Resistance strain gauge pasting robot, image processing method and glue injection amount control method

Abstract

The invention relates to the technical field of resistance strain gauges, in particular to a resistance strain gauge pasting robot, an image processing method and a glue injection amount control method. An illuminating lamp and a camera are arranged on the front side surface of the body; four mounting blocks are arranged behind the body, a mounting bottom plate is arranged below the mounting blocks, the position of the mounting bottom plate is adjusted by a position adjusting mechanism, and the position adjusting mechanism comprises a translation assembly and a lifting assembly; the four mounting bottom plates are respectively provided with a coarse grinding mechanism, a fine grinding mechanism, a glue injection mechanism and a chip mounting mechanism; the four mounting bottom plates are respectively a coarse grinding bottom plate, a fine grinding bottom plate, a glue injection bottom plate and a patch bottom plate; according to the invention, resistance strain gauges in stress tests of large-scale metal structures such as cranes and the like can be pasted.

Description

Technical Field

[0001] The invention relates to the technical field of resistance strain gauges, in particular to a resistance strain gauge pasting robot, an image processing method and a glue injection amount control method. Background Art

[0002] Cranes are important handling equipment widely used in industrial and civil fields. Their safety and reliability are directly related to the safety of people's lives and property. In order to ensure the normal operation of the crane, it is necessary to conduct stress tests on it regularly to detect whether there is local stress concentration in the metal structure of the crane and whether the structural strength meets the design requirements. At present, the main method of stress testing is to use resistance strain gauges (hereinafter referred to as strain gauges) to measure the strain and stress of the crane structure. Strain gauges are sensors used to measure the deformation and stress of structures. Their principle is to use the resistance strain effect, that is, when the material is deformed by external force, its resistance value will also change. By measuring the change in resistance value, the strain and stress of the material can be calculated. However, the pasting process of strain gauges is a very tedious and time-consuming task, which requires manual rough grinding, fine grinding, glue injection, patching and other operations, and requires the operator to have a certain level of technical skills and experience. More importantly, due to the complexity and diversity of crane structures, during the stress testing of cranes, some strain gauges need to be pasted to locations that are inaccessible to personnel or locations with extremely high safety risks. This poses great difficulties for crane structural stress testing. Therefore, it is necessary to develop a patch robot that can realize inter-plane flipping on the crane box beam structure. Summary of the invention

[0003] The purpose of the present invention is to provide a resistance strain gauge pasting robot, an image processing method and a glue injection amount control method, which can paste the resistance strain gauges used in stress testing of large metal structures such as cranes.

[0004] The technical solution of the present invention:

[0005] The resistance strain gauge pasting robot comprises a main body and a walking mechanism arranged at the bottom of the main body; the front side of the main body is provided with a lighting lamp and a camera; four mounting blocks are arranged at the rear of the main body, a mounting base plate is arranged below the mounting blocks, and the position of the mounting base plate is adjusted by a position adjustment mechanism, and the position adjustment mechanism comprises a translation component and a lifting component;

[0006] The four installation base plates are respectively provided with a rough grinding mechanism, a fine grinding mechanism, a glue injection mechanism and a patch mechanism; the four installation base plates are respectively a rough grinding base plate, a fine grinding base plate, a glue injection base plate and a patch base plate;

[0007] The rough grinding mechanism comprises a rough grinding bottom plate, a rough grinding motor is fixed to the rear end of the rough grinding bottom plate, a drill chuck is fixed to the output shaft of the rough grinding motor, and a rough grinding head is clamped and fixed on the drill chuck;

[0008] The fine grinding mechanism comprises a fine grinding base plate, a fine grinding motor is fixed to the rear end of the fine grinding base plate, and a fine grinding head is fixed to the output shaft of the fine grinding motor;

[0009] The glue injection mechanism comprises a glue injection bottom plate, a first electric push rod (hot air pipe) is fixed to the top of the rear end of the glue injection bottom plate, and a glue injection tube is fixed to the bottom of the rear end, and a push rod of the first electric push rod penetrates the glue injection bottom plate and is fixedly connected to the glue injection rod in the glue injection tube;

[0010] The patch mechanism comprises a patch base plate, a second electric push rod is fixed to the top of the rear end of the patch base plate, a patch syringe is fixed to the bottom of the rear end, a patch plug is arranged in the patch syringe, a push rod of the second electric push rod penetrates the patch base plate and is fixedly connected to the patch plug; the lower end of the inner wall of the patch syringe has a placement groove for placing the patch.

[0011] Further, the translation assembly includes a first motor fixed in the body, a first driving gear is fixed on the output shaft of the first motor, the first driving gear is meshed with a first driven gear, a screw is fixed on the first driven gear, the front end of the screw is fixed to the inner bottom of the body through a bearing seat, and the rear end is threadedly connected to the right end of a slider; the left end of the slider is sleeved on the first guide rod, the rear end of the first guide rod is fixed to the rear inner wall of the body, and the front end is fixed to a mounting box;

[0012] The slide block is provided with four oblique grooves, in which positioning pins are passed, and the lower ends of the positioning pins are fixedly connected to the mounting blocks; the four mounting blocks are sleeved on two second guide rods, and the left and right ends of the second guide rods are respectively fixedly connected to the left and right inner walls of the body.

[0013] Furthermore, the lifting assembly includes a rack arranged in the installation box, the rack meshes with the second driving gear, and the second driving gear is driven by the second motor; a guide block is fixed to the front and rear sides of the rack, the guide block is sleeved on the third guide rod, the left end of the third guide rod is fixed to the left inner wall of the installation box, and the right end is inserted into a sliding box;

[0014] The right side of the rack is fixedly connected to the left side of the sliding box via a connecting column; an L-shaped plate is arranged inside the sliding box, and sliding grooves are provided at the top and bottom of the sliding box, and the upper and lower ends of the vertical plate of the L-shaped plate pass through the upper and lower sliding grooves respectively, and the front end of the mounting base plate has a bending plate for the horizontal plate of the L-shaped plate to press down;

[0015] A guide column is fixed on the L-shaped plate in the sliding box, the front and rear inner walls of the sliding box are provided with S-shaped grooves, the front and rear inner walls of the installation box are provided with vertical grooves, and the front and rear ends of the guide column pass through the S-shaped groove and the vertical groove.

[0016] Furthermore, a movable column is arranged above the mounting block, a first upper baffle is fixed to the upper end of the movable column, and a first lower baffle is fixed to the lower end which penetrates the mounting block and the mounting base plate in sequence; a first spring is provided on the movable column.

[0017] Furthermore, the walking mechanism includes a rear bracket fixed to the rear end of the bottom of the body, and rear wheel guards are fixed to the left and right sides of the rear bracket. The inner side of the rear wheel guard is provided with a rear wheel, and the rear wheel is driven by a first reduction motor;

[0018] The front end of the bottom of the body is connected to a front bracket via a connecting frame, and front wheel guards are fixed to the left and right sides of the front bracket. A front wheel is arranged on the inner side of the front wheel guard, and the front wheel is driven by a second reduction motor.

[0019] Furthermore, telescopic columns are fixed to the four corners of the top of the front bracket, the upper ends of the telescopic columns penetrate the bottom of the main body and are fixed to a second upper baffle; a second spring is provided on the telescopic column; and the connecting frame is made of an elastomeric material.

[0020] Furthermore, the front wheels and the rear wheels are both magnetic wheels; the ratio of the chassis height to the wheelbase of the robot is not less than 5:6, so that the robot can successfully climb over a right-angled surface.

[0021] Furthermore, a magnetic suction cup is provided in the lower middle part of the rear bracket, and the magnetic suction cup is driven by a third electric push rod.

[0022] The image processing method of the resistance strain gauge pasting robot collects images through the camera. The method is as follows:

[0023] (1) Image cropping and compression: The image resolution is 188 columns × 120 rows. To improve processing speed and forward-looking vision, only the first 60 rows of images are captured. At the same time, each row of image data is subjected to non-uniform compression processing, and finally a 120 × 60 image is generated as the path information map.

[0024] (2) Dynamic threshold image segmentation: The image is divided into upper and lower parts, the average threshold is calculated for each part, and the threshold is adjusted according to formula (1);

[0025]

[0026] Among them, DTMOS is the global threshold calculated by the OTSU algorithm, j is the number of image rows, k is the number of shadow area rows, and C is a constant. The image is segmented by the optimal threshold, and a binary image with significant differences can be obtained.

[0027] (3) Route planning: Taking the second row of the image as the reference row, traverse upward row by row to the top row of the image; in each row scan, use the column shrinking technology, that is, scan every other column, to reduce the amount of data processing and improve the processing speed;

[0028] Traverse the bottom edge from top to bottom. When a boundary that meets the conditions is found normally, the continuity and trend change of the upper and lower boundaries in the image are used to optimize the boundary search of the next row. The boundary search range of the next row is determined based on the center line information center_line[i-1] of the previous row, as shown in formula (2).

[0029] line_part[i]=[center_line[i-1]-C1,center_line[i-1]] (2)

[0030] C1 is a parameter that needs to be adjusted according to the actual camera installation position. line_part[i] is the boundary position of the i-th row, that is, the position of the black-white transition point of this row.

[0031] If the pixels in the current row are all white, the boundary range is determined directly according to formula (3);

[0032] line_part[i]=[1,1+C2] (3)

[0033] C2 is a parameter that needs to be adjusted according to the actual camera installation position, that is, the boundary is at the initial position;

[0034] If the pixels in the current row are all black, it is necessary to further determine the pixel value at the midpoint of the previous row. If the midpoint of the previous row is black, the boundary range is given by formula (4).

[0035] line_part[i]=[Width / 2-C3,Width / 2-C4] (4)

[0036] Where width is the width of the cropped image, C3 and C4 are constants determined according to actual conditions;

[0037] If the midpoint of the previous row is white, the boundary range is given by formula (5), and C5 is a constant determined according to the actual situation;

[0038] line_part[i]=[1,Width / 2-C5] (5)

[0039] (4) Fitting processing of boundaries and center lines: Based on the positions of the center line, left line, and right line, the information of each row is stored in an array, as described in formula (6), where center_line[i] represents the column where the center line of the current i-th row is located, top_line[i] represents the column where the top line of the current i-th row is located, and bottom_line[i] represents the column where the bottom line of the current i-th row is located.

[0040]

[0041] At the same time, in the process of recording the centerline position, the forward travel width weight[i] of each row is also recorded, and its calculation method is shown in formula (7);

[0042]

[0043] Based on whether the row weight[i] value is close to 0, we can determine whether there is an obstacle or it is suspended in the air ahead.

[0044] The glue injection amount control method of the resistance strain gauge pasting robot controls the glue injection amount through the genetic PID algorithm. The method is as follows:

[0045] During the operation of the first electric push rod, the internal encoder sensor continuously collects displacement calculation speed, that is, the injection speed, and the PID controller continuously receives real-time data from the encoder sensor and calculates the error between the current injection speed and the target injection speed, that is, the speed error;

[0046] Then, the required current of the first electric linear actuator is calculated according to the PID algorithm, and the signal is transmitted to the controller of the first electric linear actuator through a relay; after receiving the signal, the controller of the first electric linear actuator adjusts its power supply to reduce the speed error; as the speed error gradually decreases, the output of the PID controller will gradually stabilize, and finally the speed calculated by the encoder sensor will be kept near the target speed;

[0047] Output of PID controller

[0048] Among them, u(I) is the instruction for controlling the speed error, that is, the PID controller adjusts the current output to the first electric push rod through the u(I) value, K p , K i , K d They are the gain coefficients of the proportional, integral and differential links respectively. By adjusting K p , K i , K d The value of can change the performance of the PID controller so that the system can achieve the desired speed control effect.

[0049] Beneficial effects of the present invention:

[0050] (1) The walking mechanism can drive the main body to move; the translation assembly can adjust the left and right position of the installation base plate, and the lifting assembly can adjust the up and down position of the installation base plate. The rough grinding mechanism first grinds the patch position, and then the fine grinding mechanism grinds the patch position after rough grinding. Then, the glue injection mechanism injects glue into the patch position, and finally the patch mechanism mounts the patch position after glue injection.

[0051] (2) Significantly improve the efficiency of strain gauge pasting: The robot adopts an automated operation process. Compared with manual pasting, it can work continuously and stably, greatly improving the efficiency of strain gauge pasting. This efficiency improvement is particularly significant in the stress detection of large cranes, which can significantly shorten the detection cycle.

[0052] (3) Significantly reduce labor costs and safety risks: The use of automated robots reduces the reliance on skilled workers for strain gauge pasting, thereby reducing labor costs. At the same time, robots can work in dangerous or difficult-to-reach locations, effectively reducing the safety risks of manual operations.

[0053] The invention can be used to paste resistance strain gauges in stress tests of large metal structures such as cranes. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a three-dimensional schematic diagram of the present invention (front side perspective).

[0055] Figure 2 It is a three-dimensional schematic diagram of the present invention (rear side perspective).

[0056] Figure 3 It is the internal structure diagram of the entity.

[0057] Figure 4 It is a schematic diagram of the structure of the position adjustment mechanism (top view).

[0058] Figure 5 It is a schematic diagram of the structure of the position adjustment mechanism (viewed from above).

[0059] Figure 6 This is the internal structure diagram of the installation box (first perspective looking down).

[0060] Figure 7 This is the internal structure diagram of the installation box (viewed from above).

[0061] Figure 8 This is the internal structure diagram of the installation box (second perspective looking down).

[0062] Fig. 9 This is a diagram of the internal structure of the slide box.

[0063] Fig.10 It is a structural diagram of the walking mechanism.

[0064] Fig.11 It is a three-dimensional schematic diagram of the present invention (looking up).

[0065] Fig.12 This is a calculation diagram of the robot's right-angle flip.

[0066] Fig.13 This is a schematic diagram of the robot's right-angle flip state.

[0067] Fig.14 is a flow chart of the image processing method.

[0068] Fig.15 It is a flow chart of the method for controlling the amount of glue injected.

[0069] In the figure: body 1, walking mechanism 2, lighting lamp 3, camera 4, mounting block 5, mounting base plate 6, translation assembly 7, lifting assembly 8, rough grinding mechanism 9, fine grinding mechanism 10, glue injection mechanism 11, patch mechanism 12, rough grinding motor 91, drill chuck 92, rough grinding head 93, fine grinding motor 101, fine grinding head 102, first electric push rod 111, glue injection tube 112, second electric push rod 121, syringe 122, patch plug 123, placement slot 124, first motor 71, first driving gear 72, first driven gear 73, screw 74, bearing seat 75, slider 76, first guide rod 77, installation box 81, inclined slot 7 8, positioning pin 79, second guide rod 710, rack 82, second driving gear 83, second motor 813, guide block 84, third guide rod 85, sliding box 86, connecting column 88, L-shaped plate 89, slide groove 87, guide column 810, S-shaped groove 811, vertical groove 812, first upper baffle 13, first lower baffle 14, first spring 15, rear bracket 28, rear wheel guard plate 16, rear wheel 17, first reduction motor 18, connecting frame 19, front bracket 20, front wheel guard plate 21, front wheel 22, hot air pipe 113, telescopic column 23, second upper baffle 24, second spring 25, magnetic suction cup 26, third electric push rod 27. DETAILED DESCRIPTION

[0070] The present invention will be further described below in conjunction with the accompanying drawings.

[0071] like Figure 1-Figure 10As shown, the present invention provides a first embodiment of a resistance strain gauge pasting robot, comprising a body 1 and a walking mechanism 2 arranged at the bottom of the body 1, the walking mechanism 2 can drive the body 1 to move; the front side of the body 1 is provided with a lighting lamp 3 and a camera 4, the lighting lamp 3 is used for lighting, and the camera 4 is used for viewing the front field of view. Four mounting blocks 5 are arranged at the rear of the body 1, and a mounting base plate 6 is arranged below the mounting blocks 5. The mounting base plate 6 is adjusted in position by a position adjustment mechanism, and the position adjustment mechanism includes a translation component 7 and a lifting component 8. The translation component 7 can adjust the left and right positions of the mounting base plate 6, and the lifting component 8 can adjust the upper and lower positions of the mounting base plate 6.

[0072] A rough grinding mechanism 9, a fine grinding mechanism 10, a glue injection mechanism 11 and a patch mechanism 12 are respectively installed on the four mounting bases 6; the four mounting bases 6 are respectively a rough grinding base, a fine grinding base, a glue injection base and a patch base; the patch position is first rough-ground by the rough grinding mechanism 9, and then the patch position after rough grinding is fine-ground by the fine grinding mechanism 10, and then the patch position is injected with glue by the glue injection mechanism 11, and finally the patch position after glue injection is patched by the patch mechanism 12.

[0073] The rough grinding mechanism 9 includes a rough grinding base plate, a rough grinding motor 91 is fixed to the rear end of the rough grinding base plate, a drill chuck 92 is fixed to the output shaft of the rough grinding motor 91, the model of the drill chuck 92 may be 1-16B18, but is not limited thereto, a rough grinding head 93 is clamped and fixed on the drill chuck 92, and the rough grinding head 93 may be a T-shaped plane grinding wheel with a diameter of 20 mm and a mesh number of 120#, but is not limited thereto; the rough grinding motor 91 can drive the drill chuck 92 to rotate, thereby driving the rough grinding head 93 to rotate to roughly grind the patch position, and the drill chuck 92 can clamp the rough grinding head 93, so that the rough grinding head 93 can be easily disassembled and assembled.

[0074] The fine grinding mechanism 10 includes a fine grinding base plate, a fine grinding motor 101 is fixed to the rear end of the fine grinding base plate, the fine grinding motor 101 can be a hollow cup motor, a fine grinding head 102 is fixed on the output shaft of the fine grinding motor 101, the fine grinding head 102 can be a T-shaped adhesive grinding head, and the matching sandpaper is 400# mesh, but not limited to this; the fine grinding motor 101 can drive the fine grinding head 102 to rotate, so that the patch position can be finely ground.

[0075] Alternatively, a pneumatic sandblasting gun is fixed to the rear end of the fine grinding base plate. The model of the pneumatic sandblasting gun may be Yanke W71, but is not limited thereto. The pneumatic sandblasting gun uses compressed air to spray abrasives at high speed to the patch position to achieve more thorough rust removal, decontamination and roughening treatment.

[0076] The glue injection mechanism 11 includes a glue injection base plate, a first electric push rod 111 is fixed to the top of the rear end of the glue injection base plate, and a glue injection tube 112 is fixed to the bottom of the rear end. The push rod of the first electric push rod 111 penetrates the glue injection base plate and is fixedly connected to the glue injection rod in the glue injection tube 112; a glue injection needle is installed at the bottom of the glue injection tube 112, and the first electric push rod 111 drives the glue injection rod to move downward to squeeze the glue in the glue injection tube 112, so that the glue is ejected from the glue injection needle to inject glue to the patch position.

[0077] The patch mechanism 12 includes a patch base plate, a second electric push rod 121 is fixed to the top of the rear end of the patch base plate, a patch syringe 122 is fixed to the bottom of the rear end, a patch plug 123 is arranged in the patch syringe 122, the push rod of the second electric push rod 121 penetrates the patch base plate and is fixedly connected to the patch plug 123; the lower end of the inner wall of the patch syringe 122 has a placement groove 124 for placing the patch. The patch plug 123 is driven to move downward by the second electric push rod 121. In the process of the patch plug 123 moving downward, the patch on the placement groove 124 can be squeezed downward to be separated from the placement groove 124, and then the patch is attached to the patch position.

[0078] The translation assembly 7 includes a first motor 71 fixed in the body 1, a first driving gear 72 is fixed on the output shaft of the first motor 71, the first driving gear 72 is meshed with a first driven gear 73, a screw 74 is fixed on the first driven gear 73, the front end of the screw 74 is fixed to the inner bottom of the body 1 through a bearing seat 75, and the rear end is threadedly connected to the right end of a slider 76; the left end of the slider 76 is sleeved on a first guide rod 77, the rear end of the first guide rod 77 is fixed to the rear inner wall of the body 1, and the front end is fixed to a mounting box 81;

[0079] The slider 76 is provided with four oblique grooves 78, and a positioning pin 79 is passed through the oblique groove 78. The lower end of the positioning pin 79 is fixedly connected to the mounting block 5; the four mounting blocks 5 are sleeved on two second guide rods 710, and the left and right ends of the second guide rods 710 are respectively fixedly connected to the left and right inner walls of the main body 1.

[0080] The first motor 71 drives the first driving gear 72 to rotate, thereby driving the first driven gear 73 to rotate, thereby driving the screw rod 74 to rotate, thereby driving the slider 76 to move forward and backward, thereby driving the positioning pin 79 to slide in the inclined groove 78, thereby driving the mounting block 5 to move left and right, thereby driving the mounting base plate 6 to move left and right, and further driving the rough grinding mechanism 9, the fine grinding mechanism 10, the glue injection mechanism 11 and the patch mechanism 12 to move left and right.

[0081] The lifting assembly 8 includes a rack 82 arranged in an installation box 81, and the rack 82 is meshed with a second driving gear 83, and the second driving gear 83 is driven by a second motor 813; a guide block 84 is fixed to the front and rear sides of the rack 82, and the guide block 84 is sleeved on a third guide rod 85, and the left end of the third guide rod 85 is fixed to the left inner wall of the installation box 81, and the right end is inserted into a sliding box 86; the guide block 84 cooperates with the third guide rod 85 to play a role of limiting guidance when the rack 82 moves left and right.

[0082] The right side of the rack 82 is fixedly connected to the left side of the sliding box 86 via a connecting column 88; an L-shaped plate 89 is arranged inside the sliding box 86, and a slide groove 87 is provided at the top and bottom of the sliding box 86. The upper and lower ends of the vertical plate of the L-shaped plate 89 pass through the upper and lower slide grooves 87 respectively, and the front end of the mounting base plate 6 has a bent plate for the horizontal plate of the L-shaped plate 89 to press down.

[0083] A guide column 810 is fixed on the L-shaped plate 89 inside the sliding box 86. The front and rear inner walls of the sliding box 86 are both provided with S-shaped grooves 811. The front and rear inner walls of the installation box 81 are both provided with vertical grooves 812. The front and rear ends of the guide column 810 pass through the S-shaped groove 811 and the vertical groove 812.

[0084] The second motor 813 drives the second driving gear 83 to rotate, thereby driving the rack 82 to move left and right, thereby driving the sliding box 86 to move left and right. When the sliding box 86 moves left and right, due to the action of the vertical slot 812 and the S-shaped slot 811, the guide column 810 moves up and down, thereby driving the L-shaped plate 89 to move up and down. When the L-shaped plate 89 moves downward, the horizontal plate of the L-shaped plate 89 is driven to move downward to press down the bent plate of the installation base plate 6, thereby driving the installation base plate 6 to move downward, and then driving the rough grinding mechanism 9, the fine grinding mechanism 10, the glue injection mechanism 11 and the patch mechanism 12 to move downward.

[0085] A moving column is arranged above the mounting block 5, and a first upper baffle 13 is fixed to the upper end of the moving column, and a first lower baffle 14 is fixed to the lower end which penetrates the mounting block 5 and the mounting base plate 6 in sequence; a first spring 15 is arranged on the moving column. When the horizontal plate of the L-shaped plate 89 moves downward to press the bent plate of the mounting base plate 6, so that the mounting base plate 6 moves downward, the moving column can be driven to move downward at the same time, and the first spring 15 is compressed at this time. When the L-shaped plate 89 moves upward, the compressive elastic force of the first spring 15 drives the moving column to move upward, and then drives the mounting base plate 6 to move upward for reset. The first upper baffle 13 can prevent the moving column from completely detaching from the mounting block 5 when it moves downward, and the first lower baffle 14 can prevent the moving column from completely detaching from the mounting block 5 when it moves upward.

[0086] The walking mechanism 2 includes a rear bracket 28 fixed to the rear end of the bottom of the body 1, and the left and right sides of the rear bracket 28 are fixed with rear wheel guards 16, and the inner side of the rear wheel guard 16 is provided with a rear wheel 17, and the rear wheel 17 is driven by a first reduction motor 18;

[0087] The front end of the bottom of the body 1 is connected to a front bracket 20 via a connecting frame 19, and front wheel guards 21 are fixed to the left and right sides of the front bracket 20. A front wheel 22 is arranged on the inner side of the front wheel guard 21, and the front wheel 22 is driven by a second reduction motor.

[0088] The first reduction motor 18 can drive the rear wheel 17 to rotate, and the second reduction motor can drive the front wheel 22, thereby driving the body 1 to move. Since the two front wheels 22 and the two rear wheels 17 are driven by their respective corresponding reduction motors, when the body 1 needs to turn, it only needs to adjust the speed of the front wheel 22 and the rear wheel 17 on one side to be different from the speed of the front wheel 22 and the rear wheel 17 on the other side, so that the body 1 can turn.

[0089] The working principle of this embodiment is as follows: when the walking mechanism 2 drives the main body 1 to move to the patch position, the rough grinding mechanism 9 is driven to move to the top of the patch position through the translation assembly 7, the L-shaped plate 89 of the lifting assembly 8 is moved to the top of the bending plate of the rough grinding bottom plate of the rough grinding mechanism 9, and then pressed downward, thereby driving the rough grinding bottom plate to move downward, and when the rough grinding bottom plate moves to the patch position, the patch position is rough-ground by the rough grinding mechanism 9, and after the rough grinding is completed, the above steps are repeated, and the patch position is fine-ground, glue injected and patched in turn by the fine-grinding mechanism 10, the glue injection mechanism 11 and the patch mechanism 12.

[0090] On the basis of the first embodiment, the present invention provides a second embodiment of the resistance strain gauge pasting robot, wherein a hot air duct 113 is further provided at the bottom of the glue injection base plate, and an electric heating wire is provided in the hot air duct 113. The hot air duct 113 is connected to an air supply device to blow out hot air, thereby preheating the patch position and the glue, and effectively solving the problem that the glue cannot stick in a low temperature environment.

[0091] On the basis of the second embodiment, the present invention provides a third embodiment of a resistance strain gauge pasting robot, wherein the top four corners of the front bracket 20 are fixed with telescopic columns 23, the upper ends of the telescopic columns 23 penetrate the bottom of the body 1 and are fixed with a second upper baffle 24; the telescopic columns 23 are provided with a second spring 25; the material of the connecting frame 19 is an elastomeric material, which may be TPU material, but is not limited thereto. When the front wheel 22 of the robot encounters a track or an obstacle, the front bracket 20 moves upward, thereby driving the telescopic column 23 to move upward, causing the second spring 25 to compress. After crossing the track or obstacle, the compressive elastic force of the second spring 25 causes the telescopic column 23 and the front bracket 20 to move downward for reset; since the material of the connecting frame 19 is an elastomeric material, it can be deformed when the front bracket 20 moves upward. The second upper baffle 24 can prevent the telescopic column 23 from completely detaching from the body 1 when moving downward.

[0092] Based on the second or third embodiment, the present invention provides a fourth embodiment of a resistance strain gauge pasting robot, wherein the front wheels 22 and the rear wheels 17 are both magnetic wheels, so that the robot can crawl stably on the surface of ferromagnetic materials, which not only improves the robot's movement stability and flexibility, but also allows the robot to operate reliably in an inverted or tilted state.

[0093] like Figure 11-13 As shown, the bottom surface 201 of the front bracket 20 and the bottom surface 281 of the rear bracket 28 are at the same height, so the height of the bottom surface 201 of the front bracket 20 (or the height of the bottom surface 281 of the rear bracket 28) is regarded as the chassis height of the robot, and the ratio of the chassis height to the wheelbase (i.e., the axle center distance of the front wheels and the rear wheels) of the robot is not less than 5:6, so that the robot can successfully climb over a right-angled plane, that is, the robot can successfully climb over from a vertical plane to a horizontal plane, or successfully climb over from a horizontal plane to a vertical plane.

[0094] like Fig.12 As shown, in this embodiment, the height of the front wheel and the rear wheel is the same as the height of the chassis of the robot, so the height of the front wheel or the rear wheel is regarded as the height of the chassis. When climbing at a right angle, the angle between the chassis and the plane is 45°;

[0095] Given that the wheel diameter is 80 mm, the chassis height is: 2*OA=80 mm.

[0096] The wheelbase should be less than: 2*AB=2*tan -1 ∠OBA=2*tan -1 22.5°=96mm,

[0097] 80mm:96mm=5:6;

[0098] Therefore, the chassis to wheelbase ratio is not less than 5:6.

[0099] Based on any one of the second to fourth embodiments, the present invention provides a fifth embodiment of a resistance strain gauge pasting robot, wherein a magnetic suction cup 26 is provided in the lower middle part of the rear bracket 28, and the magnetic suction cup 26 is driven by a third electric push rod 27. The third electric push rod 27 can drive the magnetic suction cup 26 to move up and down, and extend or retract the electromagnet according to work needs. During rough grinding, fine grinding and other operations, the magnetic suction cup 26 extends and adsorbs on the metal structure, and uses magnetic force to offset the reaction force generated during grinding, thereby preventing the robot from falling due to uneven force. After the operation is completed, the magnetic suction cup 26 retracts without affecting the movement and positioning of the robot.

[0100] like Fig.14 As shown, the present invention also provides an image processing method for a resistance strain gauge pasting robot, which collects images through a camera, and the method is as follows:

[0101] (1) Image cropping and compression: The image resolution is 188 columns × 120 rows. To improve processing speed and forward-looking vision, only the first 60 rows of images are captured. At the same time, each row of image data is subjected to non-uniform compression processing, and finally a 120 × 60 image is generated as the path information map.

[0102] (2) Dynamic threshold image segmentation: The image is divided into upper and lower parts, the average threshold is calculated for each part, and the threshold is adjusted according to formula (1);

[0103]

[0104] Among them, DTMOS is the global threshold calculated by the OTSU algorithm, j is the number of image rows, k is the number of shadow area rows, and C is a constant. The image is segmented by the optimal threshold, and a binary image with significant differences can be obtained.

[0105] (3) Route planning: Taking the second row of the image as the reference row, traverse upward row by row to the top row of the image; in each row scan, use the column shrinking technology, that is, scan every other column, to reduce the amount of data processing and improve the processing speed;

[0106] Traverse the bottom edge from top to bottom. When a boundary that meets the conditions is found normally, the continuity and trend change of the upper and lower boundaries in the image are used to optimize the boundary search of the next row. The boundary search range of the next row is determined based on the center line information center_line[i-1] of the previous row, as shown in formula (2).

[0107] line_part[i]=[center_line[i-1]-C1,center_line[i-1]] (2)

[0108] C1 is a parameter that needs to be adjusted according to the actual camera installation position. line_part[i] is the boundary position of the i-th row, that is, the position of the black-white transition point of this row.

[0109] If the pixels in the current row are all white, the boundary range is determined directly according to formula (3);

[0110] line_part[i]=[1,1+C2] (3)

[0111] C2 is a parameter that needs to be adjusted according to the actual camera installation position, that is, the boundary is at the initial position;

[0112] If the pixels in the current row are all black, it is necessary to further determine the pixel value at the midpoint of the previous row. If the midpoint of the previous row is black, the boundary range is given by formula (4).

[0113] line_part[i]=[Width / 2-C3,Width / 2-C4] (4)

[0114] Where width is the width of the cropped image, C3 and C4 are constants determined according to actual conditions;

[0115] If the midpoint of the previous row is white, the boundary range is given by formula (5), and C5 is a constant determined according to the actual situation;

[0116] line_part[i]=[1,Width / 2-C5] (5)

[0117] (4) Fitting processing of boundaries and center lines: Based on the positions of the center line, left line, and right line, the information of each row is stored in an array, as described in formula (6), where center_line[i] represents the column where the center line of the current i-th row is located, top_line[i] represents the column where the top line of the current i-th row is located, and bottom_line[i] represents the column where the bottom line of the current i-th row is located.

[0118]

[0119] At the same time, in the process of recording the centerline position, the forward travel width weight[i] of each row is also recorded, and its calculation method is shown in formula (7);

[0120]

[0121] Based on whether the row weight[i] value is close to 0, we can determine whether there is an obstacle or it is suspended in the air ahead.

[0122] The method is further described below in conjunction with a specific embodiment:

[0123] Vision-based edge and environment detection methods:

[0124] The patch robot is designed with an efficient image processing algorithm specifically for identifying the edges of metal structures to prevent collisions or falls. The algorithm can accurately retain edge information, eliminate redundant data, and simplify images through binarization processing to improve recognition accuracy and efficiency. The camera captures edge information and uses grayscale value differences for accurate recognition. Based on the optimized image, the robot performs edge detection to ensure stable operation in complex environments and effectively avoid the risk of collisions or falls.

[0125] (1) Cropping and compression of original images: In order to solve the problem of huge image data volume, a cropping and compression strategy is adopted to optimize processing efficiency. The resolution of the original grayscale image is 188 columns × 120 rows. In order to improve processing speed and forward-looking vision, only the first 60 rows of images are intercepted. At the same time, each row of image data is subjected to non-uniform compression processing, and finally a 120×60 image is generated as a path information map. This processing method significantly reduces the amount of data, improves the processor processing speed, and enhances predictability, while ensuring that key information is not lost, effectively preventing the robot from hitting the wall or falling off.

[0126] (2) Dynamic threshold segmentation image: In the visual processing of patch robots, the key to binary processing lies in the threshold setting. In order to solve the problem of threshold instability caused by changes in ambient light, the OTSU dynamic threshold segmentation algorithm is adopted. However, after the mechanical structure and camera position are fixed, the lower half of the image is easily affected by the front shadow, resulting in a lower threshold. To solve this problem, through a large number of experiments, an improved method is proposed: the image is divided into upper and lower parts, the average threshold is calculated respectively, and the threshold is adjusted according to formula (1).

[0127]

[0128] Among them, DTMOS is the global threshold calculated by the OTSU algorithm, j is the number of image rows, k is the number of shadow area rows, and C is a constant (this system sets C = 20). Using this method to obtain the optimal threshold for image segmentation, a binary image with significant differences can be obtained, which effectively improves the recognition ability of the patch robot in complex environments.

[0129] (3) Route planning: The second row of the image (counting from bottom to top) is used as the reference row, and the image is traversed row by row to the top row. In each row scan, the column shrinking technology is used, that is, scanning every other column, to reduce the amount of data processing and improve the processing speed.

[0130] For the recognition of the edge of a general route, a basic flow chart as shown in the figure above is proposed, and processing solutions for various situations are designed. Taking the upper edge as an example, traverse the bottom edge from top to bottom. When a boundary that meets the conditions is found normally, case 1 is met. The algorithm will use the continuity and trend change of the upper and lower boundaries in the image to optimize the boundary search of the next row. Specifically, the boundary search range of the next row will be determined based on the center line information of the previous row (center_line[i-1]), as shown in formula (2).

[0131] line_part[i]=[center_line[i-1]-C1,center_line[i-1]] (2)

[0132] C1 is a parameter that needs to be adjusted according to the actual camera installation position. Line_part[i] is the boundary position of the i-th row, that is, the position of the black-white transition point of this row. Center_line is calculated by formula (6), which can be understood as the boundary position of the road recognized by the camera, the middle line of the road, and is calculated as (upper boundary coordinate + lower boundary coordinate) / 2.

[0133] If the pixels in the current row are all white, satisfying condition 2, the boundary range is directly determined according to formula (3).

[0134] line_part[i]=[1,1+C2] (3)

[0135] C2 is a parameter that needs to be adjusted according to the actual camera installation position, that is, the boundary is at the initial position.

[0136] If the pixels in the current row are all black, it is necessary to further determine the pixel value at the midpoint of the previous row. The scan line information may be limited by the shrinking process, so a detailed analysis is required. If the midpoint of the previous row is black, it satisfies case 4, and the boundary range is given by formula (4).

[0137] line_part[i]=[Width / 2-C3,Width / 2-C4] (4)

[0138] Wherein width is the width of the cropped image, C3 and C4 are constants determined according to actual conditions (in this system, C3=1, C4=3).

[0139] If the midpoint of the previous row is white, condition 3 is satisfied, and the boundary range is given by formula (5), and C5 is also a constant determined according to the actual situation (in this system, C5=1).

[0140] line_part[i]=[1,Width / 2-C5] (5)

[0141] (4) Fitting of boundaries and center lines: Based on the positions of the center line, left line, and right line, this information of each row is stored in an array, as described in formula (6), where center_line[i] represents the column where the center line of the current i-th row is located, that is, the position where the black and white jump point first appears from top to bottom in each column of the image; top_line[i] represents the column where the top line of the current i-th row is located, and bottom_line[i] represents the column where the bottom line of the current i-th row is located, that is, the position where the black and white jump point first appears from bottom to top. Black and white jump points are considered to be boundaries.

[0142]

[0143] At the same time, in the process of recording the centerline position, the system also records the forward travel distance width weight[i] of each row, which is calculated as shown in formula (7).

[0144]

[0145] The system determines whether there is an obstacle or a suspended object ahead based on whether the row weight[i] value is close to 0.

[0146] like Fig.15 As shown, the present invention also provides a method for controlling the amount of glue injected by a resistance strain gauge pasting robot, and the amount of glue injected is controlled by a genetic PID algorithm, and the method is as follows:

[0147] During the operation of the first electric push rod, the internal encoder sensor continuously collects displacement calculation speed, that is, the injection speed, and the PID controller continuously receives real-time data from the encoder sensor and calculates the error between the current injection speed and the target injection speed, that is, the speed error;

[0148] Then, the required current of the first electric linear actuator is calculated according to the PID algorithm, and the signal is transmitted to the controller of the first electric linear actuator through a relay; after receiving the signal, the controller of the first electric linear actuator adjusts its power supply to reduce the speed error; as the speed error gradually decreases, the output of the PID controller will gradually stabilize, and finally the speed calculated by the encoder sensor will be kept near the target speed;

[0149] Output of PID controller

[0150] Among them, u(I) is the instruction for controlling the speed error, that is, the PID controller adjusts the current output to the first electric push rod through the u(I) value, K p , K i , K dThey are the gain coefficients of the proportional, integral and differential links respectively. By adjusting K p , K i , K d The value of can change the performance of the PID controller so that the system can achieve the desired speed control effect.

[0151] The method is further described below in conjunction with a specific embodiment:

[0152] (1) Injection volume control algorithm based on genetic PID algorithm:

[0153] In the injection molding production process, accurate control of the amount of glue injected is crucial to ensure the quality of the patch. The fixed-speed glue injection control method may have problems such as too fast or too slow glue injection due to the influence of the robot's position and external temperature, resulting in low patch accuracy, slow curing speed or loose adhesion. Therefore, a glue injection control algorithm based on genetic PID algorithm is proposed, which aims to automatically adjust the parameters of the PID controller through the optimization ability of the genetic algorithm to achieve accurate control of the glue injection amount.

[0154] (2) Hardware design ideas:

[0155] 1. Determine the injection volume adjustment range: Determine the required injection speed range based on the injection type and application requirements.

[0156] 2. Select the motor adjustment element for injection: select the appropriate push rod motor control system according to the requirements of injection adjustment range, injection force and response time.

[0157] 3. Design the control circuit: Design the corresponding control circuit according to the selected motor regulation controller and control requirements. The control circuit may include feedback loop, comparator, PWM (pulse width modulation) generator, etc.

[0158] 4. Protection circuit: In order to prevent damage to components in the circuit or faults such as short circuits, it is necessary to design corresponding protection circuits, such as overcurrent protection, overheating protection, etc.

[0159] (3) Control process:

[0160] 1. Initialization: Set the parameters of the genetic algorithm, such as population size, crossover probability, mutation probability, and maximum number of iterations. At the same time, initialize the parameter range of the PID controller, such as the upper and lower limits of the proportional coefficient Kp, integral coefficient Ki, and differential coefficient Kd.

[0161] Proportional term (P): Quickly adjust the injection amount based on the error between the current injection amount and the target injection amount.

[0162] Integral term (I): takes into account the cumulative effect of errors and eliminates the steady-state error of the system.

[0163] Differential term (D): predict the changing trend of the error and adjust the injection amount in advance to prevent overshoot.

[0164] Control process: During the operation of the push rod motor, the internal encoder sensor continuously collects displacement calculation speed, that is, the injection speed. The PID controller continuously receives real-time speed data from the sensor and calculates the error between the current speed and the target speed. Then, the required push rod motor current is calculated according to the PID control algorithm, and the signal is transmitted to the motor controller through an actuator (such as a relay). After receiving the signal, its power supply will be adjusted to reduce the speed error. As the error gradually decreases, the output of the PID controller will gradually stabilize, and eventually the speed calculated by the encoder sensor will remain near the target value.

[0165] Speed ​​error represents the difference between actual speed and target speed.

[0166] Output of PID controller

[0167] Among them, u(I) is the instruction to control the speed error, that is, the PID controller adjusts the current output to the push rod motor through the u(I) value, K p , K i , K d They are the gain coefficients of the proportional, integral and differential links respectively. By adjusting K p , K i , K d The value of can change the performance of the controller so that the system can achieve the desired speed control effect.

[0168] 2. Obtain the actual driving force f required in the actual glue injection process, and construct a relationship model between the actual driving force f and related observation parameters (the flip angle of the patch robot, the target position temperature, and the amount of glue in the tube) by applying the deep reinforcement learning algorithm, so as to obtain the actual driving force f through the observation parameters monitored in real time during operation.

[0169] 3. Generate PID population: Randomly generate a set of PID parameters as the initial population. Each individual represents a possible PID controller configuration.

[0170] The operating parameters of the genetic algorithm are set as follows:

[0171] Group size: M

[0172] Termination Algebra: T

[0173] Crossover probability: pc

[0174] Mutation probability: pm

[0175] 3. Simulation: For each individual in the population, its PID parameters are applied to the control system of the electric actuator motor and the response of the brake is simulated.

[0176] Individual variable: x i =[Kp i Ki i Kd i a i ], i represents the i-th population in the genetic algorithm. According to formula 1, the changes in Kp, Ki and Kd affect the changes in the output current controlled by PID. ai is the weight coefficient of each angular speed.

[0177] The corresponding driving force is:

[0178] f=f(I), where f is the input control force of the injection push rod motor system and I is the current corresponding to the control force.

[0179] Build analytical models

[0180] Objective function:

[0181]

[0182] aj is the weight coefficient of the output current of each robot at different horizontal angles. f is the input control force of the injection push rod motor system, and fj(Ij) is the thrust required to output the target propulsion speed at different horizontal angles (controlled by the motor current).

[0183] 4. Fitness evaluation: Calculate the fitness value of each individual based on the simulation results. According to the observed parameters (flip angle of the patch robot, target position temperature, glue content in the tube) obtained from real-time monitoring, the actual driving force f required by the push rod motor is obtained, and Φ(a j ,I j ), and through the fitness function: fitValue = minΦ(a j ,I) conduct evaluation.

[0184] 5. Termination judgment: Check whether the termination conditions are met, such as reaching the maximum number of iterations or finding the PID parameters that meet the requirements, that is, under the PID output current control, the input control force f is close to the actual required driving force. If the conditions are met, stop the iteration and output the optimal PID parameters; otherwise, proceed to the next step.

[0185] 6. Select excellent individuals: Select a part of excellent individuals as the parents of the next generation according to the fitness value, that is, take individuals with smaller fitness values.

[0186] 7. Crossover operation: Perform a crossover operation on the selected parent generation to generate new offspring individuals. The crossover operation can exchange some genes of the parent individuals to generate new PID parameter combinations.

[0187] 8. Mutation operation: Perform mutation operation on the newly generated offspring individuals to increase the diversity of the population. The mutation operation can randomly change a gene value of the offspring individuals to generate new PID parameters.

[0188] 9. Generate a new population: Combine the offspring individuals generated after crossover and mutation operations into a new population and replace the original population.

[0189] 10. Loop: Repeat steps 3 to 9 until the termination condition is met.

[0190] 11. Apply PID control: Apply the found optimal PID parameters to the control system of the injection push rod motor.

[0191] 12. Real-time current control: Use the optimal PID controller to control the real-time current of the injection push rod motor to achieve the desired injection effect.

[0192] The above description is only a preferred embodiment of the present invention and should not be construed as limiting the present application. All equivalent changes and modifications made within the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. Resistance strain gauge pasting robot, characterized in that: It comprises a body and a walking mechanism arranged at the bottom of the body; a lighting lamp and a camera are arranged on the front side of the body; four mounting blocks are arranged at the rear of the body, a mounting base plate is arranged below the mounting blocks, and the mounting base plate is adjusted in position by a position adjustment mechanism, and the position adjustment mechanism comprises a translation component and a lifting component; The four installation base plates are respectively provided with a rough grinding mechanism, a fine grinding mechanism, a glue injection mechanism and a patch mechanism; the four installation base plates are respectively a rough grinding base plate, a fine grinding base plate, a glue injection base plate and a patch base plate; The rough grinding mechanism comprises a rough grinding bottom plate, a rough grinding motor is fixed to the rear end of the rough grinding bottom plate, a drill chuck is fixed to the output shaft of the rough grinding motor, and a rough grinding head is clamped and fixed on the drill chuck; The fine grinding mechanism comprises a fine grinding base plate, a fine grinding motor is fixed to the rear end of the fine grinding base plate, and a fine grinding head is fixed to the output shaft of the fine grinding motor; The glue injection mechanism comprises a glue injection bottom plate, a first electric push rod (hot air pipe) is fixed to the top of the rear end of the glue injection bottom plate, and a glue injection tube is fixed to the bottom of the rear end, and a push rod of the first electric push rod penetrates the glue injection bottom plate and is fixedly connected to the glue injection rod in the glue injection tube; The patch mechanism comprises a patch base plate, a second electric push rod is fixed to the top of the rear end of the patch base plate, a patch syringe is fixed to the bottom of the rear end, a patch plug is arranged in the patch syringe, a push rod of the second electric push rod penetrates the patch base plate and is fixedly connected to the patch plug; the lower end of the inner wall of the patch syringe has a placement groove for placing the patch.

2. The resistance strain gauge pasting robot according to claim 1, characterized in that: The translation assembly includes a first motor fixed in the body, a first driving gear is fixed on the output shaft of the first motor, the first driving gear is meshed with a first driven gear, a screw is fixed on the first driven gear, the front end of the screw is fixed to the inner bottom of the body through a bearing seat, and the rear end is threadedly connected with the right end of a slider; the left end of the slider is sleeved on the first guide rod, the rear end of the first guide rod is fixed to the rear inner wall of the body, and the front end is fixed to a mounting box; The slide block is provided with four oblique grooves, in which positioning pins are passed, and the lower ends of the positioning pins are fixedly connected to the mounting blocks; the four mounting blocks are sleeved on two second guide rods, and the left and right ends of the second guide rods are respectively fixedly connected to the left and right inner walls of the body.

3. The resistance strain gauge pasting robot according to claim 2, characterized in that: The lifting assembly includes a rack arranged in the installation box, the rack meshes with the second driving gear, and the second driving gear is driven by the second motor; a guide block is fixed to the front and rear sides of the rack, the guide block is sleeved on the third guide rod, the left end of the third guide rod is fixed to the left inner wall of the installation box, and the right end is inserted into a sliding box; The right side of the rack is fixedly connected to the left side of the sliding box via a connecting column; an L-shaped plate is arranged inside the sliding box, and sliding grooves are provided at the top and bottom of the sliding box, and the upper and lower ends of the vertical plate of the L-shaped plate pass through the upper and lower sliding grooves respectively, and the front end of the mounting base plate has a bending plate for the horizontal plate of the L-shaped plate to press down; A guide column is fixed on the L-shaped plate in the sliding box, the front and rear inner walls of the sliding box are provided with S-shaped grooves, the front and rear inner walls of the installation box are provided with vertical grooves, and the front and rear ends of the guide column pass through the S-shaped groove and the vertical groove.

4. The resistance strain gauge pasting robot according to claim 3, characterized in that: A moving column is arranged above the mounting block, a first upper baffle is fixed to the upper end of the moving column, and a first lower baffle is fixed to the lower end which penetrates the mounting block and the mounting bottom plate in sequence; a first spring is arranged on the moving column.

5. The resistance strain gauge pasting robot according to claim 1, characterized in that: The walking mechanism comprises a rear bracket fixed to the rear end of the bottom of the body, the left and right sides of the rear bracket are fixed with rear wheel guards, the inner side of the rear wheel guard is provided with a rear wheel, and the rear wheel is driven by a first reduction motor; The front end of the bottom of the body is connected to a front bracket via a connecting frame, and front wheel guards are fixed to the left and right sides of the front bracket. A front wheel is arranged on the inner side of the front wheel guard, and the front wheel is driven by a second reduction motor.

6. The resistance strain gauge pasting robot according to claim 5, characterized in that: The four corners of the top of the front bracket are all fixed with telescopic columns, the upper ends of the telescopic columns penetrate the bottom of the body and are fixed with a second upper baffle; the telescopic columns are provided with a second spring; the material of the connecting frame is an elastomer material.

7. The resistance strain gauge pasting robot according to claim 5 or 6, characterized in that: The front wheels and the rear wheels are both magnetic wheels; the ratio of the chassis height to the wheelbase of the robot is not less than 5:6, so that the robot can successfully climb over a right-angled surface.

8. The resistance strain gauge pasting robot according to claim 5 or 6, characterized in that: A magnetic suction cup is arranged at the middle part of the lower part of the rear bracket, and the magnetic suction cup is driven by the third electric push rod.

9. An image processing method for a resistance strain gauge pasting robot, characterized in that: The method for collecting images through the camera is as follows: (1) Image cropping and compression: The image resolution is 188 columns × 120 rows. To improve processing speed and forward-looking vision, only the first 60 rows of images are captured. At the same time, each row of image data is subjected to non-uniform compression processing, and finally a 120 × 60 image is generated as the path information map. (2) Dynamic threshold image segmentation: The image is divided into upper and lower parts, the average threshold is calculated for each part, and the threshold is adjusted according to formula (1); Among them, DTMOS is the global threshold calculated by the OTSU algorithm, j is the number of image rows, k is the number of shadow area rows, and C is a constant. The image is segmented by the optimal threshold, and a binary image with significant differences can be obtained. (3) Route planning: Taking the second row of the image as the reference row, traverse upward row by row to the top row of the image; in each row scan, use the column shrinking technology, that is, scan every other column, to reduce the amount of data processing and improve the processing speed; Traverse the bottom edge from top to bottom. When a boundary that meets the conditions is found normally, use the continuity and trend changes of the upper and lower boundaries in the image to optimize the boundary search of the next line; the boundary search range of the next line will be based on the center line information center_line of the previous line [ i-1 ] To determine, as shown in formula (2); line_part[i]=[center_line[i-1]-C1,center_line[i-1]] (2) C1 is a parameter that needs to be adjusted according to the actual camera installation position. line_part[i] is the boundary position of the i-th row, that is, the position of the black-white transition point of this row. If the pixels in the current row are all white, the boundary range is determined directly according to formula (3); line_part [ i ] = [ 1,1+C 2] (3) C2 is a parameter that needs to be adjusted according to the actual camera installation position, that is, the boundary is at the initial position; If the pixels in the current row are all black, it is necessary to further determine the pixel value at the midpoint of the previous row. If the midpoint of the previous row is black, the boundary range is given by formula (4). line_part [ i ] = [ Width / 2-C3,Width / 2-C 4] (4) Where width is the width of the cropped image, C3 and C4 are constants determined according to actual conditions; If the midpoint of the previous row is white, the boundary range is given by formula (5), and C5 is a constant determined according to the actual situation; line_part [ i ] = [ 1,Width / 2-C 5] (5) (4) Fitting processing of boundaries and center lines: Based on the positions of the center line, left line, and right line, the information of each row is stored in an array, as described in formula (6), where center_line[i] represents the column where the center line of the current i-th row is located, top_line[i] represents the column where the top line of the current i-th row is located, and bottom_line[i] represents the column where the bottom line of the current i-th row is located. At the same time, in the process of recording the centerline position, the forward travel width weight[i] of each row is also recorded, and its calculation method is shown in formula (7); Based on whether the row weight[i] value is close to 0, we can determine whether there is an obstacle or it is suspended in the air ahead.

10. A method for controlling the amount of glue injected by a resistance strain gauge pasting robot, characterized in that: The injection amount is controlled by genetic PID algorithm as follows: During the operation of the first electric push rod, the internal encoder sensor continuously collects displacement calculation speed, that is, the injection speed, and the PID controller continuously receives real-time data from the encoder sensor and calculates the error between the current injection speed and the target injection speed, that is, the speed error; Then, the required current of the first electric linear actuator is calculated according to the PID algorithm, and the signal is transmitted to the controller of the first electric linear actuator through a relay; after receiving the signal, the controller of the first electric linear actuator adjusts its power supply to reduce the speed error; as the speed error gradually decreases, the output of the PID controller will gradually stabilize, and finally the speed calculated by the encoder sensor will be kept near the target speed; Output of PID controller Among them, u(I) is the instruction for controlling the speed error, that is, the PID controller adjusts the current output to the first electric push rod through the u(I) value, K p , K i , K d They are the gain coefficients of the proportional, integral and differential links respectively. By adjusting K p , K i , K d The value of can change the performance of the PID controller so that the system can achieve the desired speed control effect.

Images