Intelligent path planning method and system for coal mining robot in mine

Abstract

The application belongs to the technical field of intelligent path planning of coal mining robots, and specifically provides a mine intelligent path planning method and system for a coal mining robot, which mainly comprises the following steps: obtaining camera original images, contact sensor geological data and inertial measurement unit original attitude data, and preprocessing to obtain three-dimensional vision data, geological feature data and corrected attitude data; obtaining historical working face data and equipment attitude information, combining the three-dimensional vision data, the geological feature data and the corrected attitude data, and performing environment perception cooperative correction processing to obtain environment model data; obtaining motor encoder original walking distance and external positioning system original displacement data, and performing walking slip state monitoring processing to obtain a dynamic slip rate. The application can realize accurate correction of environment perception and real-time compensation of walking slip, effectively control the deviation of the actual cutting track and the theoretical path within the safety threshold, and improve the coal mining efficiency and operation safety.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent path planning of coal mining robots, and particularly relates to a method and system for intelligent path planning of a coal mining robot in a mine. BACKGROUND

[0002] In coal mining operations in a mine, the path planning of a coal mining robot relies on multiple sensing technologies, and a visual sensor is often used to obtain a working face image, a contact sensor is used to collect geological parameters, and an inertial measurement unit is used to monitor the attitude of the robot, so as to construct an environment model and plan an operation path.

[0003] However, the mine environment has high dust characteristics, and coal dust is easy to adhere to the lens of the visual sensor, resulting in distortion of image information and affecting the accuracy of environmental perception. When the walking mechanism of the coal mining robot operates on a loose coal gangue layer, dynamic slippage phenomenon is easy to occur, so that the deviation between the actual operation trajectory of the robot and the theoretical path may exceed the safety range, affecting the coal mining efficiency and operation safety. SUMMARY

[0004] The present application provides a method and system for intelligent path planning of a coal mining robot in a mine, which effectively solves the environmental perception deviation caused by distortion of the visual sensor in a high dust environment, and the dynamic slippage of the walking mechanism on a loose coal gangue layer. The existing method cannot compensate for the trajectory deviation caused by the two factors exceeding the safety range, and the present application realizes accurate correction of environmental perception and real-time compensation of walking slippage, effectively controls the deviation between the actual cutting trajectory and the theoretical path within the safety threshold, and improves the coal mining efficiency and operation safety.

[0005] In order to achieve the above purpose, the application adopts the following technical scheme:

[0006] In a first aspect, the present application provides a method for intelligent path planning of a coal mining robot in a mine, comprising:

[0007] Obtain the original image of the camera, the geological data of the contact sensor and the original attitude data of the inertial measurement unit, and perform preprocessing to obtain three-dimensional visual data, geological feature data and corrected attitude data.

[0008] Obtain historical working face data and equipment attitude information, and perform environment perception collaborative correction processing on the three-dimensional visual data, the geological feature data and the corrected attitude data to obtain environment model data.

[0009] Obtain the original walking distance of the motor encoder and the original displacement data of the external positioning system, and perform walking slippage state monitoring processing to obtain a dynamic slippage rate.

[0010] The preset cutting template is acquired, historical working face geological data in historical working face data, the corrected attitude data, the environment model data and the dynamic slip rate are combined, path initial planning and slip compensation processing are performed, and optimized path data is obtained.

[0011] Based on the optimized path data, safety conflict detection and path adjustment processing are performed, path verification flag data and adjusted path data are obtained.

[0012] Based on the path verification flag data and the adjusted path data, path output and execution control processing are performed, and actual path execution effect is obtained.

[0013] Further, preprocessing is performed to obtain three-dimensional visual data, geological feature data and corrected attitude data, including:

[0014] The original image of the camera is subjected to coal dust adhesion interference removal processing to obtain a preliminary clean image.

[0015] The preliminary clean image is subjected to edge sharpening enhancement processing to obtain enhanced visual data.

[0016] The enhanced visual data is subjected to stereoscopic depth reconstruction processing to obtain three-dimensional visual data.

[0017] The rock stratum feature extraction processing is performed on the contact type sensor geological data, and the zero point drift correction processing is performed on the original attitude data of the inertial measurement unit, to obtain the geological feature data and the corrected attitude data.

[0018] Further, environment perception collaborative correction processing is performed to obtain environment model data, including:

[0019] The three-dimensional point cloud reconstruction processing is performed on the three-dimensional visual data and the geological feature data to obtain environment point cloud data.

[0020] Based on the environment point cloud data and historical working face data, coal-rock interface mode recognition processing is performed to obtain interface recognition data.

[0021] The pose collaborative registration processing is performed on the corrected attitude data and the interface recognition data to obtain compensated perception data.

[0022] Based on the compensated perception data and device attitude information, dynamic reference surface construction processing is performed to obtain environment model data.

[0023] Further, walking slip state monitoring processing is performed to obtain a dynamic slip rate, including:

[0024] The original walking distance of the motor encoder is subjected to cumulative error reset processing to obtain reset distance data.

[0025] The reset distance data is subjected to instantaneous speed deduction processing, and the original displacement data of the external positioning system is subjected to multipath effect suppression processing to obtain theoretical walking speed data and actual displacement data.

[0026] The theoretical walking speed data and the actual displacement data are subjected to slip rate dynamic estimation processing to obtain a dynamic slip rate.

[0027] Further, path initial planning and slip compensation processing are performed to obtain optimized path data, including:

[0028] Based on the environment model data and a preset cutting template, initial path generation processing is performed to obtain theoretical cutting path data.

[0029] Based on interface recognition data and historical working face geological data, working face geological feature partitioning processing is performed to obtain geological partitioning map data.

[0030] Based on the dynamic slip rate and the geological partitioning map data, slip influence domain partitioning processing is performed to obtain regionalized slip parameter data.

[0031] The regionalized slip parameter data is subjected to trend prediction processing to obtain slip prediction data.

[0032] The theoretical cutting path data and the slip prediction data are subjected to path point cooperative compensation processing to obtain compensated path data.

[0033] The compensated path data and the corrected attitude data are subjected to dynamic attitude constraint optimization processing to obtain optimized path data.

[0034] Further, based on the optimized path data, safety conflict detection and path adjustment processing are performed to obtain path verification flag data and adjusted path data, including:

[0035] The optimized path data is subjected to safety distance conflict scanning processing to obtain conflict marker data.

[0036] The conflict marker data is subjected to local path re-planning processing to obtain adjusted path data.

[0037] The adjusted path data is subjected to efficiency evaluation processing to obtain path score data.

[0038] The path score data is subjected to executability threshold checking processing to obtain path verification flag data.

[0039] Further, based on the path verification flag data and the adjusted path data, path output and execution control processing are performed to obtain actual path execution effect, including:

[0040] Perform instruction conversion processing based on the path verification mark data and the adjusted path data to obtain control instruction data.

[0041] Perform timing scheduling allocation processing on the control instruction data to obtain timing instruction sequence data.

[0042] Perform execution mechanism compatibility adaptation processing on the timing instruction sequence data to obtain adapted control data.

[0043] Obtain actual path execution effect by sending the adapted control data to a robot driving unit.

[0044] Further, perform dynamic pose constraint optimization processing on the compensated path data and the corrected pose data to obtain optimized path data, including:

[0045] Perform dynamic pose constraint modeling to generate constraint rule data.

[0046] Perform iterative optimization calculation on the compensated path data based on the constraint rule data to obtain optimized path data.

[0047] Further, perform instantaneous speed deduction processing on the reset distance data and multi-path effect suppression processing on external positioning system original displacement data to obtain theoretical walking speed data and actual displacement data, including:

[0048] Perform instantaneous speed deduction processing on the reset distance data to obtain theoretical walking speed data.

[0049] Perform multi-path effect suppression processing on external positioning system original displacement data to obtain actual displacement data.

[0050] In a second aspect, the application also provides a mine coal mining robot intelligent path planning system, which includes:

[0051] A multi-source data preprocessing module: obtains and preprocesses camera original images, contact sensor geological data, and inertial measurement unit original pose data to obtain three-dimensional vision data, geological feature data, and corrected pose data;

[0052] An environment perception correction and modeling module: obtains historical working face data and equipment pose information, combines three-dimensional vision data, geological feature data, and corrected pose data, and performs environment perception collaborative correction processing to obtain environment model data;

[0053] A walking and sliding monitoring module: obtains motor encoder original walking distance and external positioning system original displacement data, performs walking and sliding state monitoring processing, and obtains dynamic sliding rate;

[0054] The path planning and slip compensation module obtains a preset cutting template, combines historical working face data, including historical working face geological data, corrected attitude data, environment model data and dynamic slip rate, performs path initial planning and slip compensation processing, and obtains optimized path data.

[0055] The safety conflict detection and path adjustment module performs safety conflict detection and path adjustment processing based on the optimized path data, and obtains path verification flag data and adjusted path data.

[0056] The path output and execution control module performs path output and execution control processing based on the path verification flag data and the adjusted path data, and obtains actual path execution effect.

[0057] In a third aspect, the present application provides a mine coal mining robot intelligent path planning device, which comprises a memory and a processor; the memory is used for storing a computer program; the processor is used for executing the computer program to realize the steps of the mine coal mining robot intelligent path planning method of the first aspect.

[0058] In a fourth aspect, the present application provides a storage medium, which stores computer program instructions; when the computer program instructions are read and run by a processor, the steps of the mine coal mining robot intelligent path planning method of the first aspect are executed.

[0059] The beneficial effects of the present application are as follows:

[0060] The present application effectively solves the problems of the prior art, such as the deviation of environment perception caused by the distortion of visual sensors in a high-dust environment, the dynamic slip of the walking mechanism in a loose coal and gangue layer, and the difficulty of existing methods in synchronously compensating for the two to cause the trajectory deviation to exceed the safety range, realizes accurate correction of environment perception and real-time compensation of walking slip, effectively controls the deviation of the actual cutting trajectory and the theoretical path within the safety threshold, and improves the coal mining efficiency and the operation safety.

[0061] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and obtained by the structure indicated in the specification and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0063] Figure 1 A flowchart of the intelligent path planning method of the coal mining robot in the mine of the present application is shown.

[0064] Figure 2 A module diagram of the intelligent path planning system of the coal mining robot in the mine of the present application is shown. DETAILED DESCRIPTION

[0065] In order to solve the problems raised in the background art, the present application adopts multi-sensor data fusion preprocessing, cooperates with the correction of the environment model, monitors the dynamic slip rate, combines the historical data and the preset template to plan the path and compensate for the slip, and outputs the execution after safety detection adjustment, thereby improving the coal mining efficiency and the operation safety.

[0066] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0067] In some embodiments, as shown in Figure 1 The present application provides an intelligent path planning method of a coal mining robot in a mine, comprising:

[0068] S1. Obtain the original image of the camera, the geological data of the contact sensor and the original attitude data of the inertial measurement unit, and perform preprocessing to obtain three-dimensional visual data, geological feature data and corrected attitude data.

[0069] S2. Obtain historical working face data and equipment attitude information, combine the three-dimensional visual data, the geological feature data and the corrected attitude data, and perform environment perception cooperative correction processing to obtain environment model data.

[0070] S3. Obtain the original walking distance of the motor encoder and the original displacement data of the external positioning system, perform walking slip state monitoring processing, and obtain the dynamic slip rate.

[0071] S4. Obtain a preset cutting template, combine historical working face data in the historical working face geological data, corrected attitude data, environment model data and dynamic slip rate, perform path initial planning and slip compensation processing, and obtain optimized path data.

[0072] S5. Based on the optimized path data, safety conflict detection and path adjustment processing is performed to obtain path verification flag data and adjusted path data.

[0073] S6. Based on the path verification flag data and the adjusted path data, path output and execution control processing is performed to obtain the actual path execution effect.

[0074] In some embodiments, the camera raw image in S1 represents the visual information of the coal mining face directly collected by the mine explosion-proof camera, which includes the spatial distribution information of the coal wall, hydraulic support and conveyor.

[0075] The contact sensor geological data represents the mechanical signals collected by the piezoelectric sensor array installed on the cutting pick of the coal mining machine, which includes three-axis force values (Fx, Fy, Fz) and vibration frequency spectrum characteristics, for example, the peak force of the Z-axis can reach 25kN±3kN when cutting sandstone layer, and maintain 8-12kN when cutting coal seam.

[0076] The inertial measurement unit raw attitude data refers to the motion parameters directly output by the MEMS inertial sensor carried by the coal mining machine body, including three-axis angular velocity ( ) and three-axis acceleration ( ).

[0077] Preprocessing is performed to obtain three-dimensional visual data, geological feature data and corrected attitude data, including:

[0078] S11. The camera raw image is subjected to coal dust adhesion interference removal processing to obtain a preliminary clean image.

[0079] The original image can be processed by the dark channel prior model to remove the fog-like blur and obtain a preliminary clean image.

[0080] S12. The preliminary clean image is subjected to edge sharpening enhancement processing to obtain enhanced visual data.

[0081] For the low-illumination environment in the mine, Laplacian pyramid fusion enhancement is used to obtain enhanced visual data, and the enhanced visual data is specifically a floating-point image matrix.

[0082] S13. The enhanced visual data is subjected to stereo depth reconstruction processing to obtain three-dimensional visual data.

[0083] Depth calculation can be realized based on the binocular parallax principle to obtain three-dimensional visual data, and the three-dimensional visual data is specifically a point cloud data set.

[0084] S14. Perform stratum feature extraction processing on the contact sensor geological data and zero-point drift correction processing on the original attitude data of the inertial measurement unit to obtain geological feature data and corrected attitude data.

[0085] Perform time-domain peak extraction and frequency energy analysis on the contact sensor geological data to obtain geological feature data.

[0086] For example, detect the cutting resistance peak value within a 50-millisecond window, mark the hard rock layer when the peak value exceeds the hard rock determination threshold, and calculate the energy proportion of the 200-800 Hz frequency band, mark the sandstone layer when the proportion exceeds the sandstone determination threshold; wherein the hard rock determination threshold and the sandstone determination threshold can be calibrated according to sample destruction tests.

[0087] In some embodiments, the historical working face data in S2 represents a set of geological features and path execution records of completed working faces stored in the coal mine mining database, including coal seam thickness distribution, lithology zoning map, cutting path trajectory, and equipment operating parameters.

[0088] The device attitude information represents the real-time spatial attitude parameters of the shearer body, including pitch angle, roll angle, and yaw angle.

[0089] Perform environment perception collaborative correction processing to obtain environment model data, including:

[0090] S21. Perform three-dimensional point cloud reconstruction processing on the three-dimensional vision data and geological feature data to obtain environment point cloud data.

[0091] The geological feature point cloud and the vision point cloud can be aligned through the ICP registration algorithm, and the environment point cloud data can be generated in combination with the RGB-D information.

[0092] S22. Perform coal-rock interface pattern recognition processing based on the environment point cloud data and the historical working face data to obtain interface recognition data.

[0093] The ResNet-18 network can be trained according to the historical data to recognize the coal-rock boundary pattern, and the environment point cloud data and the historical working face data can be used as input. The network extracts point cloud features through transfer learning and outputs interface recognition data. The coal-rock boundary threshold can be determined by statistical analysis of the average gray gradient of samples in the historical working face data.

[0094] S23. Perform pose collaborative registration processing on the corrected attitude data and the interface recognition data to obtain compensated perception data.

[0095] Extended Kalman filter correction can be used to perform pose collaborative registration on the attitude data and the interface recognition data to obtain compensated perception data, i.e., a spatially aligned semantic environment model.

[0096] S24. Perform dynamic datum plane construction processing based on the compensated perception data and the device pose information to obtain environment model data.

[0097] The dynamic datum plane can be constructed by least square plane fitting to obtain the environment model data.

[0098] In some embodiments, the motor encoder raw walking distance in S3 represents pulse count data directly output by a built-in encoder of a walking motor of the coal mining machine.

[0099] The external positioning system raw displacement data represents absolute position data of the coal mining machine directly output by a mine positioning system.

[0100] Perform walking and slipping state monitoring processing to obtain a dynamic slipping rate, including:

[0101] S31. Perform cumulative error resetting processing on the motor encoder raw walking distance to obtain reset distance data.

[0102] The cumulative error resetting processing is periodically calibrated by the absolute position feedback of the external positioning system to obtain the reset distance data.

[0103] S32. Perform instantaneous speed deduction processing on the reset distance data and multi-path effect suppression processing on the external positioning system raw displacement data to obtain theoretical walking speed data and actual displacement data, including:

[0104] S321. Perform instantaneous speed deduction processing on the reset distance data to obtain the theoretical walking speed data.

[0105] The instantaneous speed deduction processing specifically includes: performing differential calculation on the reset distance data to obtain an instantaneous speed, and outputting the theoretical walking speed data as a time-stamped instantaneous speed sequence.

[0106] S322. Perform multi-path effect suppression processing on the external positioning system raw displacement data to obtain the actual displacement data.

[0107] The multi-path effect suppression processing specifically includes: using Kalman filtering to eliminate roadway multi-path interference on the external positioning system raw displacement data to obtain the actual displacement data.

[0108] S33. Perform slipping rate dynamic estimation processing on the theoretical walking speed data and the actual displacement data to obtain a dynamic slipping rate.

[0109] The slipping rate is calculated by a slipping rate formula: ; wherein, represents the slipping rate, is obtained by differentiating the actual displacement data, represents the theoretical walking speed data.

[0110] In some embodiments, the preset cutting template in S4 represents a set of reference cutting paths generated based on the coal seam hosting parameters, containing path point coordinates (X, Y, Z) and a vector sequence of cutting drum height adjustment instructions.

[0111] For example, when the thickness of a certain coal seam is 2.1 m ± 0.2 m and the inclination is 15°, the template defines the drum center trajectory at a distance of 1.0 m from the roof and a cutting depth of 0.8 m.

[0112] The historical working face geological data represents the structured geological attribute features of the mined working face extracted from the historical working face data, including lithology distribution, structural features, hydrological parameters, etc.

[0113] The initial path planning and slip compensation processing are performed to obtain optimized path data, including:

[0114] S41. Based on the environmental model data and the preset cutting template, initial path generation processing is performed to obtain theoretical cutting path data.

[0115] The reference path points of the preset cutting template are aligned with the spatial coordinates of the environmental model data through a cubic spline interpolation algorithm to generate smooth and continuous theoretical cutting path data.

[0116] For example, when the inclination of a certain coal seam is 18°, the preset template path points (X, Y, Z) are interpolated to generate continuous trajectory lines with a spacing of 0.1 meters.

[0117] S42. Based on the interface recognition data and the historical working face geological data, working face geological feature partitioning processing is performed to obtain geological partition map data.

[0118] The coal-rock boundary line of the interface recognition data is used as a threshold to divide the geological units, and the attribute labeling of the geological units can refer to the lithology parameters (such as sandstone hardness 120 MPa) of similar geological conditions in the historical working face geological data to obtain a grid-based geological attribute map with lithology encoding, marked as geological partition map data.

[0119] For example, a certain working face is divided into 2 geological zones:

[0120] Zone 1: coal seam proportion > 85%, cutting parameters [rotation speed 80 rpm, feed 1.2 m / s].

[0121] Zone 2: sandstone interlayer, cutting parameters [rotation speed 60 rpm, feed 0.8 m / s].

[0122] S43. Based on the dynamic slip rate and the geological partition map data, slip influence domain segmentation processing is performed to obtain regionalized slip parameter data.

[0123] The slip rate value in the dynamic slip rate data is mapped to the grid cell of the geological zoning map data according to its spatial coordinates (X, Y), and a mapping relationship between the slip rate and the geological attribute is established.

[0124] If the slip rate is greater than or equal to 0.2, it represents a high influence domain; if the slip rate is less than or equal to 0.1, it represents a low influence threshold; and the rest represents a medium influence threshold.

[0125] The slip compensation coefficient is regenerated as follows:

[0126] 1. Calculate the base coefficient: wherein, , represents the regression coefficient calibrated through the mine site test.

[0127] 2. If the geology is soft rock, the slip compensation coefficient is corrected as follows: ; if the geology is hard rock, the slip compensation coefficient is corrected as follows: .

[0128] The generated regionalized slip parameter data includes coordinates, geological codes, slip rates , influence domain registration, and slip compensation coefficients .

[0129] S44. Trend prediction processing is performed on the regionalized slip parameter data to obtain slip prediction data.

[0130] The regionalized slip parameter data can be input into an ARIMA time series model, and after model processing, output slip prediction data is obtained, which is the slip rate change curve in the future time window (such as 5 seconds).

[0131] For example, the current slip rate is 0.18, and it is predicted to rise to 0.22 after 2 seconds.

[0132] S45. Path point collaborative compensation processing is performed on the theoretical cutting path data and the slip prediction data to obtain compensated path data.

[0133] The coordinates (X, Y) of the theoretical cutting path data are mapped to the grid coordinate system, and the path point normal offset is calculated as follows: ; wherein v represents the theoretical walking speed, and represents the prediction time window.

[0134] According to the drum axial pitch angle, the normal direction of the cutting trajectory is determined, and the coordinate points of the theoretical cutting path data are translated along the normal direction by a distance to obtain the compensated path data, i.e., the new trajectory after slip correction.

[0135] S46. Perform dynamic pose constraint optimization processing on the compensated path data and the corrected pose data to obtain optimized path data, comprising:

[0136] S461. Perform dynamic pose constraint modeling to generate constraint rule data.

[0137] Construct a curvature safety constraint through the real-time pitch angle θ and roll angle γ of the corrected pose data: ; wherein, represents the maximum allowed curvature, represents the length of the coal mining robot, represents the maximum pitch angle.

[0138] When , it is attenuated to 0.85 times, and when the roll angle , it is attenuated to 0.9 times. Constraint rule data contains a structured parameter set of dynamic threshold and attenuation rules.

[0139] S462. Perform iterative optimization calculation on the compensated path data based on the constraint rule data to obtain optimized path data.

[0140] Detect the curvature of the compensated path data,

[0141] compress the coordinates along the curvature radius at the points of ; obtain optimized path data; wherein, represents the optimized curvature radius. In some embodiments, the safety conflict detection and path adjustment processing based on the optimized path data in S5 obtains path verification flag data and adjusted path data, comprising:

[0142] S51. Perform safety distance conflict scanning processing on the optimized path data to obtain conflict marker data.

[0143] Detect the minimum distance between the optimized path data and static obstacles such as roadway boundaries and hydraulic supports through a distance field algorithm.

[0144] When the distance is less than a preset safety distance threshold (such as 0.5 meters), it is marked as a conflict, and conflict marker data containing structured data of conflict position, type and level is output; wherein, the safety distance threshold can be determined through collision accident data analysis.

[0145]

[0146] ​​For example, the minimum distance between the theoretical path point (35.6, 2.42) of the coal mining machine and the hydraulic support is 0.3 meters, which is less than 0.5 meters. Therefore, the conflict marking data record is "coordinates (35.6, 2.42), conflict type: equipment interference, level: high risk".

[0147] S52. Local path re-planning processing is performed on the conflict marking data to obtain adjusted path data.

[0148] Specifically, the local path correction can be performed by using a rapid exploration random tree algorithm: a re-planning area with a specified range (such as 1.5*1.5 meters) is established around the conflict point, an obstacle avoidance path is generated in combination with the roadway constraint conditions of the environment model data, the new path needs to ensure that the distance to the obstacle is greater than the safety distance threshold, and the generated adjusted path data is the continuous trajectory after obstacle avoidance optimization.

[0149] For example, for the conflict point (35.6, 2.42), the detour path point sequence (35.6, 2.62), (35.7, 2.58), and (35.7, 2.45) is generated.

[0150] S53. Efficiency evaluation processing is performed on the adjusted path data to obtain path score data.

[0151] The adjusted path data is subjected to path length increment , curvature change rate , and cutting time estimation quantitative analysis, the path length increment is: , the curvature change rate is: , and the cutting time estimation is: ; wherein represents the total length of the adjusted path, represents the total length of the original path, represents the curvature of the i-th path point, represents the average curvature of the path, and n represents the total number of path points, represents the preset average cutting speed.

[0152] S54. Executability threshold checking processing is performed on the path score data to obtain path verification flag data.

[0153] The itemized threshold checking is performed to determine the path executability: the checking threshold of the path length increment ΔL can be set according to the minimum cutting efficiency requirement of the mine, and when the path length increment ΔL is greater than the checking threshold, it is determined that the mining efficiency is affected; the checking threshold of the curvature change rate can be determined according to the statistical results of the equipment stability test, and when the curvature change rate is greater than the checking threshold, it is determined that the equipment stability is affected.If the value is less than the check threshold, it indicates that the equipment is in good stability; the check threshold of the cutting time T is determined by the time consumption distribution statistics of the cutting cycle in the historical working face data, and when the cutting time T is greater than the check threshold, the production rhythm will be damaged.

[0154] The final generated path verification flag data is specifically: when the path length increment ΔL, the curvature change rate , and the cutting time T all meet the threshold, output a state code representing executable; if any index exceeds the limit, output a state code representing the need for re-adjustment.

[0155] In some embodiments, based on the path verification flag data and the adjusted path data in S6, path output and execution control processing is performed to obtain the actual path execution effect, including:

[0156] S61. Based on the path verification flag data and the adjusted path data, perform execution instruction conversion processing to obtain control instruction data.

[0157] When the state code represents executable, the coordinate points of the adjusted path data are converted into robot driving instructions, specifically including coordinate instructions and motion parameters, and the motion parameters specifically include speed and acceleration.

[0158] S62. Perform timing scheduling allocation processing on the control instruction data to obtain timing instruction sequence data.

[0159] According to the path point spacing and the preset walking speed, the time interval of each instruction is calculated, and the absolute timestamp is allocated according to the specified time (such as 0.1 seconds) as the reference time slice, and finally the timing instruction sequence data is output.

[0160] S63. Perform execution mechanism compatibility adaptation processing on the timing instruction sequence data to obtain adapted control data.

[0161] Convert the timing instruction sequence data into an electrical protocol supported by the robot driving unit, and simultaneously perform electrical property adaptation, output the adapted control data, which represents an electrical signal sequence that can directly drive the execution mechanism.

[0162] S64. By sending the adapted control data to the robot driving unit processing, the actual path execution effect is obtained.

[0163] Send the adapted control data to the robot driving unit, and monitor the execution effect in real time: sample the actual position (X', Y', Z') fed back by the motor encoder at a specified frequency (such as 1 kHz frequency), calculate the error value with the theoretical coordinates; when the error value is greater than the preset error threshold, automatically trigger real-time calibration, output the actual path execution effect, which contains the structure of the theoretical coordinates, actual coordinates, error value and timestamp.

[0164] In some embodiments, as shown in FIG. 1, Figure 2 The present application also provides a mine coal mining robot intelligent path planning system, which comprises:

[0165] A multi-source data preprocessing module: obtains and pre-processes original camera images, contact sensor geological data and original attitude data of an inertial measurement unit to obtain three-dimensional vision data, geological feature data and corrected attitude data;

[0166] An environment perception correction and modeling module: obtains historical working face data and equipment attitude information, and performs environment perception collaborative correction processing on the three-dimensional vision data, the geological feature data and the corrected attitude data to obtain environment model data;

[0167] A walking and sliding monitoring module: obtains original walking distance of a motor encoder and original displacement data of an external positioning system, and performs walking and sliding state monitoring processing to obtain a dynamic sliding rate;

[0168] A path planning and sliding compensation module: obtains a preset cutting template, and performs path initial planning and sliding compensation processing on the historical working face geological data in the historical working face data, the corrected attitude data, the environment model data and the dynamic sliding rate to obtain optimized path data;

[0169] A safety conflict detection and path adjustment module: performs safety conflict detection and path adjustment processing based on the optimized path data to obtain path verification flag data and adjusted path data;

[0170] A path output and execution control module: performs path output and execution control processing based on the path verification flag data and the adjusted path data to obtain actual path execution effect.

[0171] In some embodiments, the present application provides a mine coal mining robot intelligent path planning device, which comprises a memory and a processor; the memory is used to store a computer program; and the processor is used to execute the computer program to realize the steps of the mine coal mining robot intelligent path planning method.

[0172] In some embodiments, the present application provides a storage medium, which stores computer program instructions; when the computer program instructions are read and run by a processor, the steps of the mine coal mining robot intelligent path planning method are executed.

[0173] Any reference to storage, memory, database or other medium herein includes non-volatile and / or volatile storage. Non-volatile storage can include read-only memory (ROM), programmable ROM (PROM), electronically programmable ROM (EPROM), or electrically erasable programmable ROM (EEPROM). Volatile storage can include random-access memory (RAM), or external cache memory.

[0174] It should be noted that, in the present document, relational terms such as "first" and "second", and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. In other words, any process, method, article, or apparatus that comprises a list of elements is deemed to include not only those elements but also other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0175] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that the technical solutions recorded in the foregoing embodiments can be modified or some of the technical features can be replaced equivalently, and such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for intelligent path planning of a coal mining robot in a mine, characterized in that, include: The system acquires raw images from the camera, geological data from the contact sensor, and raw attitude data from the inertial measurement unit, and performs preprocessing to obtain 3D visual data, geological feature data, and corrected attitude data. Historical working face data and equipment posture information are acquired, and combined with the three-dimensional visual data, the geological feature data and the corrected posture data, environmental perception collaborative correction processing is performed to obtain environmental model data. The original walking distance of the motor encoder and the original displacement data of the external positioning system are acquired, and the walking slip state monitoring and processing are performed to obtain the dynamic slip rate. A preset cutting template is obtained, and combined with historical working face geological data, the corrected attitude data, the environmental model data and the dynamic slip rate in the historical working face data, the path initial planning and slip compensation processing are performed to obtain optimized path data; Based on the optimized path data, security conflict detection and path adjustment processing are performed to obtain path verification flag data and adjusted path data. Based on the path verification flag data and the adjusted path data, path output and execution control processing are performed to obtain the actual path execution effect; The process involves initial path planning and slip compensation to obtain optimized path data, including: Based on the environmental model data and the preset cutting template, an initial path generation process is performed to obtain theoretical cutting path data; Geological feature zoning of working faces is performed based on interface recognition data and historical working face geological data to obtain geological zoning map data. Based on the dynamic slip rate and the geological zoning map data, the slip influence domain is segmented to obtain regionalized slip parameter data. The regionalized slip parameter data is subjected to trend prediction processing to obtain slip prediction data; The theoretical truncation path data and the slip prediction data are subjected to path point collaborative compensation processing to obtain compensated path data; The compensated path data and the corrected attitude data are subjected to dynamic attitude constraint optimization processing to obtain optimized path data.

2. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Preprocessing yields 3D visual data, geological feature data, and corrected pose data, including: The original camera image is processed to remove coal dust adhesion interference, resulting in a preliminary clean image; The initial cleaned image is subjected to edge sharpening enhancement processing to obtain enhanced visual data; The enhanced visual data is subjected to stereo depth reconstruction processing to obtain three-dimensional visual data; The geological data from the contact sensor is processed to extract rock strata features, and the original attitude data of the inertial measurement unit is processed to correct zero-point drift, thus obtaining geological feature data and corrected attitude data.

3. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Environmental perception collaborative correction processing is performed to obtain environmental model data, including: The three-dimensional visual data and the geological feature data are subjected to three-dimensional point cloud reconstruction processing to obtain environmental point cloud data; Based on the environmental point cloud data and historical working face data, coal-rock interface pattern recognition processing is performed to obtain interface recognition data. The corrected posture data and the interface recognition data are subjected to pose co-registration processing to obtain compensated perception data; Based on the compensated sensing data and device attitude information, dynamic reference surface construction is performed to obtain environmental model data.

4. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Walking slip state monitoring and processing are performed to obtain the dynamic slip rate, including: The original travel distance of the motor encoder is processed by cumulative error reset to obtain the reset distance data; The reset distance data is processed by instantaneous velocity extrapolation, and the original displacement data of the external positioning system is processed by multipath effect suppression to obtain theoretical walking speed data and actual displacement data. The theoretical walking speed data and the actual displacement data are subjected to dynamic slip ratio estimation processing to obtain the dynamic slip ratio.

5. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Based on the optimized path data, security conflict detection and path adjustment processing are performed to obtain path verification flag data and adjusted path data, including: The optimized path data is subjected to a safe distance conflict scan to obtain conflict marker data; The conflict marker data is subjected to local path replanning to obtain the adjusted path data; The adjusted path data is subjected to efficiency evaluation processing to obtain path score data; The path scoring data is subjected to an executability threshold verification process to obtain path verification flag data.

6. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Based on the path verification flag data and the adjusted path data, path output and execution control processing are performed to obtain the actual path execution effect, including: Based on the path verification flag data and the adjusted path data, execution instruction conversion processing is performed to obtain control instruction data; The control command data is processed by timing scheduling and allocation to obtain timing command sequence data; The timing instruction sequence data is subjected to actuator compatibility adaptation processing to obtain adapted control data; The actual path execution effect is obtained by sending the adapted control data to the robot drive unit for processing.

7. The intelligent path planning method for coal mining robots in mines according to claim 1, characterized in that, Dynamic attitude constraint optimization is performed on the compensated path data and the corrected attitude data to obtain optimized path data, including: Perform dynamic attitude constraint modeling to generate constraint rule data; Based on the constraint rule data, iterative optimization calculations are performed on the compensated path data to obtain optimized path data.

8. The intelligent path planning method for coal mining robots in mines according to claim 4, characterized in that, The reset distance data is processed to perform instantaneous velocity extrapolation, and the original displacement data from the external positioning system is processed to suppress multipath effects, resulting in theoretical walking speed data and actual displacement data, including: The instantaneous velocity extrapolation process is performed on the reset distance data to obtain theoretical walking speed data; Multipath effect suppression processing is performed on the original displacement data of the external positioning system to obtain the actual displacement data.

9. An intelligent path planning system for a coal mining robot in a mine, characterized in that, It includes: Multi-source data preprocessing module: Acquires raw images from cameras, geological data from contact sensors, and raw attitude data from inertial measurement units, and performs preprocessing to obtain 3D visual data, geological feature data, and corrected attitude data; Environmental perception correction and modeling module: acquires historical working face data and equipment posture information, and combines the three-dimensional visual data, the geological feature data and the correction posture data to perform environmental perception collaborative correction processing to obtain environmental model data; Walking slip monitoring module: acquires the original walking distance of the motor encoder and the original displacement data of the external positioning system, performs walking slip status monitoring and processing, and obtains the dynamic slip rate; Path planning and slip compensation module: Obtain a preset cutting template, and combine it with historical working face geological data, the corrected attitude data, the environmental model data and the dynamic slip rate in the historical working face data to perform initial path planning and slip compensation processing to obtain optimized path data; Security conflict detection and path adjustment module: Based on the optimized path data, it performs security conflict detection and path adjustment processing to obtain path verification flag data and adjusted path data; Path output and execution control module: Based on the path verification flag data and the adjusted path data, it performs path output and execution control processing to obtain the actual path execution effect; The process involves initial path planning and slip compensation to obtain optimized path data, including: Based on the environmental model data and the preset cutting template, an initial path generation process is performed to obtain theoretical cutting path data; Geological feature zoning of working faces is performed based on interface recognition data and historical working face geological data to obtain geological zoning map data. Based on the dynamic slip rate and the geological zoning map data, the slip influence domain is segmented to obtain regionalized slip parameter data. The regionalized slip parameter data is subjected to trend prediction processing to obtain slip prediction data; The theoretical truncation path data and the slip prediction data are subjected to path point collaborative compensation processing to obtain compensated path data; The compensated path data and the corrected attitude data are subjected to dynamic attitude constraint optimization processing to obtain optimized path data.

Images