Method for operating a ship plate flattening positioning and machining indication system

Abstract

The application discloses a kind of ship thin plate flattening positioning and processing indication system operating method, system includes electromagnetic heating flattening robot, the inside of the electromagnetic heating flattening robot is equipped with inertial navigation module, data processing and control assembly, front is equipped with monocular camera, LED illuminating lamp, data processing and control assembly are respectively signal connection with electromagnetic heating module assembly of the electromagnetic heating flattening robot, drive motor module assembly, inertial navigation module, monocular camera are respectively signal connection with data processing and control assembly, and coordinate correction beacon and processing indication beacon are arranged at interval on the plane to be processed, monocular camera identifies coordinate correction beacon and processing indication beacon and signal is passed to data processing and control assembly.The application has the characteristics that processing unit is independent, anti-electromagnetic interference, it is convenient to function expansion, it is convenient to maintenance and debugging, greatly reduces manpower cost, effectively improves economic efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of machine vision and metal sheet deformation flattening, in particular to a running method of a ship sheet flattening positioning and processing indication system. BACKGROUND

[0002] Luxury cruise ships need to use a large amount of welding technology to connect metal sheets to build independent cabins during production and manufacturing. The internal stress of the sheet changes due to the high-temperature welding process and the later ship splicing, causing deformation. To meet the acceptance and later processing requirements, the ship deformation sheet needs to be flattened. Electromagnetic induction heating flattening can now replace traditional flame heating for flattening, improving the flattening efficiency, but the device still needs to be manually held for flattening. Therefore, a magnetic attraction type robot is used as a carrier to carry an electromagnetic heating integrated device for automatic flattening. However, in actual use, it is found that the cabin of the cruise ship is a metal cavity, and the electromagnetic environment is complex. The traditional navigation positioning and data transmission method cannot meet the demand of the robot for automatic flattening processing in the cabin.

[0003] Patent CN110798799B "WIFI precise navigation positioning method", through traditional outdoor GPS navigation positioning and indoor WiFi signal positioning, in the construction process, the ship is regarded as a large metal cavity, effectively isolating the GPS signal. Through WiFi positioning and current commonly used indoor navigation technologies such as Bluetooth positioning, the unit point distance is calculated based on signal strength, and the position information is obtained using the three-edge positioning algorithm. However, in actual use, it is found that the signal in the metal cabin presents a rapid decay trend, is seriously affected by noise signals, and has large precision error. In the common cruise ship room cabin with a size of 6m*8m*3m, only decimeter-level positioning can be achieved, which cannot ensure accurate positioning and flattening processing of the processing point. Moreover, when relying on the signal source for cross-room positioning, the signal strength cannot be accepted at the minimum threshold, which is not suitable for the demand of large-area accurate positioning and processing.

[0004] Patent CN112183171A "A method and device for establishing a beacon map based on a visual beacon", uses an odometer combined with a visual beacon to establish a beacon map. However, the deformation of the sheet itself causes the odometer to be unable to accurately calculate the coordinate change in the plane, and the direction determination method through the differential wheel odometer will cause a large amount of error in the direction when processing an irregular undulating plane. In the case of a complex magnetic field environment inside the ship, a large number of visual beacons need to be installed to eliminate cumulative travel error and correct direction, which is not suitable for actual production conditions of metal plane processing on uneven cabin surfaces. Moreover, it cannot complete the positioning and processing indication of the processing point. SUMMARY

[0005] Invention purposes: In view of the above problems, the purpose of the present application is to provide a running method of a ship plate flattening positioning and processing indication system, to improve positioning accuracy and anti-interference, optimize operation, position the point to be processed, and process according to the deformation degree of the point to be processed, adapt to the production cycle characteristics of large ships, and have economy and practicality.

[0006] Technical scheme: A ship plate flattening positioning and processing indication system, comprising an electromagnetic heating flattening robot, an inertial navigation module and a data processing and control assembly are installed inside the electromagnetic heating flattening robot, a monocular camera and an LED illuminating lamp are installed at the front part, the data processing and control assembly is respectively signal connected with an electromagnetic heating module assembly and a driving motor module assembly of the electromagnetic heating flattening robot, the inertial navigation module and the monocular camera are respectively signal connected with the data processing and control assembly, coordinate correction beacons and processing indication beacons are arranged at intervals on a plane to be processed, the monocular camera identifies the coordinate correction beacons and the processing indication beacons and transmits signals to the data processing and control assembly.

[0007] The monocular camera obtains visual beacon detection data of the coordinate correction beacons and the processing indication beacons,

[0008] Further, the inertial navigation module is a ten-axis high-precision inertial navigation ROSIMU module, comprising a three-axis inertial sensor, a three-axis magnetometer and a redundant one-axis gyroscope, the three-axis inertial sensor outputs three-axis angular velocity and three-axis acceleration according to the signal of the redundant one-axis gyroscope, the three-axis magnetometer outputs three-axis magnetometer data, measures the intensity of the earth's magnetic field to calculate the magnetic heading, and transmits the data to the data processing and control assembly.

[0009] Optimally, the redundant one-axis gyroscope is a heading gyroscope with a zero bias stability of 2° / h.

[0010] The inertial navigation module obtains real-time three-axis measurement data and heading zero bias, and the three-axis inertial sensor outputs three-axis angular velocity and three-axis acceleration based on the redundant one-axis gyroscope and the accelerometer.

[0011] Such inertial navigation module has high integration degree, small size and high precision, and the related precision parameters are dynamic heading accuracy <2° / h, angle resolution <0.01°, accelerometer resolution <0.5mg, and position static error ±0.4m / h.

[0012] Further, the coordinate correction beacon is a two-dimensional code card placed in a transparent acrylic card slot and adsorbed on the surface of the metal sheet to be processed by a rubidium magnet behind the card slot, and the coordinate correction beacon is provided with at least one two-dimensional code containing cabin identification number, beacon serial number and positioning point three-axis coordinate string information.

[0013] The coordinate correction beacon takes a two-dimensional code as an information carrier, has the characteristics of convenient arrangement, easy adjustment and easy information replacement.

[0014] Further, the processing indication beacon is a color block beacon, which is printed on the position needing to be flattened on the surface to be processed by laser.

[0015] The processing indication beacon is set after measuring the metal deformation degree, and has a form of a color block beacon located in an identification frame. The processing indication beacon is set by laser printing, and the color of the stainless steel metal after the vaporization of the primer can form a color difference with the color of the primer of the unprocessed part, which can be identified by gray scale. The processing indication beacon has the characteristics of simple and fast setting, no influence on the robot movement, and simple structure and pollution resistance.

[0016] The marking process of the processing indication beacon can be measured by manual or flatness detection instrument, and the processing indication beacon with a suitable process is marked at a suitable position by using a fiber laser marking and engraving machine.

[0017] Further, the data processing and control assembly includes a processor, a memory, an inertial navigation module, a monocular camera, an electromagnetic heating module assembly, a driving motor module assembly, and the memory is signal connected with the processor.

[0018] The memory is used to store a computer program, and the processor is used to execute the program stored on the memory to realize the method of identifying and reading various beacons, processing the output data of the inertial navigation, and controlling the robot driving motor and the electromagnetic flattening module based on the information output.

[0019] The data processing and control assembly further includes a computer readable storage medium, and the storage medium stores a computer program, a cabin information database and a processing technology database, and the computer program is executed by the processor to perform the above work.

[0020] After setting the initial coordinates and the traversal path, the processor carried by the robot controls the movement and turning of the robot based on the tracking path and the global coordinates currently located by the robot.

[0021] The inertial navigation module outputs real-time three-axis acceleration and heading zero offset, and the integral algorithm is used for the acceleration data, the first integral obtains three-axis velocity, and the second integral obtains three-axis displacement.

[0022] The processor calculates the real-time coordinates currently located in the global domain according to the three-axis displacement, and corrects the path according to the coordinates and the heading angle, so that the robot moves according to the predetermined route.

[0023] When the robot moves to the coordinate correction beacon, the monocular camera collects the image data of each beacon, and the computer extracts the visual beacon detection data of the beacon image frame, and the specific method is as follows:

[0024] Firstly, the image frame is converted into a gray image, converted into a binary image through the OTSU algorithm (Otsu threshold segmentation method), and the contour data is obtained by using the contour discovery function on the binary image.

[0025] The data obtained by decoding the two-dimensional code is extracted, and the three-axis absolute coordinates of the positioning points in the data are used to replace the real-time coordinate data currently calculated by the computer based on the initial coordinates and inertial navigation data, and the coordinate initialization is completed based on the current global coordinate information, and the cumulative error caused by inertial navigation is eliminated.

[0026] When the robot travels to the processing indication beacon, the monocular camera collects visual beacon image data, and the computer extracts visual beacon detection data of the visual angle beacon image frame, converts the image frame into a gray image, converts it into a binary image through the OTSU algorithm (Otsu threshold segmentation method), and obtains contour data by using the contour discovery function on the binary image.

[0027] Based on the processing process database, the data obtained by visual recognition of the processing indication beacon is compared, and the processing process indicated by the beacon is selected, and the corresponding subroutine of the process is called.

[0028] In actual use, electromagnetic heating flattening single processing can heat the metal wallboard to 650-700℃ in 5-6s, change its internal stress by rapid temperature difference, and the single processing range is 5cm*14cm of round rectangular. Compared with the metal sheet which deforms to different degrees due to the change of internal stress, the processing technology for different deformation degrees has been summarized in actual production, and the 5cm*14cm round rectangular generated by the single processing range is taken as the basic image in the regular array of the deformation area. The deformation error after flattening is less than 2mm, which meets the acceptance requirements, and realizes the standardized flattening workflow of the corresponding deformation plane.

[0029] The computer runs the subroutine, controls the flattening robot to execute the processing process with the processing indication beacon as the base point.

[0030] After completing the flattening work, the flattening robot continues to track the path, and executes the above work according to the image data collection and processing, until the traversal tracking of the local area to be processed is completed.

[0031] An operation method of the above-mentioned ship plate flattening positioning and processing indication system, comprising the following steps:

[0032] Step 1: Preprocessing based on the plane to be processed;

[0033] Enter the cabin, select the surface to be processed and set the coordinate correction beacon, obtain the thickness N of the metal sheet to be processed, measure the flatness of the surface of the metal sheet to be processed, and set the processing instruction beacon of the corresponding specification according to the fluctuation at the place where the fluctuation is greater than N-1;

[0034] Step two: the electromagnetic heating flattening robot realizes positioning tracking based on the inertial navigation module;

[0035] The electromagnetic heating flattening robot enters the cabin to work, and the inertial navigation module outputs real-time three-axis acceleration, three-axis angular velocity and heading zero offset angle. The data processing and control assembly obtains the body running attitude data through integral algorithm based on the obtained data, obtains the global coordinates of the body in the processing domain based on the displacement amount, and obtains the body direction based on the three-axis steering angle and the heading angle. Real-time adjustment of the output of the drive motor module assembly is realized by combining the path data, so as to realize real-time positioning of the electromagnetic heating flattening robot.

[0036] Step three: identify the processing instruction beacon and locate the processing point;

[0037] The electromagnetic heating flattening robot tracks along the path, and the monocular camera collects image data. The data processing and control assembly screens whether the key frame contains a beacon image. When an image frame containing a complete processing instruction beacon is identified, the current coordinates are marked and the current coordinate information is stored, and the electromagnetic heating flattening robot stops tracking along the path.

[0038] Step four: select the corresponding processing technology;

[0039] The data obtained by visual recognition of the processing instruction beacon is compared based on the beacon contour data in the processing technology database, the processing technology indicated by the beacon is selected, and the corresponding subprogram is called;

[0040] Step five: the data processing and control assembly sends a signal to the electromagnetic heating module assembly, and the electromagnetic heating flattening robot completes the flattening of the sheet;

[0041] Step six: eliminate the coordinate error caused by processing;

[0042] After the flattening is completed, the data processing and control assembly reads the stored coordinate information in step three, and eliminates the coordinate error generated by the inertial navigation module during the flattening,

[0043] Step seven: the electromagnetic heating flattening robot continues to track along the path, and when an image frame containing a complete processing instruction beacon is identified again, steps three to six are repeated;

[0044] Step eight: identify the coordinate correction beacon;

[0045] The monocular camera acquires image data, and the data processing and control assembly filters key frames to see if they contain beacon images. When an image frame containing a complete coordinate correction beacon is identified, the data processing and control assembly extracts the coordinate correction beacon information to obtain the current fixed coordinates.

[0046] Step 9: The data processing and control assembly updates the coordinate parameters to correct for accumulated errors;

[0047] Step 10: The electromagnetic heating leveling robot continues to follow the path. When a processing instruction beacon is detected, repeat steps 3 to 6; when a coordinate correction beacon is detected, repeat steps 8 to 9.

[0048] Step 11: Based on the image data acquisition, the electromagnetic heating leveling robot completes the path traversal, and the leveling work of the deformed thin plate on the surface to be processed is completed.

[0049] Furthermore, in step one, the deformation is measured manually or by a flatness testing instrument, and the corresponding processing indicator beacon is selected and marked using a fiber laser marking and engraving machine.

[0050] Furthermore, the processing indicator beacon is divided into four grids in a grid pattern. Each grid has two forms: filled with color and transparent. Based on the degree of undulation, there are three specifications from small to large: low undulation with only one filled grid, medium undulation with two diagonally filled grids, and high undulation with three filled grids. By changing the distribution of the color blocks to correspond to the metal deformation measurement value, the distribution adjustment span is 1.5mm.

[0051] Beneficial Effects: Compared with existing technologies, the advantages of this invention are: enabling high-precision navigation and positioning within the nearly fully enclosed metal cavity of a ship's hull, without requiring navigation signals or a local area network connection, relying solely on its own inertial navigation module and a small number of external visual beacons for correction. After setting a few positioning visual beacons, inputting initial coordinates, and planning the path, path traversal can be performed. Simultaneously, it enables precise positioning by recognizing visual beacon marks, allowing for accurate electromagnetic heating and leveling of deformed welded thin metal plates without manual control, wireless signal control, or wired signal control. It solves the significant impact of the ship's hull, as a complex electromagnetic interference source, on equipment positioning and information transmission, thus automating the high-precision ship deformation thin plate leveling process.

[0052] The use of machine vision and visual beacon enables the relevant robot to realize accurate machining point positioning and process indication in the case that wireless connection communication cannot be used and data cannot be uploaded to the server in real time, simplifies the process flow and actual use difficulty. The independent machining unit only exchanges visual information with the outside world, is not affected by the remaining units and online system downtime failure, and can realize simultaneous machining of multiple machining and flattening units according to the visual beacon of any detection unit in the case of no connection, improve machining efficiency, and reduce equipment cost. It is beneficial to later function expansion and is convenient for executing subsequent processing of the flattening process, such as that the corresponding functional robot can directly determine the primer repair range and the machining area metal flaw detection range according to the visual beacon.

[0053] The present application effectively solves the problems of positioning difficulty of the flattening robot and processing wireless control difficulty of a large ship represented by a luxury cruise ship in actual production, which has a large size, a large number of internal cabins, and high flatness requirement of cabin wall surface, and realizes accurate positioning and processing indication of the sheet flattening robot in the ship cabin during the ship manufacturing cycle without receiving other information from the outside world except visual information. The machining unit is independent, anti-electromagnetic interference, easy to expand functions, easy to maintain and debug, greatly reduces the labor cost, and effectively improves the economic benefit. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 The workflow diagram of the present application is shown in the figure;

[0055] Figure 2 The structural schematic diagram of the present application is shown in the figure;

[0056] Figure 3 The traversal path schematic diagram of the electromagnetic heating flattening robot in an embodiment is shown in the figure;

[0057] Figure 4 The data transmission processing and control schematic diagram of the present application is shown in the figure;

[0058] Figure 5 The processing indication beacon and the corresponding electromagnetic flattening domain array schematic diagram is shown in the figure;

[0059] Figure 6 The coordinate correction beacon structure schematic diagram is shown in the figure. DETAILED DESCRIPTION

[0060] The present application will be further illustrated in conjunction with the drawings and specific embodiments, and it should be understood that these embodiments are only used to illustrate the present application and not to limit the scope of the present application.

[0061] A marine sheet flattening positioning and processing indication system, such as Figures 2-6As shown, including electromagnetic heating leveling robot 202, the inside of electromagnetic heating leveling robot 202 is mounted with inertial navigation module 102, data processing and control assembly 104, the front is mounted with monocular camera 101, LED light 103.

[0062] Data processing and control assembly 104 includes processor 1041, memory 1042, inertial navigation module 102, monocular camera 101, electromagnetic heating module assembly 105, drive motor module assembly 106, memory 1042 is respectively connected with processor 1041 signal. The coordinate correction beacon 206 and the processing instruction beacon 203 are arranged on the plane to be processed. The monocular camera 101 identifies the coordinate correction beacon 206 and the processing instruction beacon 203 and transmits the signal to the data processing and control assembly 104.

[0063] Inertial navigation module 102 is a ten-axis high-precision inertial navigation ROSIMU module, including three-axis inertial sensor, three-axis magnetometer and redundant one-axis gyroscope. The redundant one-axis gyroscope is a heading gyroscope with a zero bias stability of 2° / h. The three-axis inertial sensor outputs three-axis angular velocity and three-axis acceleration according to the signal of the redundant one-axis gyroscope. The three-axis magnetometer outputs three-axis magnetometer data, measures the earth's magnetic field intensity, calculates the magnetic heading, and transmits them to the data processing and control assembly 104 respectively.

[0064] The coordinate correction beacon 206 is a two-dimensional code card placed in a transparent acrylic card slot and adsorbed on the surface of the metal sheet to be processed by a rubidium magnet behind the card slot. The coordinate correction beacon 206 is provided with at least one two-dimensional code containing cabin identification number, beacon serial number and three-axis coordinate string information.

[0065] The processing instruction beacon 203 is a color block beacon printed on the surface to be processed by laser.

[0066] The operation method of the above-mentioned ship plate leveling positioning and processing indication system, as shown in Figure 1 The implementation process includes the following steps:

[0067] Step 1, preprocessing based on the processing plane.

[0068] Referring to the accompanying Figure 3 As shown, the 6*8m cabin commonly used in luxury cruise ships is selected as an example. The cabin includes cabin entrance 201, support structure steel 205, and 3mm stainless steel plate welded plane for leveling processing. The plane is measured for flatness, and the thin plate deformation fluctuation >2mm is marked with laser processing indication beacon based on the deformation amount.

[0069] In addition, the marking process can be measured by manual or flatness detection instrument, and the appropriate process beacon is selected for marking by using the fiber laser marking engraving machine. In the embodiment, the tracking path 204 is a rotary S-shaped curve, and the tracking path 204 can also be a spiral line and the like.

[0070] Based on the structural characteristics of the processing area, two coordinate correction beacons are arranged at 25m and 40m of the tracking route. The beacons are arranged near the wall, and the thin plate deformation at this structure is small, and there is no need for flattening processing, and the flattening robot will not be affected.

[0071] The traversal path data corresponding to the processing plane in the memory is selected, and the global coordinate where the current flattening robot is located is set as the initial coordinate.

[0072] Step two, the flattening robot realizes positioning tracking based on inertial navigation.

[0073] The inertial navigation module outputs real-time three-axis acceleration, three-axis angular velocity and heading zero offset angle. The processor obtains the body running attitude data including three-axis displacement, three-axis turning angle and heading angle offset data through integral algorithm based on the obtained data.

[0074]

[0075] The processor calculates the global coordinate of the body in the processing domain based on the displacement, obtains the direction of the body based on the three-axis turning angle and the heading angle, and adjusts the output of the drive motor with the Hall encoder in real time to realize the accurate positioning of the flattening robot without relying on external signal sources.

[0076] Compared with the odometer, the odometer is affected by the errors such as slippage generated during movement, insufficient flatness of the driving surface, differential loss of the transmission system, and the like, which leads to difficult-to-adjust offset error in the heading direction, and cannot meet the production requirements. The inertial navigation module only collects and outputs its own motion attitude data. The related accuracy parameters are dynamic heading accuracy <2° / h, angle resolution <0.01°, accelerometer resolution <0.5mg, and position static error ±0.4m / h. The mature ten-axis inertial navigation technology initially has a traditional three-axis acceleration, angular velocity, and magnetometer, and a redundant heading gyroscope with a bias stability of 2° / h, which can realize small offset error in direction and distance for a long time, but there is a problem of position offset caused by non-movement over time, and the coordinate still needs to be corrected within a certain time.

[0077] Step three, identify the processing instruction beacon and locate the processing point.

[0078] A monocular camera acquires image data, and a computer filters keyframes to see if they contain beacon images. When identifying an image frame containing a complete processing instruction beacon, the computer marks the current coordinates and stores the current coordinate information.

[0079] Visual beacon detection data is extracted from viewpoint beacon image frames. The image frames are converted into grayscale images and then into binary images using the OTSU algorithm (Otsu thresholding method). Contour detection algorithms are then used on the binary images to obtain contour data.

[0080] Step 4: Select the appropriate processing technology.

[0081] Reference Appendix Figure 5 As shown, the data obtained from the visual recognition of the processing instruction beacon is compared with the beacon contour data in the processing technology database, the processing technology indicated by the beacon is selected, and the corresponding subroutine of the process is called.

[0082] The processing instruction beacon 203 is divided into four grids in a grid pattern. Each grid has two forms: filled with color and transparent. According to the degree of undulation, there are three specifications from small to large: low undulation with only one grid filled with color, medium undulation with two diagonally filled grids, and high undulation with three grids filled with color.

[0083] Step 5: Complete the thin plate leveling.

[0084] Reference Appendix Figure 5 As shown, in practical applications, electromagnetic heating leveling can heat a metal panel to 650-700℃ in 5-6 seconds during a single processing step. This rapid temperature difference alters the internal stress, and the processing range for a single step is a 5cm x 14cm rounded rectangle. Compared to thin metal sheets that deform to varying degrees due to changes in internal stress, processing techniques for different deformation levels have been developed in actual production. These techniques utilize the 5cm x 14cm rounded rectangles generated in a single processing step as the basic unit, arranged in a regular array within the deformation area. Actual measurements show that the deformation error after leveling is <2mm, meeting acceptance requirements and achieving a standardized leveling workflow for the corresponding deformed plane.

[0085] The subroutine is a fixed process flow control program. The processor controls the motor output, and the leveling robot moves to move the electromagnetic heating module to the designated position. The electromagnetic heating module is powered on to heat-treat the 5cm*14cm metal panel below. The processing time for the 3mm thin plate is set to 6 seconds. After the time is up, the electromagnetic heating module is powered off, and the single processing cycle ends. The above process is repeated until a regular array of the single processing range is completed within the designated area. The program then returns to the beacon point, and the subroutine ends.

[0086] Step six: Eliminate coordinate errors caused by machining.

[0087] Step three, read the coordinate information, eliminate the coordinate error generated by the inertial navigation during the flattening process, and continue to track along the path.

[0088] It should be noted that during the flattening of the sheet, the displacement of the robot is directly controlled by the fixed parameters in the subroutine based on the encoder feedback signal of the motor to drive the servo motor, which can be applied to low-speed short-distance movement in a small range during the machining process, and the error of the beacon point after returning is less than 5mm, which avoids the position deviation caused by the inertial navigation module during the long time flow of the sheet flattening stage.

[0089] Step seven, identify the coordinate correction beacon.

[0090] The monocular camera collects image data, and the computer screens whether the key frame contains a beacon image. When the image frame containing the complete coordinate correction beacon is identified, the computer extracts the visual beacon detection data of the visual beacon image frame.

[0091] Step eight, coordinate replacement to eliminate errors.

[0092] First, convert the image frame to a grayscale image, convert it to a binary image through the OTSU algorithm (Otsu threshold segmentation method), use the contour finding function to get the contour algorithm by finding the contour of the binary image, filter the area and proportion of the contour according to the characteristics of the three regions of the two-dimensional code, and decode the data code word to obtain the string information.

[0093] For example, the string obtained by decoding the two-dimensional code is A03 004X17684 Y54236 Z00350, the processing area code is A03, the beacon serial number is 004, and the global coordinates of the correction point are X axis 26700, Y axis 56732, and Z axis 00350.

[0094] Reference to the attached Figure 6 Extract the data obtained by decoding the two-dimensional code, replace the real-time relative coordinate data calculated by the computer based on the initial coordinates and inertial navigation data with the three-axis absolute coordinates in the data, complete the coordinate initialization based on the current global coordinate information, and eliminate the cumulative error caused by inertial navigation.

[0095] It should be noted that based on the calculation and actual use, in the actual production environment, the appropriate installation distance of the coordinate correction beacon is 20-30m, and the real-time coordinate error calculated based on the output data of the inertial navigation module during this period is less than 0.05m, which meets the detection and recognition requirements of the monocular camera for the beacon in the region. The measured heading angle zero offset is less than 1° after completing the flattening task of a single cabin, which meets the direction confirmation requirement.

[0096] Step nine, based on the image data collection, execute the above steps to complete the path traversal. That is, the deformation sheet flattening work in this area is completed.

Claims

1. A method for operating a marine thin plate leveling, positioning, and processing instruction system, characterized in that... Includes the following steps: Step 1: Preprocess the plane to be processed; Enter the chamber, select the surface to be processed and set the coordinate correction beacon (206), obtain the thickness N of the metal sheet to be processed, measure the surface flatness of the metal sheet to be processed, and set the corresponding processing indicator beacon (203) according to the degree of undulation where the undulation is greater than N-1. Step 2: The electromagnetic heating leveling robot (202) achieves positioning and tracking based on the inertial navigation module (102); The electromagnetic heating and leveling robot (202) enters the cabin to work. The inertial navigation module (102) outputs real-time three-axis acceleration, three-axis angular velocity and zero heading angle. The data processing and control assembly (104) obtains the body's running posture data through an integral algorithm based on the obtained data, calculates the global coordinates of the body in the processing domain based on the displacement, obtains the body's direction based on the three-axis steering angle and heading angle, and adjusts the output of the drive motor module assembly (106) in real time in combination with the path data to realize the real-time positioning of the electromagnetic heating and leveling robot (202). Step 3: Identify the processing instruction beacon (203) and locate the processing point; The electromagnetic heating leveling robot (202) follows the path, the monocular camera (101) collects image data, the data processing and control assembly (104) filters whether the key frame contains a beacon image, when the image frame containing the complete processing instruction beacon (203) is identified, the current coordinates are marked and the current coordinate information is stored, and the electromagnetic heating leveling robot (202) stops following the path. Step 4: Select the appropriate processing technology; The processing instruction beacon visual recognition data is compared with the beacon contour data in the processing technology database. The processing technology indicated by the beacon is selected and the corresponding subroutine is called. Step 5: The data processing and control assembly (104) sends a signal to the electromagnetic heating module assembly (105), and the electromagnetic heating leveling robot (202) completes the leveling of the thin plate; Step Six: Eliminate coordinate errors caused by machining; After the leveling process is completed, the data processing and control assembly (104) reads the coordinate information stored in step three to eliminate the coordinate error generated by the inertial navigation module (102) during the leveling process. Step 7: The electromagnetic heating leveling robot (202) continues to follow the path. When it recognizes an image frame containing the complete processing instruction beacon (203) again, it repeats steps 3 to 6. Step 8: Identify coordinate correction beacons (206); The monocular camera (101) acquires image data, and the data processing and control assembly (104) filters key frames to see if they contain beacon images. When an image frame containing a complete coordinate correction beacon (206) is identified, the data processing and control assembly (104) extracts the coordinate correction beacon (206) information to obtain the current fixed coordinates. Step 9: The data processing and control assembly (104) updates the coordinate parameters to correct for accumulated errors; Step 10: The electromagnetic heating leveling robot (202) continues to follow the path. When the processing instruction beacon (203) is detected, steps 3 to 6 are repeated. When the coordinate correction beacon (206) is detected, steps 8 to 9 are repeated. Step 11: Based on the image data acquisition, the electromagnetic heating leveling robot (202) completes the path traversal, and the leveling work of the deformed thin plate on the surface to be processed is completed.

2. The operation method of the marine thin plate leveling, positioning and processing indication system according to claim 1, characterized in that: In step one, the deformation is measured manually or by a flatness testing instrument, and the corresponding processing indicator beacon (203) is selected and marked using a fiber laser marking and engraving machine.

3. The operation method of the marine thin plate leveling, positioning and processing indication system according to claim 1, characterized in that: The processing indicator beacon (203) is divided into four grids in a grid pattern. Each grid has two forms: filled with color and transparent. According to the degree of undulation, there are three specifications from small to large: low undulation has only one grid filled with color, medium undulation has two grids filled with color diagonally, and high undulation has three grids filled with color.

Images