A robotic arm control method and system for peeling adhesive from hydraulic hoses

Abstract

This invention relates to the field of robotic arm control technology, specifically to a robotic arm control method and system for peeling adhesive from hydraulic hoses. The method includes: acquiring an image sequence and a tactile contact signal sequence of the hydraulic hose as the robotic arm contacts it along a preset peeling path; positioning block groups are provided at both ends of the hydraulic hose; the tactile contact signal sequence includes contact feedback signals between the robotic arm and the positioning block groups; extracting edge features of the adhesive layer based on the hose image sequence; and determining a sequence of target positioning points corresponding to the positioning block groups based on the edge features of the adhesive layer and the tactile contact signal sequence. This invention, through comprehensive analysis of the hose image sequence and the tactile contact signal sequence, can accurately extract the edge features of the adhesive layer and, based on quasi-de Bruin sequences, accurately determine the contact position between the robotic arm and the hydraulic hose. This precise positioning capability ensures that the robotic arm can accurately peel off the adhesive layer according to actual needs, avoiding misoperation or damage to the hose.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manipulator control, in particular to a manipulator control method and system for stripping hydraulic oil pipe. BACKGROUND

[0002] In the past few decades, manufacturing industry has gradually transformed from manual operation to automation and intelligence. Traditional hydraulic oil pipe stripping operation mainly relies on manual operation or semi-automatic machinery, which has problems of low efficiency, insufficient precision and high risk of injury. With the development of industrial and intelligent manufacturing concepts, automated production lines require efficient, accurate and repeatable operation links, especially for links with high precision and safety requirements such as hydraulic oil pipes. Manipulator control method can replace manual operation to achieve precise, stable and controllable stripping process, which meets the general trend of industrial upgrading.

[0003] Currently, traditional stripping methods usually rely on manual operation or simple mechanical control systems, which lack precise positioning and real-time feedback. This can lead to inaccurate contact position between the manipulator and the oil pipe, increasing the risk of misoperation, which may damage the oil pipe or fail to accurately strip the adhesive layer as required. Moreover, there is usually a lack of real-time adjustment mechanism based on the thickness of the adhesive layer and the characteristics of the oil pipe. The stripping path and force often cannot adapt to the changes of different adhesive layers, resulting in uneven stripping process, incomplete adhesive layer stripping or damage to the oil pipe surface.

[0004] In addition, in the traditional method, once an error occurs, manual adjustment or shutdown correction is usually required, which not only wastes time but also affects efficiency. The lack of automatic adjustment capability based on comprehensive errors makes the stripping operation often unable to be optimized to the best state. Moreover, it often only relies on one aspect of visual or tactile signals for operation. Such a single perception mode makes the operation system lack intelligent and flexible response capability. SUMMARY

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solution: a manipulator control method for stripping hydraulic oil pipe, comprising:

[0006] obtaining an oil pipe image sequence and a tactile contact signal sequence when the manipulator contacts the hydraulic oil pipe along a preset stripping path; the hydraulic oil pipe is provided with a positioning block group at both ends, and the tactile contact signal sequence includes contact feedback signals between the manipulator and the positioning block group;

[0007] extracting adhesive layer edge features based on the oil pipe image sequence, determining a target positioning point sequence corresponding to the positioning block group according to the adhesive layer edge features and the tactile contact signal sequence, generating an adhesive layer positioning sequence and a contact pressure deviation sequence;

[0008] According to the quasi-de Bruijn sequence coding rule of the positioning block group, the target positioning point sequence is preliminarily decoded to determine the spatial distribution characteristics corresponding to the target positioning point sequence;

[0009] According to the spatial distribution characteristics, a target grid is established to obtain a reference point sequence in the target grid; based on the offset information of the target positioning point sequence and the reference point sequence, the contact position deviation sequence of the manipulator and the hydraulic oil pipe is determined through complete decoding of the quasi-de Bruijn sequence;

[0010] According to the contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness data in the glue layer positioning sequence, the glue stripping control comprehensive error is calculated;

[0011] Based on the glue stripping control comprehensive error, the glue stripping path, the glue stripping force, and the glue stripping tool speed of the manipulator are cooperatively adjusted to control the manipulator to complete the glue layer stripping operation of the hydraulic oil pipe.

[0012] Preferably, the glue layer edge features are extracted based on the oil pipe image sequence, the target positioning point sequence corresponding to the positioning block group is determined according to the glue layer edge features and the tactile contact signal sequence, and the glue layer positioning sequence and the contact pressure deviation sequence are generated, including:

[0013] Each frame of the oil pipe image sequence is subjected to denoising processing and gray scale enhancement processing;

[0014] The glue layer edge in the processed oil pipe image is extracted by using an edge contour extraction algorithm, the coverage area of the glue layer on the oil pipe surface is determined, and the glue layer positioning sequence is formed in the stripping order;

[0015] The actual contact pressure value at each contact time is extracted from the tactile contact signal sequence, the difference between the actual contact pressure value and the preset standard contact pressure value is calculated, and the contact pressure deviation sequence is arranged in the contact time order;

[0016] The positioning block corresponding to the tactile contact signal in the oil pipe image is identified, the center point coordinates of the positioning block are extracted, and the target positioning point sequence is composed in the contact order.

[0017] Preferably, according to the quasi-de Bruijn sequence coding rule of the positioning block group, the target positioning point sequence is preliminarily decoded to determine the spatial distribution characteristics corresponding to the target positioning point sequence, including:

[0018] The position of each positioning block in the positioning block group is uniquely coded by the quasi-de Bruijn sequence, and the coding information is pre-stored in the manipulator control system;

[0019] The coordinate parameters of each positioning point in the target positioning point sequence are obtained, the pre-stored quasi-de Bruijn coding information is matched, and the positioning block number corresponding to each positioning point is determined;

[0020] According to the arrangement order of the positioning block numbers, distribution rules of the positioning points in the axial direction and the radial direction of the oil pipe are analyzed to form a spatial distribution feature, wherein the spatial distribution feature includes an axial interval distance of the positioning points and a radial deviation trend.

[0021] Preferably, a target grid is established according to the spatial distribution feature, and a reference point sequence in the target grid is obtained, including:

[0022] A target direction of the target grid is determined based on the axial interval distance in the spatial distribution feature, and the target direction is consistent with the axial direction of the hydraulic oil pipe.

[0023] A reference point in the target positioning point sequence located near the center of the oil pipe cross section is selected as a reference point, a distance average of all positioning points to the reference point is calculated, and a target orthogonal point of the target grid is determined.

[0024] According to the radial deviation trend in the spatial distribution feature, in combination with a preset interval of the positioning block group, a grid width of the target grid is determined.

[0025] A two-dimensional target grid is established with the target orthogonal point as the origin, the target direction as the X axis and the radial direction as the Y axis, the grid points of the target grid are taken as reference points, and a reference point sequence is formed in the coordinate order.

[0026] Preferably, according to the radial deviation trend in the spatial distribution feature, in combination with the preset interval of the positioning block group, the grid width of the target grid is determined, including:

[0027] The center interval of adjacent positioning blocks in the positioning block group is obtained, the average of all adjacent intervals is calculated, and a basic interval value is obtained.

[0028] The radial deviation trend in the spatial distribution feature is analyzed, and the maximum deviation range of the target positioning point in the radial direction is counted.

[0029] The ratio of the basic interval value to the maximum radial deviation range is rounded to obtain the grid width of the target grid; the numerical value of the grid width is positively correlated with the pipe diameter specification of the hydraulic oil pipe.

[0030] Preferably, based on the deviation information of the target positioning point sequence and the reference point sequence, the contact position deviation sequence of the manipulator and the hydraulic oil pipe is determined through quasi-De Bruijn sequence complete decoding, including:

[0031] The X-axis deviation and the Y-axis deviation between each target positioning point and the nearest reference point are calculated to form a deviation information set.

[0032] The deviation information set is input into a quasi-De Bruijn decoding model, and the model reversely deduces the actual spatial position of the positioning block based on a pre-stored encoding rule.

[0033] The axial deviation value and the radial deviation value of each contact position are calculated by comparing the actual spatial position derived with the theoretical position of the positioning block, and a contact position deviation sequence is formed in the order of contact.

[0034] The stripping control comprehensive error is calculated based on the contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness data in the glue layer positioning sequence, including:

[0035] The average thickness value of each glue layer region is extracted from the glue layer positioning sequence to form a glue layer thickness sequence;

[0036] The contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness sequence are respectively assigned a weight coefficient, and the stripping control comprehensive error is obtained by weighted summation.

[0037] The stripping path, the stripping force, and the stripping tool speed of the manipulator are adjusted based on the stripping control comprehensive error, including:

[0038] The X-axis compensation amount and the Y-axis compensation amount of the stripping path are calculated based on the axial deviation value and the radial deviation value in the contact position deviation sequence, and the motion trajectory of the manipulator is adjusted;

[0039] The stripping force adjustment amount is calculated based on the contact pressure deviation sequence, and the force adjustment is realized by changing the oil supply pressure of the manipulator drive cylinder;

[0040] The speed adjustment amount of the stripping tool is calculated based on the glue layer thickness sequence and the stripping control comprehensive error, and the speed adjustment is realized by changing the pulse frequency of the tool drive motor;

[0041] The path compensation amount, the force adjustment amount, and the speed adjustment amount are input into the manipulator control module simultaneously, the stripping parameters are updated, and the stripping operation is performed.

[0042] The edge profile extraction algorithm is used to extract the glue layer edge in the processed oil pipe image, and the coverage area of the glue layer on the oil pipe surface is determined, including:

[0043] The Gaussian filter algorithm is used to denoise the oil pipe image, and the salt and pepper noise in the image is eliminated;

[0044] The histogram equalization algorithm is used to enhance the gray scale of the denoised image, and the gray scale contrast between the glue layer and the oil pipe body is improved;

[0045] The Canny edge detection algorithm is used to extract the edge pixel points in the enhanced image, and the isolated edge points are removed by connected component analysis;

[0046] The polynomial fitting algorithm is used to curve fit the effective edge pixel points to form a continuous glue layer edge contour line;

[0047] According to the closed area of the edge contour line, the coverage area of the adhesive layer on the surface of the oil pipe is determined, and the coordinate range of the coverage area is recorded.

[0048] A mechanical hand control system for hydraulic oil pipe stripping is suitable for the mechanical hand control method for hydraulic oil pipe stripping described above, comprising:

[0049] A data acquisition unit is configured to acquire an oil pipe image sequence and a tactile contact signal sequence when the mechanical hand contacts the hydraulic oil pipe along the preset stripping path; the hydraulic oil pipe is provided with a positioning block group at both ends, and the tactile contact signal sequence includes contact feedback signals between the mechanical hand and the positioning block group;

[0050] A feature extraction unit is configured to extract edge features of the adhesive layer based on the oil pipe image sequence, determine a target positioning point sequence corresponding to the positioning block group based on the edge features of the adhesive layer and the tactile contact signal sequence, and generate an adhesive layer positioning sequence and a contact pressure deviation sequence;

[0051] A distribution coding unit is configured to preliminarily decode the target positioning point sequence based on the quasi-De Bruijn sequence coding rule of the positioning block group, and determine the spatial distribution features corresponding to the target positioning point sequence;

[0052] A deviation calculation unit is configured to establish a target grid based on the spatial distribution features, acquire a reference point sequence in the target grid, and determine a contact position deviation sequence between the mechanical hand and the hydraulic oil pipe through complete decoding of the quasi-De Bruijn sequence based on the offset information between the target positioning point sequence and the reference point sequence;

[0053] An error calculation unit is configured to calculate a stripping control comprehensive error based on the contact position deviation sequence, the contact pressure deviation sequence, and adhesive layer thickness data in the adhesive layer positioning sequence;

[0054] A stripping control unit is configured to cooperatively adjust the stripping path, stripping force, and stripping tool speed of the mechanical hand based on the stripping control comprehensive error, and control the mechanical hand to complete the adhesive layer stripping operation of the hydraulic oil pipe.

[0055] Compared with the prior art, the present application has the following advantages:

[0056] (1) The present application can accurately extract the edge features of the adhesive layer through comprehensive analysis of the oil pipe image sequence and the tactile contact signal sequence, and accurately determine the contact position between the mechanical hand and the hydraulic oil pipe based on the decoding of the quasi-De Bruijn sequence. This precise positioning capability ensures that the mechanical hand can accurately strip the adhesive layer according to actual needs, avoiding misoperation or damage to the oil pipe. By calculating the stripping control comprehensive error and cooperatively adjusting the stripping path, stripping force, and stripping tool speed based on the error, this dynamic adjustment enables the stripping operation to be optimized according to actual conditions, improving the operation accuracy and efficiency.

[0057] (2) The present application combines visual information and tactile signals to ensure that the robot has higher response capability in the process of stripping the adhesive layer, the visual information provides the edge features of the adhesive layer, and the tactile signal provides real-time contact feedback, this double perception makes the operation more intelligent and flexible, through accurate stripping path and force control, the uneven stripping of the adhesive layer or the damage of the oil pipe caused by excessive stripping force or path deviation can be effectively avoided, in addition, through the analysis of the thickness sequence of the adhesive layer, the reasonable matching of the force and the thickness in the stripping process can be ensured, and the quality of the stripping operation is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 It is a step flow diagram of the overall method in an embodiment of the present application.

[0059] Figure 2 It is a system architecture diagram of the overall system in an embodiment of the present application.

[0060] In the figure: 1, data acquisition unit; 2, feature extraction unit; 3, distribution coding unit; 4, deviation calculation unit; 5, error calculation unit; 6, stripping control unit. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part 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 fall within the scope of protection of the present application.

[0062] Embodiment one, please refer to Figure 1 The present application provides a technical solution: a robot control method for stripping hydraulic oil pipe, comprising:

[0063] S1, obtaining the oil pipe image sequence and the tactile contact signal sequence when the robot contacts the hydraulic oil pipe according to the preset stripping path; the two ends of the hydraulic oil pipe are provided with a positioning block group, and the tactile contact signal sequence includes the contact feedback signal of the robot and the positioning block group;

[0064] S2, extracting the adhesive layer edge features based on the oil pipe image sequence, determining the target positioning point sequence corresponding to the positioning block group according to the adhesive layer edge features and the tactile contact signal sequence, generating the adhesive layer positioning sequence and the contact pressure deviation sequence;

[0065] S3, based on the quasi-De Bruijn sequence coding rule of the positioning block group, preliminarily decoding the target positioning point sequence to determine the spatial distribution features corresponding to the target positioning point sequence;

[0066] S4, establishing a target grid according to the spatial distribution characteristics, obtaining a reference point sequence in the target grid; based on the offset information of the target positioning point sequence and the reference point sequence, determining a contact position deviation sequence of the manipulator and the hydraulic oil pipe through quasi-de Bruijn sequence complete decoding;

[0067] S5, calculating a stripping control comprehensive error according to the contact position deviation sequence, the contact pressure deviation sequence, and the adhesive layer thickness data in the adhesive layer positioning sequence;

[0068] S6, cooperatively adjusting the stripping path, stripping force, and stripping tool rotation speed of the manipulator based on the stripping control comprehensive error, to control the manipulator to complete the adhesive layer stripping operation of the hydraulic oil pipe.

[0069] It should be noted that when the manipulator strips the adhesive layer, it first needs to collect environmental information through the visual and tactile perception systems; the image sequence records the appearance of the hydraulic oil pipe and its adhesive layer edge information; the tactile signal sequence records the contact condition (such as pressure, contact time, etc.) of the manipulator and the positioning block at both ends of the oil pipe; specific example: assuming that the surface of the hydraulic oil pipe has a rubber adhesive layer, when the manipulator contacts the surface of the oil pipe, it uses a high-definition camera to take an image of the oil pipe and obtains the edge line of the adhesive layer; in terms of tactile feedback, the manipulator senses the pressure signal of the positioning block to determine the force change of its contact with the positioning block; for example, if the manipulator applies too much force, the tactile sensor will detect a strong feedback, so that the force can be adjusted in the subsequent steps;

[0070] The edge features of the adhesive layer are extracted from the image sequence, which helps to determine the shape, thickness, and other information of the adhesive layer; at the same time, the tactile signal sequence helps to determine the contact position of the manipulator and the positioning block, and then the target positioning point of the hydraulic oil pipe can be calculated; specific example: through image processing algorithms, the edge of the adhesive layer of the hydraulic oil pipe is identified; for example, if the adhesive layer has irregular thickness, the algorithm will find different edges in the image and mark the area where the thickness changes; the tactile signal sequence can help determine the relative position of the positioning block at both ends of the oil pipe; by analyzing these data, the position where the manipulator should contact is determined, thereby ensuring the accuracy of the stripping action;

[0071] The quasi-de Bruijn sequence is a mathematical encoding method for converting the target positioning point sequence into spatial distribution characteristics; in this stage, the spatial distribution characteristics required for the manipulator operation are obtained through the preliminary decoding of the target positioning points; specific example: assuming that the target positioning point sequence is the starting point position of multiple adhesive layer stripping on the oil pipe, the quasi-de Bruijn sequence can help convert these points into more uniformly distributed coordinates; for example, on a circular hydraulic oil pipe, the quasi-de Bruijn sequence can help determine the optimal stripping path, making the manipulator's action more uniform and error-free;

[0072] According to the spatial distribution characteristics, a target grid is generated, and reference points are extracted therefrom to help determine the contact position of the manipulator with the oil pipe; then, through the complete decoding of the quasi-De Bruijn sequence, the contact position deviation between the manipulator and the oil pipe can be calculated; specific example: assuming that the surface of the oil pipe is irregular, the manipulator needs to generate a virtual grid on the surface of the oil pipe to accurately locate each contact point; for example, the manipulator can make fine adjustments by calculating the deviation of the reference points (such as the center points of the grid) in the grid from the target points (the stripping area); if the manipulator deviates from the ideal trajectory, the contact position deviation sequence can feedback this information to help correct the path of the manipulator;

[0073] According to the contact position deviation of the manipulator with the oil pipe, the thickness data of the adhesive layer positioning, and the contact pressure deviation, the comprehensive error of the stripping control is calculated; the goal of this step is to ensure the accuracy and quality of the stripping process by analyzing all control errors; specific example: assuming that during the stripping process, the contact position of the manipulator deviates slightly, or the contact pressure feedback by the tactile sensor exceeds the expected range; by calculating these deviations, the stripping control system can derive a comprehensive error value; for example, if the thickness of the adhesive layer of the oil pipe is uneven, the comprehensive error will increase accordingly, and the system will calculate the best adjustment scheme;

[0074] The stripping control comprehensive error is used to adjust the operating parameters of the manipulator, such as the stripping path, the stripping force, and the speed of the stripping cutter; through these adjustments, the manipulator can more accurately complete the task of stripping the adhesive layer of the hydraulic oil pipe; specific example: assuming that during the stripping process, the comprehensive error indicates that the stripping force of the manipulator is too large, which may cause damage to the surface of the oil pipe; the system will automatically adjust the stripping force and the speed of the cutter to reduce the contact force and optimize the speed of the cutter to ensure that the adhesive layer can be stripped smoothly without damaging the oil pipe; at the same time, the stripping path will also be corrected according to the error, so that the manipulator strips along the optimal path.

[0075] In an optional embodiment, based on the oil pipe image sequence, the adhesive layer edge features are extracted, the target positioning point sequence corresponding to the positioning block group is determined according to the adhesive layer edge features and the tactile contact signal sequence, and the adhesive layer positioning sequence and the contact pressure deviation sequence are generated, including:

[0076] The noise removal processing and the gray scale enhancement processing are performed on each frame of the oil pipe image in the oil pipe image sequence;

[0077] The edge contour extraction algorithm is used to extract the adhesive layer edge in the processed oil pipe image, the coverage area of the adhesive layer on the surface of the oil pipe is determined, and the adhesive layer positioning sequence is formed in the stripping order;

[0078] The actual contact pressure value of each contact time is extracted from the tactile contact signal sequence, the difference between the actual contact pressure value and the preset standard contact pressure value is calculated, and a contact pressure deviation sequence is formed in the order of contact time;

[0079] The positioning block corresponding to the tactile contact signal in the oil pipe image is identified, the center point coordinates of the positioning block are extracted, and a target positioning point sequence is formed in the contact order.

[0080] It should be noted that each frame of image in the oil pipe image sequence may have noise or uneven brightness problems; in order to more clearly identify the edge of the adhesive layer, it is necessary to first perform denoising processing (remove unnecessary noise) and gray scale enhancement (enhance the contrast and details of the image); specific example: assuming that the oil pipe image is taken, there are some environmental disturbances such as light reflection or low camera resolution, resulting in noise in the image; at this time, the noise is removed by a denoising algorithm (such as median filtering), and then gray scale enhancement is performed to make the edge of the adhesive layer of the image more obvious, thereby improving the accuracy of subsequent edge extraction;

[0081] After image preprocessing, an edge detection algorithm (such as Canny edge detection) is used to extract the edge of the adhesive layer in the image; the edge contour extraction algorithm can accurately identify the boundary of the adhesive layer on the surface of the oil pipe, and then help determine the coverage area of the adhesive layer on the surface of the oil pipe; specific example: assuming that the adhesive layer of the oil pipe is uniform, and the surface of the oil pipe has obvious adhesive layer coverage; when the image is enhanced by gray scale, the Canny edge detection algorithm can clearly detect the contour of the adhesive layer; for example, the edge of the adhesive layer may be an irregular curve, which can be accurately positioned after extraction by the algorithm, providing data support for the planning of the adhesive stripping path;

[0082] According to the extracted adhesive layer edge, a positioning sequence is formed in the adhesive stripping order (usually from the starting point to the end point of the adhesive layer); this sequence can be used to guide the movement path of the mechanical hand; this sequence indicates the key positions of the adhesive layer being stripped from the surface of the oil pipe; specific example: assuming that the adhesive layer covers part of the oil pipe and is gradually stripped from one end of the oil pipe; through the edge extraction algorithm, the system can record the starting point, end point and each stripping point of the adhesive layer in order to form an adhesive layer positioning sequence; for example, the positioning points may be points on the surface of the oil pipe at regular intervals (such as every 10 mm); these points will help the mechanical hand accurately perform the adhesive stripping task;

[0083] The tactile sensor can record the actual contact pressure when the robot contacts the oil pipe; by analyzing these pressure signals, the actual pressure value of each contact point can be obtained; these values will be compared with the preset standard contact pressure value to calculate the pressure deviation; specific example: assuming that the robot is in the process of stripping, the tactile sensor will record the pressure value of the contact point every time it touches the oil pipe; for example, in a stripping operation, the contact pressure value is 5N, and the preset standard pressure value is 4N; then, the pressure deviation is 5N-4N=1N; in this way, the system can monitor the contact of the robot in real time, avoiding uneven stripping or damage to the oil pipe due to excessive or insufficient pressure;

[0084] The difference between the actual contact pressure at each contact moment and the preset pressure value will be arranged in chronological order to form a contact pressure deviation sequence; this sequence can help analyze the trend of pressure control during stripping and optimize the control strategy of the robot; specific example: assuming that the robot records the contact pressure every time it contacts a position during stripping; suppose the pressure deviation at different contact points is 1N, 0.8N, 0.5N, 1.2N, etc.; by arranging these pressure deviations in order, a contact pressure deviation sequence is formed; this sequence can help the system determine whether the stripping operation is smooth, and if the pressure deviation is too large in some places, the robot's force can be adjusted to correct it;

[0085] By analyzing the image of the oil pipe, the positioning block corresponding to the tactile contact signal is identified; the positioning block refers to the key reference point used for positioning when the robot operates, which can be a specific area on the surface of the oil pipe, helping the robot to determine its position; specific example: for example, the two ends of the oil pipe may have some markings or fixed devices, which can be used as positioning blocks; in the image, the center point coordinates of these positioning blocks will be extracted for matching with the tactile signal; for example, the center point coordinates of the two end positioning blocks of the oil pipe are (50mm, 0mm) and (150mm, 0mm), which can be used as the target positioning points of the robot in the subsequent stripping process;

[0086] The center point coordinates of the positioning block corresponding to the tactile contact signal are arranged in contact order to form a target positioning point sequence; this sequence is used to guide the robot to accurately position during stripping; specific example: assuming that there are multiple positioning blocks on the surface of the oil pipe (for example, several key points are marked on the oil pipe as contact references), these points are arranged in contact order; for example, the tactile signal records the order as (50mm, 0mm), (80mm, 0mm), (120mm, 0mm), (150mm, 0mm); these points form a target positioning point sequence, and the robot can move according to this sequence to ensure correct trajectory stripping along the oil pipe.

[0087] In an optional embodiment, the target positioning point sequence is preliminarily decoded based on the quasi-De Bruijn sequence coding rule of the positioning block group, and the spatial distribution characteristics corresponding to the target positioning point sequence are determined, including:

[0088] The position of each positioning block in the positioning block group is uniquely coded by the quasi-De Bruijn sequence, and the coding information is pre-stored in the manipulator control system;

[0089] The coordinate parameters of each positioning point in the target positioning point sequence are obtained, and the pre-stored quasi-De Bruijn coding information is matched to determine the positioning block number corresponding to each positioning point.

[0090] According to the arrangement order of the positioning block number, the distribution law of the positioning point in the axial and radial directions of the oil pipe is analyzed to form the spatial distribution characteristics; wherein the spatial distribution characteristics include the axial interval distance and the radial offset trend of the positioning point.

[0091] It should be noted that the quasi-De Bruijn sequence is a special sequence, and its characteristic is that each possible permutation combination appears in the sequence once and is arranged in a certain specific order; using this sequence to code the positioning block means that the number of each positioning block is unique and can be identified and positioned according to the order of the quasi-De Bruijn sequence; specific example: assuming that 5 positioning blocks are set on the surface of the oil pipe, which are located in different areas of the oil pipe; these positioning blocks are coded by the quasi-De Bruijn sequence, for example: the first positioning block is coded as "101", the second positioning block is coded as "110", the third positioning block is coded as "011", the fourth positioning block is coded as "100", and the fifth positioning block is coded as "010"; these coding information will be pre-stored in the control system of the manipulator for subsequent use in the matching of the target positioning point;

[0092] The target positioning point sequence is the path that the robot needs to track, involving the specific coordinates of each positioning point. These coordinates are generated in the system and recorded through sensors or other positioning technologies. Specific example: suppose there are 5 points in the target positioning point sequence, located at different positions of the oil pipe, with the following coordinate parameters: positioning point 1: coordinates (10mm, 0mm), positioning point 2: coordinates (30mm, 0mm), positioning point 3: coordinates (50mm, 0mm), positioning point 4: coordinates (70mm, 0mm), positioning point 5: coordinates (90mm, 0mm). According to the coordinate information of each target positioning point, the system matches with the pre-stored quasi-De Bruijn code information to determine which positioning block each positioning point corresponds to. In this way, the system knows the specific position of each positioning point, ensuring accurate positioning of the robot. Specific example: positioning point 1 (10mm, 0mm) corresponds to code "101", i.e. the 1st positioning block, positioning point 2 (30mm, 0mm) corresponds to code "110", i.e. the 2nd positioning block, positioning point 3 (50mm, 0mm) corresponds to code "011", i.e. the 3rd positioning block, positioning point 4 (70mm, 0mm) corresponds to code "100", i.e. the 4th positioning block, positioning point 5 (90mm, 0mm) corresponds to code "010", i.e. the 5th positioning block. Through this matching method, each target positioning point can be associated with its corresponding positioning block.

[0093] By analyzing the order of positioning block numbers, the system can understand the distribution pattern of these positioning points on the oil pipe, including their axial (along the length of the oil pipe) and radial (cross-sectional direction of the oil pipe) spacing and offset trends. Axial spacing refers to the distance between adjacent positioning points, and radial offset trend refers to the offset change of the positioning points in the cross-sectional direction of the oil pipe. Specific example: axial spacing: suppose the distance between positioning point 1 and positioning point 2 is 20mm, the distance between positioning point 2 and positioning point 3 is 20mm, and so on. This spacing is part of the axial distribution. After analysis, it can be found that the spacing between the target positioning points in the axial direction is uniform (every 20mm a point). Radial offset: suppose the radial coordinate of positioning point 1 is (0mm), the radial coordinate of positioning point 2 is (1mm), the radial coordinate of positioning point 3 is (2mm), the radial coordinate of positioning point 4 is (2mm), and the radial coordinate of positioning point 5 is (3mm). By analyzing these coordinates, it can be seen that the radial coordinates gradually shift in the first few positioning points, forming a trend, i.e. the radial offset of the positioning points shows a gradually increasing trend.

[0094] In an optional embodiment, a target grid is established according to the spatial distribution characteristics, and a sequence of reference points in the target grid is obtained, including:

[0095] The target direction of the target grid is determined based on the axial interval distance in the spatial distribution feature, and the target direction is consistent with the axial direction of the hydraulic oil pipe;

[0096] A positioning point near the center of the oil pipe cross section in the target positioning point sequence is selected as a reference point, the distance average of all positioning points to the reference point is calculated, and a target orthogonal point of the target grid is determined;

[0097] The grid width of the target grid is determined according to the radial deviation trend in the spatial distribution feature and in combination with the preset interval of the positioning block group;

[0098] A two-dimensional target grid is established with the target orthogonal point as the origin, the target direction as the X axis, and the radial direction as the Y axis, the grid points of the target grid are taken as reference points, and a reference point sequence is formed in the order of coordinates.

[0099] It should be noted that in the foregoing steps, the distribution features of the positioning points in the axial and radial directions have been analyzed; the axial interval distance can help determine the target direction of the target grid, that is, along the length direction of the oil pipe (usually consistent with the axial direction of the hydraulic oil pipe); this means that the target direction of the grid will be the axial direction of the oil pipe; specific example: assuming that the axial interval between every two positioning points is 20 mm through the analysis of the axial interval between the positioning points; this interval can be taken as the step length of the grid, and the grid is established in the axial direction of the oil pipe; the target direction is the axial direction of the oil pipe, which can be taken as a reference along the central axis of the oil pipe to determine the direction of the grid;

[0100] The target orthogonal point of the target grid is the origin of the grid, and this point is usually a positioning point near the center of the oil pipe cross section; in order to ensure the symmetry and accuracy of the grid, the distance average of all positioning points to the reference point is calculated and taken as the basis for determining the target orthogonal point of the grid; specific example: assuming that the coordinates of the positioning points in the target positioning point sequence are as follows: positioning point 1: coordinates (10 mm, 0 mm), positioning point 2: coordinates (30 mm, 0 mm), positioning point 3: coordinates (50 mm, 0 mm), positioning point 4: coordinates (70 mm, 0 mm), and positioning point 5: coordinates (90 mm, 0 mm); positioning point 3 (50 mm, 0 mm) is selected as the reference point; the axial distances of other positioning points to the reference point are calculated: the distance of positioning point 1 to the reference point is |10 mm-50 mm|=40 mm, the distance of positioning point 2 to the reference point is |30 mm-50 mm|=20 mm, the distance of positioning point 4 to the reference point is |70 mm-50 mm|=20 mm, and the distance of positioning point 5 to the reference point is |90 mm-50 mm|=40 mm; the average of all distances is calculated: the distance average=(40 mm+20 mm+20 mm+40 mm) / 4=30 mm; through the calculation, the position of the target orthogonal point can be determined, and the point is usually the geometric center of the oil pipe or a point close to the center;

[0101] The radial offset trend in the spatial distribution feature reflects the change in the position of the positioning point on the tubing cross-section; according to the offset trend and in combination with the preset spacing of the positioning block group (such as the fixed distance between the positioning blocks), the width of the target grid can be determined; specific example: assuming that the change in the radial coordinate is analyzed, and it is found that the radial offset trend of the positioning point is gradually increasing; for example: the radial coordinate of positioning point 1 is (0 mm), the radial coordinate of positioning point 2 is (1 mm), the radial coordinate of positioning point 3 is (2 mm), the radial coordinate of positioning point 4 is (2 mm), and the radial coordinate of positioning point 5 is (3 mm); in combination with the preset spacing (for example, the spacing of the positioning block group is 1 mm), it can be concluded that the width of the grid can be set to 1 mm, so as to better track the change of the positioning point in the radial direction;

[0102] According to the positioning of the target orthogonal point, a two-dimensional coordinate system is established, in which the target direction (axial direction) is the X axis and the radial direction is the Y axis; then the grid is gradually constructed according to the grid width, and the grid points in the grid are generated as reference points; these reference points can be used for positioning, control and path planning; specific example: assuming that the target orthogonal point (reference point) is located at the origin (0 mm, 0 mm) of the tubing coordinate system, the target direction (X axis) is the axial direction of the tubing, and the radial direction (Y axis) is the radial direction on the tubing cross-section; the grid width is set to 1 mm, so that the grid points of the two-dimensional grid can be distributed as follows: the X axis direction (axial) step is 1 mm, and the Y axis direction (radial) step is 1 mm; the reference point sequence will be as follows: (0 mm, 0 mm), (1 mm, 0 mm), (2 mm, 0 mm), (3 mm, 0 mm), (4 mm, 0 mm), (0 mm, 1 mm), (1 mm, 1 mm), (2 mm, 1 mm), (3 mm, 1 mm), (4 mm, 1 mm), these reference points as reference points in the grid, arranged in coordinate order, form a complete target grid sequence for subsequent positioning and control.

[0103] In an optional embodiment, the grid width of the target grid is determined according to the radial offset trend in the spatial distribution feature and in combination with the preset spacing of the positioning block group, comprising:

[0104] Obtaining the center spacing of adjacent positioning blocks in the positioning block group, calculating the average value of all adjacent spacings to obtain a basic spacing value;

[0105] Analyzing the radial offset trend in the spatial distribution feature, and counting the maximum offset range of the target positioning point in the radial direction;

[0106] The grid width of the target grid is obtained by rounding the ratio of the basic spacing value to the radial maximum deviation range; the numerical value of the grid width is positively correlated with the pipe diameter specification of the hydraulic oil pipe.

[0107] It should be noted that the average value of the center spacing of all adjacent positioning blocks can be obtained by calculating the average value of the center spacing of all adjacent positioning blocks; this value will be used as a reference for the preliminary setting of the grid width; specific example: assuming that a group of positioning blocks have the following center coordinates (unit: mm): positioning block 1: position (10mm, 0mm), positioning block 2: position (20mm, 0mm), positioning block 3: position (30mm, 0mm), positioning block 4: position (40mm, 0mm), the center spacing between adjacent positioning blocks is calculated: the spacing between positioning block 1 and positioning block 2: |20mm-10mm|=10mm, the spacing between positioning block 2 and positioning block 3: |30mm-20mm|=10mm, the spacing between positioning block 3 and positioning block 4: |40mm-30mm|=10mm, the average spacing of all adjacent spacings is: average spacing=(10mm+10mm+10mm) / 3=10mm, therefore, the basic spacing value is 10mm;

[0108] Next, the deviation trend of the target positioning point in the radial direction needs to be considered; the radial deviation is the deviation of the positioning point relative to the center axis of the oil pipe; the distribution of the positioning points on the oil pipe cross section can be understood by statistically analyzing the maximum deviation range of the target positioning point in the radial direction; specific example: assuming that the radial coordinate data of the target positioning point is as follows: positioning point 1: radial coordinate (0mm), positioning point 2: radial coordinate (2mm), positioning point 3: radial coordinate (5mm), positioning point 4: radial coordinate (8mm), positioning point 5: radial coordinate (10mm), the maximum radial deviation of these positioning points is: 10mm (i.e. the radial coordinate of positioning point 5); therefore, the maximum deviation range of the target positioning point in the radial direction is 10mm;

[0109] The ratio of the basic spacing value (10mm obtained in step 1) to the radial maximum deviation range (10mm obtained in step 2) is calculated; the calculated ratio will tell the reasonable range of the grid width; then, the ratio is rounded to determine an integer value as the grid width of the target grid; specific example: according to the previous calculation, the basic spacing value is 10mm and the radial maximum deviation range is 10mm; the ratio of the two values is calculated: ratio=10mm / 10mm=1, after rounding, the grid width is 1mm;

[0110] The grid width is closely related to the pipe diameter specification of the hydraulic oil pipe; generally, when the pipe diameter of the hydraulic oil pipe is large, the grid width of the target grid should also be large; through the grid width calculated above, a suitable grid step can be selected to ensure that the grid meets the requirements in actual application; specific example: assuming that the pipe diameter specification of the hydraulic oil pipe is 10 mm; through the above calculation, the grid width obtained is also 1 mm, which meets the common application requirements of the hydraulic oil pipe; if the pipe diameter specification of the oil pipe is larger, for example, 20 mm, it may lead to the grid width needing to be increased, for example, 2 mm or higher.

[0111] In an optional embodiment, based on the offset information of the target positioning point sequence and the reference point sequence, the contact position deviation sequence of the manipulator and the hydraulic oil pipe is determined through the quasi-de Bruijn sequence complete decoding, comprising:

[0112] Calculate the X-axis offset and Y-axis offset between each target positioning point and the nearest reference point to form an offset information set;

[0113] Input the offset information set into the quasi-de Bruijn decoding model, and the model reversely deduces the actual spatial position of the positioning block based on the pre-stored encoding rule;

[0114] Compare the actual spatial position deduced with the theoretical position of the positioning block, calculate the axial deviation value and the radial deviation value of each contact position, and arrange them in the contact order to form the contact position deviation sequence.

[0115] It should be noted that for each target positioning point, find its nearest reference point in the reference point sequence; then, calculate the offset of each target positioning point and the corresponding reference point on the X axis and Y axis; the offset is the difference between the target point and the reference point on the respective coordinate axis, the X axis offset represents the horizontal distance, and the Y axis offset represents the vertical distance; specific example: assuming the coordinates of the target positioning points and reference points are as follows (unit: mm): target positioning point: positioning point 1: (5mm, 10mm), positioning point 2: (20mm, 30mm), positioning point 3: (35mm, 50mm), reference point: reference point 1: (0mm, 0mm), reference point 2: (15mm, 25mm), reference point 3: (40mm, 60mm), calculate the offset of the target positioning point and the reference point: the offset of positioning point 1 and reference point 1: X axis offset: |5mm-0mm|=5mm, Y axis offset: |10mm-0mm|=10mm, the offset of positioning point 2 and reference point 2: X axis offset: |20mm-15mm|=5mm, Y axis offset: |30mm-25mm|=5mm, the offset of positioning point 3 and reference point 3: X axis offset: |35mm-40mm|=5mm, Y axis offset: |50mm-60mm|=10mm, offset information set: positioning point 1 offset: X axis=5mm, Y axis=10mm, positioning point 2 offset: X axis=5mm, Y axis=5mm, positioning point 3 offset: X axis=5mm, Y axis=10mm;

[0116] The offset information set (i.e. the offset of each target positioning point relative to the nearest reference point) will be input into the quasi-de Bruijn decoding model; the quasi-de Bruijn sequence is an encoding method used in signal processing to compress data, which can effectively deduce the actual spatial position of the target positioning point; the decoding model will deduce the actual spatial position based on pre-stored encoding rules (such as the encoding algorithm of the de Bruijn sequence); these rules can help accurately determine the actual position of each positioning block, especially in complex spatial distribution cases; specific example: assuming there is a quasi-de Bruijn decoding model that can deduce the actual spatial position of the target positioning point based on pre-stored rules; after the offset information set is input into the decoding model, the actual spatial position deduced by the model is: the actual spatial position of positioning point 1: (6mm, 11mm), the actual spatial position of positioning point 2: (21mm, 32mm), the actual spatial position of positioning point 3: (36mm, 52mm);

[0117] In this step, the actual spatial position derived by the decoding model is compared with the theoretical position (expected position) of the positioning block, and the axial deviation (deviation in the X-axis direction) and radial deviation (combined deviation in the X-axis and Y-axis, i.e., distance from the origin) of each contact position are calculated; specific example: assuming that the theoretical positions of the positioning blocks are as follows: the theoretical position of positioning point 1 is (5mm, 10mm), the theoretical position of positioning point 2 is (20mm, 30mm), and the theoretical position of positioning point 3 is (35mm, 50mm), the actual spatial position derived by the decoding model is compared with the theoretical position.

[0118] Finally, according to the calculated axial deviation and radial deviation, the deviation values are arranged in contact order to form a contact position deviation sequence; this sequence can help understand the deviation of the contact position of the manipulator and the hydraulic oil pipe, and optimize future operations; specific example: deviation sequence (arranged in contact order): positioning point 1: X-axis deviation = 1mm, Y-axis deviation = 1mm, radial deviation = 1.03mm, positioning point 2: X-axis deviation = 1mm, Y-axis deviation = 2mm, radial deviation = 2.03mm, positioning point 3: X-axis deviation = 1mm, Y-axis deviation = 2mm, radial deviation = 3.08mm.

[0119] In an optional embodiment, based on the contact position deviation sequence, the contact pressure deviation sequence, and the glue layer positioning sequence, the glue stripping control comprehensive error is calculated, including:

[0120] The average thickness value of each glue layer region is extracted from the glue layer positioning sequence to form a glue layer thickness sequence;

[0121] The contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness sequence are respectively assigned a weight coefficient, and the weighted sum is obtained to obtain the glue stripping control comprehensive error.

[0122] It should be noted that the average thickness of each glue layer region is extracted from the glue layer positioning sequence; the glue layer positioning sequence refers to the data describing the position of the glue layer in space, and the average thickness of each glue layer region refers to the average value of the glue layer thickness of all points in the region; a weight coefficient is assigned to each deviation sequence (contact position deviation, contact pressure deviation, and glue layer thickness deviation), and the weight coefficient represents the importance of each factor to the final glue stripping control comprehensive error; then, the deviation of each sequence is multiplied by the corresponding weight coefficient to obtain the weighted error value, and finally the weighted sum is obtained to obtain the comprehensive error.

[0123] In an optional embodiment, based on the glue stripping control comprehensive error, the glue stripping path of the manipulator is adjusted in coordination with the glue stripping force and the glue stripping tool speed, including:

[0124] According to the axial deviation value and the radial deviation value in the contact position deviation sequence, the X-axis compensation amount and the Y-axis compensation amount of the stripping path are calculated, and the motion trajectory of the manipulator is adjusted;

[0125] Based on the contact pressure deviation sequence, the stripping force adjustment amount is calculated, and the force adjustment is realized by changing the oil supply pressure of the manipulator driving cylinder;

[0126] According to the glue layer thickness sequence and the stripping control comprehensive error, the rotational speed adjustment amount of the stripping cutter is calculated, and the rotational speed adjustment is realized by changing the pulse frequency of the cutter driving motor;

[0127] The path compensation amount, the force adjustment amount and the rotational speed adjustment amount are input into the manipulator control module synchronously, the stripping parameters are updated and the stripping operation is performed.

[0128] It should be noted that the contact position deviation sequence provides the axial and radial deviations of the contact position during the motion of the manipulator; the axial deviation and the radial deviation represent the position errors of the manipulator in the X-axis and Y-axis directions respectively; in order to maintain the accuracy of the stripping operation, the path of the manipulator needs to be adjusted according to these deviations, i.e. the X-axis and Y-axis are compensated to ensure that the cutter operates along the correct trajectory; specific example: assuming that there is a certain deviation in the path of the manipulator during the execution of the stripping operation; the following contact position deviation data is obtained through the sensor: contact position deviation sequence: axial deviation: 0.3mm, radial deviation: 0.2mm, through the control algorithm, the compensation amount of X-axis and Y-axis can be calculated; assuming that the compensation amount needs to be adjusted according to a certain proportion, the proportion coefficient is 0.5 (i.e. half of the compensation amount): X-axis compensation amount: 0.3mm×0.5=0.15mm, Y-axis compensation amount: 0.2mm×0.5=0.1mm, which means that the manipulator needs to adjust its X-axis and Y-axis positions, respectively offset by 0.15mm and 0.1mm, to ensure the accuracy of the path; finally, the compensation amount is input through the motion control system of the manipulator to update the motion trajectory;

[0129] The contact pressure deviation sequence describes the change of the contact pressure in the stripping operation; the stripping force refers to the force exerted by the robot during the stripping process; if the contact pressure deviation is too large (i.e. too high or too low), the stripping force needs to be adjusted; the force adjustment amount is calculated, and the force adjustment is realized by changing the oil supply pressure of the robot driving cylinder; specific example: suppose the contact pressure deviation sequence obtained through the sensor is as follows: contact pressure deviation sequence: deviation 1: 0.1 MPa, deviation 2: 0.2 MPa, suppose the force adjustment algorithm is: if the contact pressure deviation is greater than 0.1 MPa, increase the stripping force; if the contact pressure deviation is less than 0.1 MPa, decrease the stripping force; according to this rule, the calculated force adjustment amount is as follows: force adjustment amount: for deviation 1 (0.1 MPa), no adjustment; for deviation 2 (0.2 MPa), the force needs to be increased by 0.1 MPa; by adjusting the oil supply pressure of the robot driving cylinder, it is increased to the target pressure, ensuring that the stripping force meets the requirements;

[0130] The glue layer thickness sequence and the stripping control comprehensive error affect the adjustment of the tool speed; the thickness change of the glue layer will affect the cutting efficiency and cutting load of the tool, and then the tool speed needs to be adjusted according to different thicknesses; the comprehensive error is the error calculated according to the contact position, contact pressure, glue layer thickness and other factors, reflecting the possible error range in the stripping operation; according to these information, the speed of the tool needs to be adjusted, usually by changing the pulse frequency of the motor; specific example: suppose the following glue layer thickness data and stripping control comprehensive error are obtained through the sensor: glue layer thickness sequence: thickness 1: 2.5 mm, thickness 2: 3.0 mm, stripping control comprehensive error: 0.6 mm, suppose the relationship between the tool speed and the glue layer thickness and the comprehensive error is: thicker glue layer and larger error require higher speed, thinner glue layer and smaller error require lower speed; the speed adjustment algorithm is as follows: speed adjustment amount = thickness × comprehensive error coefficient thickness × comprehensive error coefficient, suppose the comprehensive error coefficient is 1000 (i.e. for every 1 mm of error, the speed increases by 1000 pulses); for thickness 1 (2.5 mm), the speed adjustment amount is: 2.5 mm × 0.6 mm × 1000 = 1500, for thickness 2 (3.0 mm), the speed adjustment amount is: 3.0 mm × 0.6 mm × 1000 = 1800, the speed of the tool is adjusted by increasing or decreasing the pulse frequency of the motor, so that it adapts to different thicknesses of the glue layer and the comprehensive error;

[0131] The calculated path compensation, force adjustment, and speed adjustment values ​​need to be synchronously input into the robot's control module. These parameters will be updated together to ensure that the robot can adjust the path, force, and speed based on real-time feedback, thereby ensuring the accuracy and efficiency of the peeling operation. For example, assuming the following have been calculated: path compensation: X-axis compensation 0.15mm, Y-axis compensation 0.1mm, force adjustment: increase the oil supply pressure by 0.1MPa, and speed adjustment: increase the frequency by 1500 pulses, after inputting these parameters into the robot's control module, the control module will adjust the robot's motion trajectory, applied force, and cutter speed in real time, and begin to execute the peeling operation, ensuring the accuracy and stability of the operation.

[0132] In an optional embodiment, an edge contour extraction algorithm is used to extract the edges of the adhesive layer in the processed tubing image to determine the coverage area of ​​the adhesive layer on the tubing surface, including:

[0133] Gaussian filtering algorithm is used to denoise the tubing image to eliminate salt-and-pepper noise in the image;

[0134] The grayscale contrast between the adhesive layer and the tubing body is improved by using a histogram equalization algorithm to enhance the grayscale of the denoised image.

[0135] The Canny edge detection algorithm is used to extract edge pixels in the enhanced image, and connected component analysis is performed on the edge pixels to remove isolated edge points.

[0136] A polynomial fitting algorithm is used to fit curves to the effective edge pixels to form a continuous adhesive layer edge contour line.

[0137] Based on the closed area of ​​the edge contour line, determine the coverage area of ​​the adhesive layer on the surface of the tubing, and record the coordinate range of the coverage area.

[0138] Example 2, please refer to Figure 2 This invention provides a technical solution: a robotic arm control system for peeling adhesive from hydraulic hoses, applicable to the aforementioned robotic arm control method for peeling adhesive from hydraulic hoses, comprising:

[0139] Data acquisition unit 1 is used to acquire the hydraulic hose image sequence and tactile contact signal sequence when the robot arm contacts the hydraulic hose according to the preset peeling path; the hydraulic hose is equipped with positioning block groups at both ends, and the tactile contact signal sequence includes the contact feedback signal between the robot arm and the positioning block group;

[0140] Feature extraction unit 2 is used to extract the edge features of the adhesive layer based on the oil pipe image sequence, determine the target positioning point sequence corresponding to the positioning block group according to the edge features of the adhesive layer and the tactile contact signal sequence, and generate the adhesive layer positioning sequence and the contact pressure deviation sequence.

[0141] The distribution coding unit 3 is configured to preliminarily decode the target positioning point sequence based on the quasi-De Bruijn sequence coding rule of the positioning block group, and determine the spatial distribution characteristics corresponding to the target positioning point sequence.

[0142] The deviation calculation unit 4 is configured to establish a target grid according to the spatial distribution characteristics, acquire a reference point sequence in the target grid, and determine a contact position deviation sequence of the manipulator and the hydraulic oil pipe based on the offset information of the target positioning point sequence and the reference point sequence through complete decoding of the quasi-De Bruijn sequence.

[0143] The error calculation unit 5 is configured to calculate a stripping control comprehensive error according to the contact position deviation sequence, the contact pressure deviation sequence, and the adhesive layer thickness data in the adhesive layer positioning sequence.

[0144] The stripping control unit 6 is configured to cooperatively adjust the stripping path, the stripping force, and the stripping tool rotating speed of the manipulator based on the stripping control comprehensive error, and control the manipulator to complete the adhesive layer stripping operation of the hydraulic oil pipe.

[0145] The embodiments of the present application are described in detail above with reference to the drawings, but the present application is not limited thereto, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present application.

Claims

1. A method for controlling a robot for hydraulic tubing stripping, characterized in that, The method comprises the following steps: obtaining an oil pipe image sequence and a tactile contact signal sequence when the manipulator contacts the hydraulic oil pipe along a preset stripping path; the two ends of the hydraulic oil pipe are provided with a positioning block group, and the tactile contact signal sequence comprises contact feedback signals between the manipulator and the positioning block group; extracting a rubber layer edge feature based on the oil pipe image sequence, determining a target positioning point sequence corresponding to the positioning block group according to the rubber layer edge feature and the tactile contact signal sequence, generating a rubber layer positioning sequence and a contact pressure deviation sequence; based on the quasi-de Bruijn sequence coding rule of the positioning block group, the target positioning point sequence is preliminarily decoded to determine the spatial distribution feature corresponding to the target positioning point sequence; establishing a target grid according to the spatial distribution feature, and obtaining a reference point sequence in the target grid; based on the offset information between the target positioning point sequence and the reference point sequence, the contact position deviation sequence between the manipulator and the hydraulic oil pipe is determined through complete decoding of the quasi-de Bruijn sequence; based on the contact position deviation sequence, the contact pressure deviation sequence and the rubber layer thickness data in the rubber layer positioning sequence, a stripping control comprehensive error is calculated; based on the stripping control comprehensive error, the stripping path, the stripping force and the stripping tool rotating speed of the manipulator are cooperatively adjusted to control the manipulator to complete the rubber layer stripping operation of the hydraulic oil pipe; wherein, based on the oil pipe image sequence, the rubber layer edge feature is extracted, the target positioning point sequence corresponding to the positioning block group is determined according to the rubber layer edge feature and the tactile contact signal sequence, and the rubber layer positioning sequence and the contact pressure deviation sequence are generated, which comprises: performing denoising processing and gray enhancement processing on each frame of oil pipe image in the oil pipe image sequence; an edge contour extraction algorithm is used to extract the rubber layer edge in the processed oil pipe image, determine the coverage area of the rubber layer on the surface of the oil pipe, and form the rubber layer positioning sequence in stripping order; the actual contact pressure value at each contact time is extracted from the tactile contact signal sequence, the difference between the actual contact pressure value and the preset standard contact pressure value is calculated, and the contact pressure deviation sequence is arranged in contact time order; the positioning block corresponding to the tactile contact signal in the oil pipe image is identified, the center point coordinates of the positioning block are extracted, and the target positioning point sequence is composed in contact order.

2. The method of claim 1, wherein, based on the quasi-de Bruijn sequence coding rule of the positioning block group, the target positioning point sequence is preliminarily decoded to determine the spatial distribution feature corresponding to the target positioning point sequence, which comprises: the position of each positioning block in the positioning block group is uniquely coded by the quasi-de Bruijn sequence, and the coding information is pre-stored in the manipulator control system; the coordinate parameters of each positioning point in the target positioning point sequence are obtained, the pre-stored quasi-de Bruijn coding information is matched, and the positioning block number corresponding to each positioning point is determined; according to the arrangement order of the positioning block number, the distribution rule of the positioning point in the axial and radial directions of the oil pipe is analyzed to form the spatial distribution feature; wherein, the spatial distribution feature comprises the axial interval distance and the radial offset trend of the positioning point.

3. The method of claim 2, wherein, based on the spatial distribution feature, a target grid is established, and a reference point sequence in the target grid is obtained, which comprises: based on the axial interval distance in the spatial distribution feature, a target direction of the target grid is determined, and the target direction is consistent with the axial direction of the hydraulic oil pipe; Selecting a positioning point near the center of the oil pipe cross section as a reference point, calculating the average distance of all positioning points to the reference point, and determining the target orthogonal point of the target grid; According to the radial deviation trend in the spatial distribution characteristics, and in combination with the preset interval of the positioning block group, the grid width of the target grid is determined; Taking the target orthogonal point as the origin, the target direction as the X-axis, and the radial direction as the Y-axis, a two-dimensional target grid is established according to the grid width, the grid points of the target grid are taken as the reference points, and a reference point sequence is formed according to the coordinate order.

4. The method of claim 3, wherein, According to the radial deviation trend in the spatial distribution characteristics, and in combination with the preset interval of the positioning block group, the grid width of the target grid is determined, including: Obtaining the center interval of adjacent positioning blocks in the positioning block group, calculating the average value of all adjacent intervals to obtain a basic interval value; Analyzing the radial deviation trend in the spatial distribution characteristics, and statistically analyzing the maximum deviation range of the target positioning point in the radial direction; The ratio of the basic interval value to the maximum radial deviation range is rounded to obtain the grid width of the target grid; the numerical value of the grid width is positively correlated with the pipe diameter specification of the hydraulic oil pipe.

5. The method of claim 4, wherein, Based on the deviation information of the target positioning point sequence and the reference point sequence, the contact position deviation sequence of the manipulator and the hydraulic oil pipe is determined through complete decoding of the quasi-De Bruijn sequence, including: Calculating the X-axis deviation and Y-axis deviation between each target positioning point and the nearest reference point to form a deviation information set; The deviation information set is input into the quasi-De Bruijn decoding model, and the model reversely deduces the actual spatial position of the positioning block based on the pre-stored encoding rule; By comparing the deduced actual spatial position with the theoretical position of the positioning block, the axial deviation value and the radial deviation value of each contact position are calculated, and the contact position deviation sequence is arranged according to the contact order.

6. The method of claim 5, wherein, According to the contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness data in the glue layer positioning sequence, the stripping control comprehensive error is calculated, including: Extracting the average thickness value of each glue layer region from the glue layer positioning sequence to form a glue layer thickness sequence; The contact position deviation sequence, the contact pressure deviation sequence, and the glue layer thickness sequence are respectively assigned weight coefficients, and the stripping control comprehensive error is obtained by weighted summation.

7. The method of claim 6, wherein, Based on the stripping control comprehensive error, the stripping path, the stripping force, and the stripping tool speed of the manipulator are cooperatively adjusted, including: According to the axial deviation value and the radial deviation value in the contact position deviation sequence, the X-axis compensation amount and the Y-axis compensation amount of the stripping path are calculated, and the motion trajectory of the manipulator is adjusted; Based on the contact pressure deviation sequence, the stripping force adjustment amount is calculated, and the force adjustment is realized by changing the oil supply pressure of the manipulator driving cylinder; According to the glue layer thickness sequence and the stripping control comprehensive error, the speed adjustment amount of the stripping tool is calculated, and the speed adjustment is realized by changing the pulse frequency of the tool driving motor; The path compensation amount, the force adjustment amount, and the speed adjustment amount are synchronously input into the manipulator control module, the stripping parameters are updated, and the stripping operation is performed.

8. The method of claim 7, wherein, The edge profile extraction algorithm is used to extract the glue layer edge in the processed oil pipe image, and the coverage area of the glue layer on the oil pipe surface is determined, including: Gaussian filtering algorithm is used to denoise the oil pipe image to eliminate the salt and pepper noise in the image; The denoised image is subjected to gray scale enhancement by a histogram equalization algorithm to improve the gray scale contrast between the adhesive layer and the oil pipe body; The Canny edge detection algorithm is used to extract the edge pixel points in the enhanced image, and the connected domain analysis is performed on the edge pixel points to remove the isolated edge points; The polynomial fitting algorithm is used to perform curve fitting on the effective edge pixel points to form a continuous adhesive layer edge contour line; According to the closed region of the edge contour line, the coverage area of the adhesive layer on the surface of the oil pipe is determined, and the coordinate range of the coverage area is recorded.

9. A mechanical hand control system for hydraulic tubing stripping, which is suitable for the mechanical hand control method for hydraulic tubing stripping according to any one of claims 1-8, characterized in that, It comprises: A data acquisition unit is configured to acquire an oil pipe image sequence and a tactile contact signal sequence when a manipulator contacts a hydraulic oil pipe along a preset adhesive stripping path; the hydraulic oil pipe is provided with a positioning block group at both ends; the tactile contact signal sequence comprises contact feedback signals between the manipulator and the positioning block group; A feature extraction unit is configured to extract adhesive layer edge features based on the oil pipe image sequence, determine a target positioning point sequence corresponding to the positioning block group based on the adhesive layer edge features and the tactile contact signal sequence, and generate an adhesive layer positioning sequence and a contact pressure deviation sequence; A distribution coding unit is configured to preliminarily decode the target positioning point sequence based on a quasi-De Bruijn sequence coding rule of the positioning block group, and determine spatial distribution features corresponding to the target positioning point sequence; A deviation calculation unit is configured to establish a target grid based on the spatial distribution features, acquire a reference point sequence in the target grid, and determine a contact position deviation sequence between the manipulator and the hydraulic oil pipe through complete decoding of the quasi-De Bruijn sequence based on offset information between the target positioning point sequence and the reference point sequence; An error calculation unit is configured to calculate a comprehensive adhesive stripping control error based on the contact position deviation sequence, the contact pressure deviation sequence, and adhesive layer thickness data in the adhesive layer positioning sequence; An adhesive stripping control unit is configured to cooperatively adjust an adhesive stripping path, adhesive stripping force, and adhesive stripping tool rotation speed of the manipulator based on the comprehensive adhesive stripping control error, and control the manipulator to complete adhesive layer stripping of the hydraulic oil pipe; The extraction of the adhesive layer edge features based on the oil pipe image sequence, the determination of the target positioning point sequence corresponding to the positioning block group based on the adhesive layer edge features and the tactile contact signal sequence, and the generation of the adhesive layer positioning sequence and the contact pressure deviation sequence comprise: Each oil pipe image in the oil pipe image sequence is subjected to denoising and gray scale enhancement; An adhesive layer edge in the processed oil pipe image is extracted by an edge contour extraction algorithm, a coverage area of the adhesive layer on the surface of the oil pipe is determined, and an adhesive layer positioning sequence is formed in the order of adhesive stripping; Actual contact pressure values at each contact time are extracted from the tactile contact signal sequence, a difference between the actual contact pressure values and a preset standard contact pressure value is calculated, and a contact pressure deviation sequence is arranged in the order of contact time; The positioning blocks corresponding to the tactile contact signals in the oil pipe image are identified, the center point coordinates of the positioning blocks are extracted, and a target positioning point sequence is formed in the order of contact.

Images