A control method for automatic cap recognition and fastening of a hexagonal screw of a component

Abstract

The present application relates to the field of assembling aerospace engine, in particular to a control method for automatic cap recognition and fastening of hexagonal screws of parts, specifically: completing the calibration of camera internal and external parameters and the calculation of the relationship with the end execution coordinate system of the robot; calibrating and checking the radial and circumferential reference pose of the tightening tool; setting and detecting the angle of the edge feature and the position of the center of the reference screw; completing the cap recognition and butt joint of the tightening tool and the reference screw at the reference position calibration by the robot; detecting the angle and the position of the center of the edge of any screw; calculating the angle difference and the center position offset of any screw in the robot base coordinate system according to the pose of the reference screw, guiding the robot to complete the cap recognition and butt joint of any screw. By setting the torque and angle upper limit conditions, the tightening machine is started to drive the tightening tool to complete the fastening of the screw. The present application reduces the use of personnel in the tightening assembly process of certain engine parts, and improves the quality of products.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of automatic assembly, in particular, a control method for automatic cap recognition and fastening of hexagonal screws of a certain part of a space engine. BACKGROUND

[0002] With the development of domestic space industry, the demand for space supporting products has greatly increased, and space industry chain related enterprises pay more and more attention to the improvement of product assembly quality and efficiency.

[0003] ATTACHED Figure 1 The figure is a top view of a certain type of space engine part. The assembly of the part involves the fastening of three hexagonal screws, and the upper surface of the screw is a regular hexagon. The traditional fastening method is to use a fixed torque open wrench to achieve the fastening process of the screw. Since one side of the three screws is close to the center cylindrical protruding part of the part, during the manual fastening process, after rotating by an angle, the wrench needs to be detached from the screw, rotated counterclockwise, and then reconnected to the screw for fastening. Each time a fastening task is completed, the process needs to be repeated about 6-8 times. Since each space engine has dozens of such parts during assembly, there are hundreds of hexagonal screws to be tightened, which means that hundreds of times of manual fastening are required for each engine. Due to manual fatigue, the surface of some screws is often scratched, affecting the quality of the whole machine, so an automatic control method is urgently needed to reduce personnel input and improve assembly quality.

[0004] In some automatic tightening applications, an industrial robot is used to position an electric tightening shaft, which can reduce personnel input and ensure the assembly quality of the bolt. The application of vision technology can achieve the acquisition of bolt position information and guide the robot to complete precise positioning of the screw. Although vision-guided robots can achieve automatic tightening, the cap recognition and fastening process of the sleeve and the screw involves relative motion. In this assembly process, relative motion between the screw and the sleeve during tightening is not allowed to avoid damaging the paint layer on the upper surface of the screw and affecting the performance of the product. Due to the special relative position of the screw to be tightened in the part, the application of ordinary hexagonal sleeves will interfere with other positions of the part. SUMMARY

[0005] In order to improve the screw fastening quality of the above space engine part, the present application proposes a control method for automatic assembly of hexagonal screw fastening process.

[0006] The technical solution adopted by the present application to achieve the above-mentioned purpose is: a control method for automatic cap recognition and fastening of hexagonal screws of a part, comprising the following steps:

[0007] 1) Use a checkerboard calibration plate to complete the calibration of camera intrinsic and extrinsic parameters;

[0008] 2) Set the camera coordinate system as C, the robot end execution coordinate system as E, and solve the transformation matrix between the two coordinate systems

[0009] 3) Preliminarily calibrate the radial reference pose of the tightening tool through the L-shaped calibration plate;

[0010] 4) Preliminarily calibrate whether the radial reference pose of the tightening tool is qualified through tool detection; when qualified, execute the next step;

[0011] 5) Determine the circumferential reference position of the tightening tool through the laser sensor and the reflective plate;

[0012] 6) Detect the edge features of the reference screw through the industrial camera and set the edge angle α formed by any two adjacent edges and the position (x, y) of the center of the reference screw;

[0013] 7) Embed the reference screw into the split head of the tightening tool that has completed the radial and circumferential reference position calibration through the teaching robot, save the pose of the robot at this time and set the pose as the basic pose of the robot for screw tightening;

[0014] 8) Detect the edge features of any screw through the industrial camera and determine the edge feature angle α ′ and the position (x ′ ,′) of the center of the reference screw;

[0015] 9) Calculate the angle difference Δα and the position offset (Δx, Δy0;

[0016] 10) According to the position offset and the angle difference between the arbitrary screw and the reference screw, adjust the pose of the tightening tool through the robot and complete the cap recognition and docking of the screw;

[0017] 11) Start tightening the screw through the electric tightening machine driving the tightening tool, and judge whether the tightening process is completed through the rotation torque and angle detection constraints.

[0018] Step 1) Place the checkerboard calibration plate on the assembly plane, move the robot to drive the camera to capture images of the calibration plate from 3 different position angles, and obtain the position information of 4 groups of corner points on the checkerboard calibration plate plane each time.

[0019] In step 2), the method for solving the transformation matrix between the camera coordinate system and the robot end execution coordinate system is as follows: solve the camera internal and external parameter matrices of the 3 different position angle images of the checkerboard calibration plate obtained in step 1), obtain and record the pose parameters of the robot end coordinate system TCP0 of the 3 images.

[0020] Step 3) Preliminary setting of the radial reference position of the tightening tool by the calibration plate, the process is as follows:

[0021] 3.1) First, the robot drives the tightening tool to enter the inside of the 90-degree corner from the open side of the L-shaped calibration plate at a lower set speed, so that the side of the tightening tool is tangent to the two ends of the calibration plate, and the lower end of the tool contacts the assembly plane;

[0022] 3.2) Then twist the wedge and observe whether the wedge axis is parallel to the calibration plate; if it is considered that the parallelism meets the set standard, the preliminary calibration of the radial reference position of the tightening tool is completed; otherwise, adjust the position of the wedge by the robot and return to step 3.1).

[0023] Step 4) is specifically: detecting whether the axis of the tightening tool in the reference position is perpendicular to the tightening plane by the tool, including the following steps:

[0024] First, place the tool vertically on the assembly table so that its axis is perpendicular to the assembly plane, and then drive the tightening tool that has completed the preliminary reference position calibration by the robot to move near the detection tool through the X, Y, Z axes of the Cartesian coordinate system, so that the tool can just be embedded in the tightening tool wedge;

[0025] Then, the robot vertically lifts the tightening tool in the Z+ direction, and finally rotates the end of the sixth axis of the robot to ±60°, ±120° and ±180° directions respectively with the position as the circumferential reference, and observes whether the tool can be embedded in the tightening wedge at the six positions without rotating the tool;

[0026] If it can be embedded, it is considered that the axis of the tightening tool is perpendicular to the assembly plane, that is, the reference position is qualified; otherwise, re-execute step 3) operation.

[0027] Step 5) is specifically: first rotate the tightening tool clockwise by the tightening machine, and when the laser sensor detects the reflected light beam of the reflector, immediately stop rotating, then rotate the tightening tool counterclockwise at a lower set speed, and when the light beam emitted by the laser sensor just escapes from the reflector, immediately stop rotating, at this time the position is the reference position of the tightening tool circumferential.

[0028] In step 6), the reference screw setting rule is: first manually twist the screw to make it in the seated state and confirm that the screw axis is perpendicular to the assembly plane, then adjust the angle of the screw so that one of the edges is tangent to the cylindrical edge of the center protrusion of the workpiece, then rotate the screw counterclockwise to set the angle, and finally confirm that the tightening wedge at this time is just not interfered with the edge of the center protrusion of the workpiece after the screw is docked, and set it as the state of the reference screw;

[0029] The edge angle α of the two adjacent edges formed by the edge features of the reference screw and the position (x, y) of the center of the reference screw are detected by the industrial camera, and the specific method is as follows:

[0030] The robot is positioned to the industrial camera to make the reference screw image in the center of the camera field of view, and then the robot posture is adjusted to make the camera detection plane parallel to the assembly plane. The pose data of the robot positioning the camera at this time is recorded and saved, then the screw edge features are identified, and finally the angle α of the edge and the coordinates (x, y) of the center of the edge in the camera coordinate system are calculated.

[0031] In step 7), the specific method is as follows: the teaching robot is moved in the X, Y and Z axis directions of the Cartesian coordinate system, and the Z direction rotation freedom of the TCP0 at the end of the robot is used to make the reference screw embedded into the wedge head of the tightening tool for radial and circumferential reference position calibration. The pose of the robot at this time is saved and set as the basic pose of the robot for tightening the screw.

[0032] In step 11), the specific method is as follows: in order to prevent interference between the tool and the workpiece during tightening, the tightening program in the electric motor tightening machine is set to set the upper limit of the angle of rotation for each execution according to the actual screw and the angle deviation of the reference screw in addition to the target torque of the tightening; when the target torque of the tightening is reached, the tightening is completed; when the upper limit of the rotation angle is reached but the target torque is not reached, the robot returns to the initial position and then executes steps 3), 5), 8), 9), 10) and 11) in turn again to continue the automatic cap recognition and tightening process.

[0033] A control system for automatic cap recognition and tightening of a hexagonal screw of a part, comprising:

[0034] A robot for adjusting the pose and positioning of the tightening tool;

[0035] An electric tightening machine fixed to the end of the sixth axis of the robot for tightening the screw under the condition of setting the target torque and angle limit;

[0036] An industrial camera fixed to one side of the electric tightening machine for detecting the position and angle of the screw;

[0037] A wedge head, which is an open hexagonal tightening wedge head, is inserted into the end of the electric tightening machine for cap recognition and butt joint with the edge of the screw;

[0038] A calibration plate with an L-shaped cross section composed of two plates with a 90-degree angle, which is fixed vertically to the assembly operation table for setting the radial reference of the tightening tool during calibration;

[0039] The detection tool is composed of an upper part, a middle part and a lower end in sequence; the upper part is a cuboid, the end face is a rectangle, the length of the rectangle is the distance between two opposite sides of a hexagonal screw, and the width is the length of one side of the hexagonal screw; when the detection tool is rotated to a certain angle, the upper part can be embedded into the hexagonal split head; the middle part is a circular table, used for connecting the upper part and the lower end; and the lower end is a cylindrical base, used for being placed in the matching groove on the assembly table.

[0040] The laser sensor is fixed on the electric tightening machine through a support, and the reflecting plate is fixed on the tightening tool, and both are used for setting the circumferential reference position of the tightening tool.

[0041] The present application has the following advantages and benefits:

[0042] 1.The present application aims to develop an automatic control method for the docking and fastening of the tightening tool and the screw on the assembly line of a certain part of a certain aerospace engine.

[0043] 2.The application of the present application can reduce the use of personnel in the assembly process of a certain type of aerospace engine, and improve the quality of the product. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 It is a top view schematic diagram of the part.

[0045] Figure 2 It is a schematic diagram of the hardware platform structure.

[0046] Figure 3 It is a schematic diagram of the calibration plate for setting the reference position of the tightening tool.

[0047] Figure 4 It is a schematic diagram of the tool for checking the reference position of the tightening tool.

[0048] Figure 5 It is a schematic diagram for judging the angle of the hexagonal screw. DETAILED DESCRIPTION

[0049] The present application will be further described in detail below in combination with the drawings and examples.

[0050] The device of the present application comprises a collaborative robot for adjusting the posture positioning, an industrial camera for detecting the position of the screw, an electric tightening machine capable of detecting the tightening torque and angle, an opening hexagonal split head for docking with the edge of the screw, a calibration plate for setting the reference position of the tightening tool, a tool for detecting the reference position of the tightening tool, and a laser sensor and a reflecting plate for circumferential reference positioning.

[0051] The robot is used for adjusting the posture positioning of the tightening tool; a six-axis collaborative robot is adopted, which is the prior art;

[0052] Electric tightening machine, as shown in the attached Figure 1 figure, is fixed with the end of the sixth axis of the robot through the connecting mechanism, for detecting and fixing the target torque fastening screw within a certain range of angles;

[0053] Industrial camera, fixed on one side of the electric tightening machine through the connecting mechanism, for detecting the position and angle of the screw;

[0054] Split head, as an open hexagonal tightening split head, is inserted into the end of the electric tightening machine through the pin, for the cap interface with the edge of the screw;

[0055] Calibration plate, composed of two plates with a 90-degree angle, is fixed on the assembly operation table vertically to the assembly plane, for setting the reference position of the tightening tool;

[0056] Detection tool, as shown in the attached Figure 4 figure. The upper part is a cuboid with a rectangular end face, the length is the distance between two opposite sides of the hexagonal screw, and the width is the length of one side of the hexagonal screw. When rotated to the corresponding angle, it can be embedded into the hexagonal split head. The middle part is a circular truncated cone for connecting the upper end and the base, and the lower end is a cylindrical base for stable placement in the matching groove of the assembly table. When verifying the reference position of the tightening tool, first manually place it on the assembly table so that the tool axis is perpendicular to the assembly plane, then move the tightening tool that has completed the preliminary reference position calibration by the robot through the X, Y, Z axes of the Cartesian coordinate system to the periphery of the verification tool, so that the verification tool can be embedded into the tightening tool split head (this process can rotate the detection tool), then make the robot vertically lift the tightening tool in the Z+ direction, finally rotate the end of the sixth axis of the robot to ±60°, ±120° and ±180° directions and observe whether the verification tool can still be embedded into the tightening split head at these six positions (this process cannot rotate the detection tool). If it can be basically embedded, it is considered that the tightening tool axis is perpendicular to the assembly plane, i.e. the reference position verification is qualified;

[0057] Laser sensor, fixed on the bracket of the electric tightening machine, and the reflector fixed on the tightening tool, are used for setting the circumferential reference position of the tightening tool.

[0058] A control method for automatic cap recognition and tightening of hexagonal screws, comprising the following steps:

[0059] 1) Use the standard checkerboard calibration plate (existing product) to complete the calibration of the camera intrinsic and extrinsic parameters.

[0060] 2) Set the camera coordinate system as C and the robot end execution coordinate system as E, and solve the transformation matrix between the two coordinate systems

[0061] 3) Through the L-shaped calibration plate (as shown in the attached Figure 3(As shown) Initially calibrate the radial reference pose of the tightening tool;

[0062] 4) Use tooling to preliminarily calibrate whether the radial reference position of the tightening tool is qualified;

[0063] 5) Determine the circumferential reference position of the tightening tool using a laser sensor and a reflector;

[0064] 6) Detect the edge features of the reference screw using an industrial camera and set its angle α and the position (x,y) of its center;

[0065] 7) By teaching the robot, the reference screw is embedded in the wedge of the tightening tool after the radial and circumferential reference positions have been calibrated. The robot's current pose is saved and set as the basic pose for the robot to tighten the screw.

[0066] 8) Detect the edge features of any hexagonal screw using an industrial camera and determine the angle α of the edge features. ′ and the location of its center (x) ′ ,′);

[0067] 9) Calculate the angular difference Δα and position offset (Δx, Δy) between any screw and the reference screw in the robot base coordinate system;

[0068] 10) Based on the position and angle deviation between the current screw and the reference screw, the robot adjusts the posture of the tightening tool and completes the cap docking.

[0069] 11) The electric tightening machine drives the tightening tool to start tightening the screws. The tightening process is judged to be completed by detecting the tightening torque and angle.

[0070] In step 1, the method for calibrating the camera's intrinsic and extrinsic parameters is as follows: a standard checkerboard calibration plate is placed on the assembly plane, and a mobile robot drives the camera to take clear images of the calibration plate from three different positions and angles. Each time, the position information of four sets of corner points on the calibration plate plane is acquired.

[0071] In step 2, the transformation matrix between the camera coordinate system and the robot end effector coordinate system is calculated. The method is as follows: calculate the camera intrinsic and extrinsic parameter matrices of the calibration board obtained by shooting clear images at three different positions and angles in step 1, and record the pose parameters of the robot end effector coordinate system (i.e., TCP0) for obtaining these three images.

[0072] In step 3, the fixture plate for calibrating the reference position of the tightening tool is as follows: Figure 3 As shown.

[0073] In step 3, the process of setting the radial reference position of the tightening tool by the calibration plate is as follows: first, the robot drives the tightening tool to slowly approach the 90-degree corner from the open side of the L-shaped calibration plate, so that the side of the tightening tool is tangent to the two ends of the calibration plate, and the lower end of the tool contacts the assembly plane, then the wrench head is rotated and it is observed whether the axis of the wrench head is parallel to the calibration plate. If the parallelism is good, the radial reference position of the tightening tool is preliminarily calibrated, otherwise the position of the wrench head is adjusted by the robot.

[0074] In step 4, the basic form of the detection tool is as shown in Figure 4 , which is placed in the groove that matches the base of the assembly plane, so that the axis of the conical tool can be perpendicular to the assembly plane.

[0075] In step 4, the method for detecting whether the axis of the tightening tool in the reference position is perpendicular to the tightening plane by the tool is as follows: first, the tool is placed vertically on the assembly table so that its axis is perpendicular to the assembly plane, then the robot drives the tightening tool that has completed the preliminary reference position calibration to move near the detection tool through the X, Y, Z axes of the Cartesian coordinate system, so that the detection tool can be embedded in the wrench head (the detection tool can be rotated at the same time), then the robot vertically lifts the tightening tool in the Z+ direction, finally the end of the sixth axis of the robot is rotated to ±60°, ±120° and ±180° directions respectively and it is observed whether the detection tool can still be embedded in the wrench head at these six positions (the detection tool cannot be rotated). If it can be embedded at all positions, it is considered that the axis of the tightening tool is perpendicular to the assembly plane, i.e. the reference position is qualified. Otherwise, the operation of step 2 is re-executed, as shown in Figure 4 .

[0076] In step 5, the method for setting the circumferential reference position of the tightening tool by the laser sensor and the reflector plate is as follows: first, the tightening tool is quickly rotated clockwise by the tightening machine, when the laser sensor detects the light beam reflected by the reflector plate, the rotation is immediately stopped, then the tightening tool is slowly rotated counterclockwise, when the light beam emitted by the laser sensor just escapes from the reflector plate, the rotation is immediately stopped, at this time the position is the circumferential reference position of the tightening tool.

[0077] In step 6, the setting rule of the reference screw is as follows: first, the screw is manually rotated to be in the seated state and it is confirmed that the axis of the screw is basically perpendicular to the assembly plane, then the angle of the screw is adjusted so that one of the edges is tangent to the cylindrical edge of the center protrusion of the workpiece (such as the positions of screws 1 or 2 in Figure 1 ), then the screw is counterclockwise rotated by about 15 degrees (such as the position of screw 3 in Figure 1 ), finally it is confirmed that the screw just does not interfere with the edge of the center protrusion of the part after the screw is butt-jointed with the wrench head, and the screw is set as the reference screw. Among them, the workpiece is as shown in Figure 1As shown in the top view, it includes a body and a protruding part extending upward along the axis of the body, which is a cylinder; the body can be a cylinder, a cuboid or other cubic structures; the central protruding part is like Figure 1 As shown in the top view of the workpiece (A part), the upper surface of the three screws 1, 2, 3 is higher than the center (A part), and one side of the three screws is close to the edge of the central protruding part.

[0078] In step 6, the angle a of the edge feature of the reference screw and the position (x, y) of the center are obtained by the industrial camera, and the method is as follows: the robot is positioned to the industrial camera to make the reference screw image in the center of the camera field of view, and the robot pose is fine-tuned to make the camera detection plane and the assembly plane in a substantially parallel state, the robot pose data of the camera positioning at this time is recorded and saved, then the screw edge feature is identified, and finally the angle a of the edge and the center position (x, y) in the camera coordinate system are calculated.

[0079] In step 7, the teaching robot is moved in the X, Y, Z axis direction of the Cartesian coordinate system and cooperates with the Z direction rotation freedom of the TCP0 at the end of the robot to make the reference screw embedded into the head of the tightening tool for radial and circumferential reference position calibration, the robot pose at this time is saved and set as the basic pose of the robot for screw tightening.

[0080] In step 9, as shown in the figure, Figure 5 In the actual production process, the screw pose detected by the vision system is calculated to obtain the position offset (Δx, Δy) and angle difference Δa of the screw from the reference screw, wherein

[0081]

[0082] In steps 6 and 9, the screw position and angle are all data in the camera coordinate system, and through the conversion relationship solved in step 2 and the robot pose data of positioning the camera in step 6, the data in the robot base coordinate system can be finally calculated.

[0083] In step 10, according to the angle deviation Δa and the position offset (Δx, Δy) of the current screw from the edge of the reference screw, the robot adjusts the pose of the tightening tool according to the reference pose point set in step 7 and completes the cap docking of the screw.

[0084] In step 11, in order to prevent interference between the tool and the parts during tightening, the tightening program in the electric motor tightening machine needs to set the upper limit of the angle of each execution of screwing according to the actual screw and the reference screw angle deviation in addition to the target torque of the set tightening. When the target torque of the tightening is reached, the tightening is completed; when the upper limit of the screwing angle is reached but the target torque is not reached, the robot returns to the original position and then executes steps 3, 5, 8, 9, 10 and 11 in sequence again to continue the automatic cap recognition and tightening process.

[0085] The system builds a hardware platform as shown in Figure 2 An electric tightening gun, a monocular industrial camera (including a lens and a light source, etc.) and a laser sensor (including a reflector) are fixed on the end effector of the collaborative robot. The camera, the tightening shaft, the laser sensor and the relative position of the 6th axis end of the robot are fixed through the corresponding mechanism, so that the robot can position the camera to identify the screw pose and adjust the tightening tool pose to realize automatic cap recognition and tightening. The reflector of the laser sensor is fixed on the side of the tightening tool, and the height is consistent with the laser sensor. In the design, the tightening shaft axis and the Z axis of the TCP0 of the end effector of the robot are also required to coincide, so that the circumferential pose adjustment of the tightening tool can be realized by rotating the 6th axis of the robot.

[0086] The method comprises the following steps:

[0087] 1) Set the world coordinate system W on the chessboard calibration plate plane, the camera coordinate system as C, and the camera pixel coordinate system as L, The transformation relationship matrix between the world coordinate system on the calibration plate and the camera pixel coordinate system Wherein is the intrinsic matrix of the camera, is the extrinsic matrix of the camera, and the solving method is:

[0088] A certain feature point M = [X, Y, Z] on the calibration plate T is mapped to m = [u, v] in the pixel coordinate system Y Since the Z coordinate of the world coordinate system on the calibration plate plane is 0, the following equation can be obtained:

[0089]

[0090] Wherein, H is the combined matrix of the intrinsic and extrinsic parameters, is the homography matrix from the world coordinate system plane of the calibration plate to the pixel plane of the camera, and can be solved by mapping any four corner points of the calibration plate and the corresponding camera pixel coordinate system. r1 and r2 are column vectors of the camera extrinsic matrix R and are orthogonal to each other, representing the coordinate system rotation transformation. t is also a column vector of the camera extrinsic matrix R, representing the coordinate system translation transformation.

[0091] The mobile robot drives the camera to take clear images of the calibration board from three different position angles to form three homography matrices H1, H2 and H3, and the internal parameter matrix and the external parameter matrix are solved by matrix decomposition

[0092] 2) Let C and D be the transformation matrix between the two camera coordinate systems and the transformation matrix between the robot end execution coordinate systems when the camera internal and external parameters are solved at different positions to take images of the calibration board, is the transformation matrix between the camera coordinate system and the robot end execution coordinate system, then The solving method is as follows:

[0093] The camera internal and external parameter matrices in step 1 are solved by taking clear images of the calibration board from three different positions, and the external parameters of the three different position cameras are A1, A2 and A3, respectively, so the transformation matrices between the first and second camera coordinate systems and the second and third camera coordinate systems are

[0094]

[0095]

[0096] The pose parameter matrices of the robot end coordinate system (i.e. TCP0) of the three images are obtained and recorded, and the transformation matrices between the first and second robot end execution coordinate systems and the second and third robot end execution coordinate systems are

[0097]

[0098]

[0099] The above is substituted into to solve

[0100] 3) The process of preliminarily calibrating the radial reference pose of the tightening tool through the L-shaped calibration board is as follows:

[0101] (1) The robot drives the tightening tool to slowly approach the 90-degree corner from the open side of the L-shaped calibration board, so that the side of the tightening tool is tangent to the two ends of the calibration board, and the lower end of the tool contacts the assembly plane, as shown in Figure 3

[0102] (2) Twist the wedge and observe whether the wedge axis is parallel to the calibration board. If the parallelism is good, the radial reference pose of the tightening tool is preliminarily calibrated, otherwise the pose of the wedge is adjusted by the robot.

[0103] 4) The specific process of detecting whether the axis of the tightening tool in the preliminary adjustment reference pose is perpendicular to the tightening plane through the tool is as follows:​​

[0104] (1) As shown in the attached Figure 4 , first place the conical tool vertically on the assembly table, with its axis perpendicular to the assembly plane;

[0105] (2) Let the robot drive the tightening tool that has completed the preliminary reference position calibration to move near the detection tool through the X, Y, Z axes of the Cartesian coordinate system, so that the verification tool can just be embedded in the tightening tool split head (the detection tool can be rotated at the same time);

[0106] (3) Let the robot vertically lift the tightening tool in the Z+ direction, and finally rotate the end of the sixth axis of the robot to ±60°, ±120° and ±180° directions respectively and observe whether the verification tool can still be embedded in the tightening split head at these 6 positions (the detection tool cannot be rotated);

[0107] (4) If it can be embedded, it is considered that the axis of the tightening tool is perpendicular to the assembly plane, that is, the reference position verification is qualified, otherwise the second step operation is re-executed.

[0108] 5) The method for setting the circumferential reference position of the tightening tool through the laser sensor and the reflector plate is: first rotate the tightening tool clockwise quickly by the tightening machine, stop rotating immediately when the laser sensor detects the light beam reflected by the reflector plate, then rotate the tightening tool counterclockwise slowly, stop rotating immediately when the light beam emitted by the laser sensor just escapes from the reflector plate, at this time the position is the circumferential reference position of the tightening tool.

[0109] 6) Obtain the angle α of the edge feature of the reference screw and the position (x, y) of the center through the industrial camera. Here, the setting method of the reference screw is:

[0110] (1) Manually screw the screw onto the workpiece to make it in the seated state and confirm that the screw axis is basically perpendicular to the assembly plane, then adjust the angle of the screw so that one of its edges is tangent to the edge of the cylindrical protrusion in the center of the workpiece (such as the positions of screws 1 or 2 in the attached Figure 1 );

[0111] (2) Rotate the screw counterclockwise by about 15 degrees (such as the position of screw 3 in the attached Figure 1 );

[0112] (3) Finally confirm that the screw just does not interfere with the edge of the protruding part in the center after the screw is docked with the tightening split head, and set it as the state of the reference screw.

[0113] In addition, the method for obtaining the angle of the edge feature of the reference screw and the position of the center is:

[0114] (1) Position the industrial camera by the robot so that the image of the reference screw is in the center of the camera's field of view;

[0115] (2) Fine-tune the robot pose to make the camera detection plane and the assembly plane substantially parallel, record and save the pose data of the robot positioning the camera at this time;

[0116] (3) Identify the screw edge features, and finally calculate the angle a of the edge and the coordinates (x, y) of the center position in the camera coordinate system.

[0117] 7) Move the teaching robot in the X, Y, Z axis direction of the Cartesian coordinate system and cooperate with the Z direction rotation freedom of the TCP0 at the end of the robot to make the reference screw embedded into the head of the tightening tool for radial and circumferential reference position calibration, save the pose of the robot at this time and set it as the basic pose of the robot tightening the screw.

[0118] 8) Detect the edge features of any hexagonal screw in production through an industrial camera and determine the angle a of the edge features and the position (x ′ ,′) of the center. ′

[0119] 9) According to the pose of any screw detected by the industrial camera in production, calculate the angle deviation Δa and the position offset (Δx, Δy) of the reference screw, wherein

[0120]

[0121] Let the base coordinate system of the robot be B, the camera coordinate system be C, and the end effector coordinate system of the robot be E. Convert the position offset of the screw from the data in the camera coordinate system to the data in the robot base coordinate system:

[0122]

[0123] Where Δx b , Δy b represent the offset of the screw center position in the robot base coordinate system, and Δx c , Δy c represent the offset of the screw center position in the camera coordinate system. is the conversion relationship between the camera coordinate system and the end effector coordinate system of the robot, which has been calculated in step 2. is the conversion relationship between the end effector of the robot and the robot base coordinate system, that is, the pose data of the robot positioning the camera to detect the reference screw recorded and saved in step 6 (which can be obtained from the robot system).

[0124] 10) According to the angle deviation Δa and the position offset (Δx, Δy) of the current screw and the reference screw in the robot base coordinate system, the robot adjusts the pose of the tightening tool for reference position calibration to make the head of the tightening tool complete the docking and cap recognition with the screw. ​

[0125] 11) The electric screwdriver drives the tightening tool to start tightening the screw. In order to prevent the tool from interfering with the parts during the tightening process, in addition to setting the target torque for the tightening, the angle upper limit for each execution of the screwing needs to be set according to the actual screw and the reference screw angle deviation. During the tightening process, when the target torque for tightening is reached, the tightening is completed; when the screwing angle upper limit is reached but the target torque is not reached, the robot returns to the original position and then executes steps 3, 5, 8, 9, 10 and 11 in turn, to continue the automatic cap recognition and tightening process. The angle deviation Δα of the screw and the screwing upper limit parameter β in the tightening program are set as follows: β = 60° - Δα + 15°, 0 ≤ Δα < 60°.

Claims

1. A control method for automatic cap tightening of a hexagonal screw of a component, characterized by, The method comprises the following steps: 1) camera intrinsic and extrinsic parameters are calibrated by using a chessboard calibration plate; 2) Set the camera coordinate system as , the robot end execution coordinate system , and solve the transformation matrix between the two coordinate systems ; 3) the radial reference position of the tightening tool is preliminarily calibrated by using an L-shaped calibration plate; 4) whether the radial reference position of the tightening tool is qualified is detected by using a tool, and when the radial reference position is qualified, the next step is performed; 5) the circumferential reference position of the tightening tool is determined by using a laser sensor and a reflective plate; 6) Detect the edge feature of the fiducial screw by the industrial camera and set the edge angle of any two adjacent edges , the position of the center of the fiducial screw ; 7) the reference screw is embedded into the split head of the tightening tool by using a teaching robot, the position of the robot at this time is saved, and the position is set as the basic position of the robot for tightening the screw; 8) Detect any screw edge feature by industrial camera and determine edge feature angle and position of the reference screw center ; 9) Calculate the angle difference between any screw and the reference screw position in the robot base coordinate system and the positional offset ; 10) according to the position offset and the angle difference between the arbitrary screw and the reference screw, the posture of the tightening tool is adjusted, and the screw is recognized and docked; 11) the tightening tool is driven by an electric tightening machine to start tightening the screw, and whether the tightening process is completed is judged by using the rotation torque and the angle detection constraint condition. Step 3) The radial reference position of the tightening tool is preliminarily set by using a calibration plate, and the process is as follows: 3.1) first, the robot drives the tightening tool to enter the inside of the 90-degree corner from the open side of the L-shaped calibration plate at a speed lower than the set speed, so that the side surface of the tightening tool is tangent to the two ends of the calibration plate, and the lower end of the tool contacts the assembly plane; 3.2) then, the split head is rotated, and whether the split head axis is parallel to the calibration plate is observed; if it is considered that the parallelism meets the set standard, the preliminary calibration of the radial reference position of the tightening tool is completed; otherwise, the position of the split head is adjusted by using the robot, and step 3.1) is returned to; Step 4) is specifically: whether the axis of the tightening tool in the reference position is perpendicular to the tightening plane is detected by using a tool, comprising the following steps: First, the tool is vertically placed on the assembly table, so that the axis is perpendicular to the assembly plane, and then the robot drives the tightening tool in the preliminary reference position to move to the vicinity of the detection tool through the X, Y and Z axes of the Cartesian coordinate system, so that the tool can be embedded into the split head of the tightening tool; Then let the robot... Lift the tightening tool vertically in the + direction, and finally rotate the end of the robot's sixth axis to ±60°, ±120° and ±180° respectively with this position as the circumferential reference. Observe whether the tool can be inserted into the tightening wedge in these 6 positions without rotating the tool. If all can be embedded, it is considered that the axis of the tightening tool is perpendicular to the assembly plane, that is, the reference position is qualified; otherwise, the operation of step 3) is re-executed; Step 5) is specifically: first, the tightening tool is rotated clockwise by using the tightening machine, and when the laser sensor detects the light beam reflected by the reflective plate, the rotation is immediately stopped; then, the tightening tool is rotated counterclockwise at a speed lower than the set speed, and when the light beam emitted by the laser sensor is just out of the reflective plate, the rotation is immediately stopped; at this time, the position is the reference position of the tightening tool in the circumferential direction; In step 6), the reference screw setting rule is: first, the screw is manually rotated to be in the seated state and the screw axis is confirmed to be perpendicular to the assembly plane; then, the angle of the screw is adjusted so that one edge is tangent to the cylindrical edge of the center protrusion of the workpiece; then, the screw is rotated counterclockwise to set the angle; finally, after confirming that the split head of the tightening tool is just not interfered with the edge of the center protrusion of the workpiece after being docked with the screw, the screw is set as the reference screw state.

2. The control method of automatic cap tightening of a hexagonal screw of a component according to claim 1, characterized in that, Step 1) is: the chessboard calibration plate is placed on the assembly plane, the camera is moved by the robot to capture images of the calibration plate from three different positions and angles, and the position information of four groups of corner points on the plane of the chessboard calibration plate is obtained each time.

3. The control method of automatic cap tightening of a hexagonal screw of a component according to claim 1, characterized in that, In step 2), the camera coordinate system and the robot end execution coordinate system transformation matrix is solved The method is as follows: the camera internal and external parameter matrix of the chessboard calibration plate image taken at 3 different position angles in step 1) is solved, and the pose parameters of the robot end coordinate system TCP0 of the 3 images are obtained and recorded.

4. The control method of automatic cap tightening of a hexagonal screw of a component according to claim 1, characterized in that, In step 6), the edge angle formed by the edge feature of the reference screw and any two adjacent edges is detected by the industrial camera The position of the center of the reference screw Specifically as follows: The robot positions the industrial camera to image the fiducial screw in the center of the camera's field of view, adjusts the robot pose to bring the camera detection plane parallel to the assembly plane, records and saves the robot pose data for this camera position, identifies the edge features of the screw, and finally calculates the angle of the edge and the center position of the screw in the camera coordinate system has coordinates .

5. The control method of automatic cap tightening of a hexagonal screw of a component according to claim 1, characterized in that, In step 7), the specific process is as follows: The robot is taught to move in the X, Y, Z axis direction of the Cartesian coordinate system and cooperate with the Z direction rotation freedom of the TCP0 at the end of the robot to make the reference screw embedded into the head of the tightening tool, save the pose of the robot at this time and set the pose as the basic pose of the robot for tightening the screw.

6. The control method of automatic cap tightening of a hexagonal screw of a component according to claim 1, characterized in that, In step 11), the following is specifically as follows: In order to prevent interference between the tool and the workpiece during tightening, the tightening program in the electric motor tightening machine sets the upper limit of the angle of each execution of screwing according to the actual angle deviation of the screw from the reference screw in addition to the target torque of the tightening; when the target torque of the tightening is reached, the tightening is completed; when the upper limit of the screwing angle is reached but the target torque is not reached, the robot returns to the initial position and then executes steps 3), 5), 8), 9), 10) and 11) in sequence again to continue the automatic cap recognition and tightening process.

7. A control system for automatic cap- tightening of a hexagonal screw of a component, the system being used to implement a control method for automatic cap- tightening of a hexagonal screw of a component according to any one of claims 1-6, characterized in that, Comprise: A robot for adjusting the pose of a tightening tool; An electric tightening machine fixed at the end of the sixth axis of the robot for tightening the screw under the condition of setting the target torque and angle limit; An industrial camera fixed on one side of the electric tightening machine for detecting the position and angle of the screw; A head, which is an open hexagonal tightening head, is inserted into the end of the electric tightening machine for cap recognition and docking with the edge of the screw; A calibration plate with an L-shaped cross section is composed of two plates with a 90-degree angle and is fixed vertically to the assembly operation table for setting the radial reference of the tightening tool during calibration; A detection tool is composed of an upper part, a middle part and a lower end connected in sequence; the upper part is a cuboid with a rectangular end face, the length of the rectangle is the distance between two opposite sides of the hexagonal screw, and the width is the length of one side of the hexagonal screw; when the detection tool is rotated to a certain angle, the upper part can be embedded into the hexagonal head; the middle part is a circular table for connecting the upper part and the lower end; the lower end is a cylindrical base for placing in the matching groove on the assembly table; A laser sensor is fixed on the electric tightening machine through a bracket, and a reflector is fixed on the tightening tool, which are used for setting the circumferential reference position of the tightening tool.

Images