twin control method provided in the present example is shown in FIG. 1. Figure 3 As shown in FIG. 1, the tunneling equipment cutting operation production process digital twin control method of the application includes the following steps: cutting operation process digital twin simulation model creation; setting load and environment input parameters from geological data; obtaining optimal control parameters of the cutting operation process; starting the cutting operation; cutting mechanism action, monitoring multi-source data back transmission; simulation model synchronous operation, optimizing control parameters; cutting parameter compensation body posture offset; cutting process and trajectory, three-dimensional display of both sides of the roadway; obtaining cutting boundary; one cutting cycle is completed; cutting process data storage and analysis.

[0110] In order to realize any of the above embodiments, the application provides a tunneling equipment cutting operation production process digital twin system, which comprises: A data acquisition module is configured to acquire multi-source sensor data collected by a plurality of types of sensors deployed on the tunneling equipment, wherein the plurality of types of sensors include device perception type, environment perception type, audio perception type and video perception type. A data processing module is configured to acquire cutting arm state information, cutting drum state information, body pose information, relative position information of the cutting arm relative to the roadway space and cutting arm boundary line information based on the multi-source sensor data. The front-end visualization module is configured to control a cutting arm three-dimensional digital model to perform a consistent operation state with the cutting arm in a real scene in a digital twin three-dimensional scene based on the cutting arm state information, the cutting drum state information, and the machine body pose information; and visualize a front coal roadway section and a side coal wall in the digital twin three-dimensional scene based on the relative position information of the cutting arm relative to a roadway space and a roadway three-dimensional digital model; the roadway three-dimensional digital model includes a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model. The device control module is configured to obtain the latest simulation-derived optimal cutting control parameter, obtain a current cutting control parameter based on the latest simulation-derived optimal cutting control parameter, the multi-source sensor data, and the cutting arm boundary line information, and control the cutting arm operation based on the current cutting control parameter; wherein the latest simulation-derived optimal cutting control parameter is obtained through simulation of a cutting operation digital simulation control model. The life prediction module is configured to identify fault information of the tunneling equipment cutting and health evaluation results of a plurality of components based on the multi-source sensor data.

[0111] The tunneling equipment cutting operation production process digital twinning method and system can construct a visualized real operation scene of the cutting arm drum in the coal roadway section, a digitalized cutting arm control based on variable load and roadway boundary and a simulation model thereof, and a life prediction model based on multi-source data, to provide optimal control parameters and a real three-dimensional working scene for automatic operation of the cutting arm, improve cutting section control precision and prolong the working life of the cutting arm. The state information of each joint arm, working data of the cutting drum, posture of the cutting arm and machine body, and working environment data are obtained; based on the multi-source sensor data, a digitalized three-dimensional mechanical model of the cutting arm, a cutting arm control and simulation model, a cutting arm operation data mining self-diagnosis model, and a roadway section model, a real working scene of the cutting arm in the roadway is obtained, and the running track of the cutting drum in the coal wall section is realized by constructing virtual coal walls on both sides of the roadway and the coal roadway section in front; the operation process of the tunneling equipment cutting arm in the coal wall section and the positional relationship between the tunneling equipment machine body and the coal wall are visualized in the digital twinning virtual scene; based on the cutting arm operation data and its working environment, the cutting arm control amount is obtained in combination with the variable load amount, the load impact amount on the cutting arm, and the dynamic cutting boundary value, to achieve optimal control speed and trajectory change and reduce the impact of the load on the cutting arm; based on the load and geological condition uncertainty and the heavy load hydraulic mechanical arm with large inertia, an online simulation digital model integrating the joint arm, the load, the boundary condition, and the control model is established, each control parameter is adjusted in real time, and optimal control is achieved. According to the simulation digital model, different parameters are set for the model to determine the optimal control parameters under the condition. Based on the equipment self-data and coupled environment data record samples in the cutting arm operation process, accumulation and analysis are performed to realize life prediction of the cutting arm and key moving parts thereof.

[0112] In the foregoing embodiment descriptions, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0113] In addition, the terms "first", "second", etc. are used only for the purpose of description, and should not be understood as indicating or implying relative importance or a specific number of the technical features indicated. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically limited.

[0114] Although the embodiments of the present application have been shown and described above, it is understood that the above-described embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above-described embodiments within the scope of the present application.

Claims

1. A method for a digital twin of a production process of a cutting operation of a mining equipment, the mining equipment comprising a machine body, a cutting arm and a cutting drum connected to a front end of the cutting arm, the cutting arm being located at a front end of the machine body; characterized in that, The method comprises the following steps: acquiring multi-source sensor data collected by multiple types of sensors deployed on the tunneling equipment, the multiple types including device perception type, environment perception type, audio perception type, and video perception type; based on the multi-source sensor data, acquiring cutting arm state information, cutting drum state information, machine body pose information, cutting arm relative position information relative to the tunnel space, and cutting arm boundary line information; based on the cutting arm state information, cutting drum state information, and machine body pose information, controlling a cutting arm three-dimensional digital model to perform a consistent running state with the cutting arm in the real scene in a digital twin three-dimensional scene; and based on the cutting arm relative position information relative to the tunnel space and a tunnel three-dimensional digital model, visualizing the front coal roadway section and the side coal wall in the digital twin three-dimensional scene; the tunnel three-dimensional digital model comprises a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model; acquiring the latest simulation-derived optimal cutting control parameters, obtaining current cutting control parameters based on the latest simulation-derived optimal cutting control parameters, based on the multi-source sensor data, and based on the cutting arm boundary line information; and controlling the cutting arm operation based on the current cutting control parameters; wherein the latest simulation-derived optimal cutting control parameters are obtained through simulation of a cutting operation digital simulation control model; based on the multi-source sensor data, identifying fault information of the tunneling equipment cutting and health evaluation results of multiple components.

2. The method of claim 1, wherein, The multi-source sensor data includes hydraulic cylinder displacement, hydraulic cylinder inlet hydraulic pressure, cutting motor operating current, cutting motor speed, cutting motor power supply voltage, motor winding temperature, motor bearing temperature, reducer oil temperature, reducer lubricating oil pressure, cutting arm vibration signal, cutting motor vibration signal, audio signal, infrared signal, acoustic emission signal, ultrasonic radar signal, millimeter wave radar signal, machine body attitude, and infrared thermal imager signal.

3. The method of claim 2, wherein, The method comprises the following steps: based on the millimeter wave radar signal in the multi-source sensor data, acquiring a front head-on reflection signal and a first distance between the machine body or the cutting drum and the two side coal walls; correcting dust interference in the front head-on reflection signal through an attenuation model to obtain a corrected front head-on reflection signal; and based on the corrected front head-on reflection signal, acquiring front coal roadway section information and the boundary of the upper and lower tunneling coal seams; based on the infrared thermal imager signal in the multi-source sensor data, acquiring temperature field distribution and abnormal smoke of the contact area between the cutting drum and the front coal roadway section; based on the temperature field distribution, acquiring a second distance between the cutting drum and the front coal roadway section in combination with the pyroelectric effect; taking the first distance and the second distance as the relative position information of the cutting arm relative to the tunnel space and a specific boundary.

4. The method of claim 1, wherein, The method comprises the following steps: Obtaining a cutting boundary line position at the last cutting based on displacement sensor data of a lifting cylinder of a cutting arm at the last cutting and a device posture; When a distance between a position of the cutting drum and the cutting boundary line position at the last cutting meets a first distance threshold, obtaining related data of reaching a cutting boundary; Determining a cutting boundary line position at the current cutting based on the related data of reaching the cutting boundary and a preset judgment rule; Determining the cutting arm boundary line information based on the cutting boundary line position at the current cutting.

5. The method of claim 4, wherein, The method for obtaining the latest simulation optimal cutting control parameter comprises: Based on different input variables, a plurality of cutting trajectories and corresponding cutting control parameters are obtained by simulating the cutting operation digital simulation control model; The optimal cutting trajectory is selected from the plurality of cutting trajectories based on the cutting drum operating at a constant power, the vibration intensity of the cutting drum being less than a vibration intensity threshold, and the cutting arm boundary line information meeting a preset condition, with the highest cutting efficiency as the target; The cutting control parameter corresponding to the optimal cutting trajectory is taken as the latest simulation optimal cutting control parameter.

6. The method of claim 1, wherein, The method for obtaining the latest simulation optimal cutting control parameter comprises: Obtaining multi-source sensor data of adjacent multiple cutting operations; Based on the multi-source sensor data of the multiple cutting operations, a cutting current, a vibration intensity and a cutting cross-section trend graph are obtained with operation time and a coal rock coefficient as the horizontal coordinates, respectively; Based on the cutting current, the vibration intensity and the cutting cross-section trend graph, an optimal cutting trajectory is determined; The cutting control parameter corresponding to the optimal cutting trajectory is taken as the latest simulation optimal cutting control parameter.

7. The method of claim 1, wherein, The current cutting control parameter is obtained based on the latest simulation optimal cutting control parameter, the multi-source sensor data and the cutting arm boundary line information; comprising: Based on the multi-source sensor data and the cutting arm boundary line information, a geological condition mutation, overbreak or underbreak and equipment self-problem are judged; Based on the geological condition mutation, overbreak or underbreak, equipment self-problem and an online adjustment strategy, the latest simulation optimal cutting control parameter is adjusted to obtain the current cutting control parameter; wherein the online adjustment strategy comprises a parameter adjustment strategy for different geological condition mutations, overbreak or underbreak and equipment self-problems.

8. The method of claim 1, wherein, The current cutting control parameter is obtained based on the latest simulation optimal cutting control parameter, the multi-source sensor data and the cutting arm boundary line information; comprising: Based on the multi-source sensor data and the cutting arm boundary line information, a next-cut cutting trajectory of the cutting drum and a corresponding cutting control parameter are predicted by combining a cutting drum control model; By comparing the cutting trajectory corresponding to the latest simulation optimal cutting control parameter and the cutting trajectory corresponding to the next-cut of the cutting drum, the cutting control parameter corresponding to the cutting trajectory with higher cutting efficiency is determined as the current cutting control parameter.

9. The method of claim 2, wherein, The method comprises the following steps: Based on the multi-source sensor data, the health assessment results of the components and the fault information of the tunneling equipment cutting are identified; Based on the multi-source sensor data, the service life of the cutting arm structure and its key components is predicted by combining with the equipment health degree evaluation model; The frequency spectrum features of the cutting arm vibration signal and the cutting motor vibration signal are extracted, and the fault information of the tunneling equipment is obtained and early warning is performed based on the frequency spectrum features and the prediction judgment rules; The change trend of the cutting motor operating current is counted, and the fault information of the tunneling equipment is obtained and early warning is performed based on the change trend of the cutting motor operating current; The temperature change trend and amplitude information are obtained from the motor winding temperature and the cutting motor vibration signal respectively, and the motor bearing fault information is identified and early warning is performed based on the temperature change trend and amplitude information and in combination with the motor bearing fault judgment rules; 10. The method of claim 2, wherein, Based on the cutting motor operating current, the cutting motor vibration signal, the motor winding temperature and the motor bearing temperature, the fault information of the cutting arm is obtained and early warning is performed in combination with the cutting arm key component deterioration identification model. The cutting operation digital simulation control model comprises an action control model of the oil cylinder, and the action control model of the oil cylinder comprises an oil cylinder speed acquisition formula, which is used for: Based on the cutting arm vibration acceleration, the cutting current and the oil cylinder pressure, the cutting load is predicted by combining with a trained neural network (NN), and a cutting load prediction value is obtained; Based on the cutting load prediction value, the machine body angle and the oil cylinder displacement, a proportional valve control signal is inferred by combining with a fuzzy inference device (Fuzzy), and the proportional valve control signal is obtained; 11. A digital twin system for production process of a cutting operation of a mining equipment, the mining equipment comprising a machine body, a cutting arm and a cutting drum connected to a front end of the cutting arm, the cutting arm being located at a front end of the machine body; characterized in that, Based on the proportional valve control signal, the valve flow gain coefficient, the oil cylinder pressure, the flow-pressure coefficient and the effective area of the oil cylinder piston, the oil cylinder speed is obtained. The system comprises: A data acquisition module is configured to acquire multi-source sensor data collected by various types of sensors deployed on the tunneling equipment, wherein the various types of sensors include device perception type, environment perception type, audio perception type and video perception type; A data processing module is configured to acquire cutting arm state information, cutting drum state information, machine body pose information, relative position information of the cutting arm relative to the roadway space and specific boundaries, cutting arm relative position information relative to the roadway space and cutting arm boundary line information based on the multi-source sensor data; A front-end visualization module is configured to control a cutting arm three-dimensional digital model to perform a running state consistent with the cutting arm in a real scene in a digital twin three-dimensional scene based on the cutting arm state information, the cutting drum state information and the machine body pose information, and visualize a front coal roadway section and a side coal wall in the digital twin three-dimensional scene based on the cutting arm relative position information relative to the roadway space and a roadway three-dimensional digital model, wherein the roadway three-dimensional digital model comprises a cutting section three-dimensional digital model and a side coal wall three-dimensional digital model. An equipment control module is configured to obtain the latest simulation-derived optimal cutting control parameter, obtain a current cutting control parameter based on the latest simulation-derived optimal cutting control parameter, based on the multi-source sensor data and the cutting arm boundary line information, and control the cutting arm operation based on the current cutting control parameter; wherein the latest simulation-derived optimal cutting control parameter is obtained through simulation of a cutting operation digital simulation control model. A life prediction module is configured to identify fault information of the tunneling equipment cutting and health assessment results of a plurality of components based on the multi-source sensor data.

Images