A method for identifying the weld seams and trajectory planning of the corrugated oil tank of an oil-immersed transformer

Abstract

The present invention relates to the field of manufacturing oil-immersed transformers, and provides a method for identifying weld seams and planning trajectories of corrugated oil tanks of oil-immersed transformers, which includes a calibration line laser sensor, teaching the scanning path of the line laser sensor and a welding robot; scanning the weld seam features between the corrugated oil tank and the frame through the line laser sensor, and outputting the point cloud data of the weld seam; segmenting the point cloud data of the weld seam into multiple point cloud segments, setting the interval threshold of the point cloud segments, defining the gap of the point cloud data greater than the interval threshold as a large weld seam gap, and obtaining the position data of the large weld seam gap; planning the reciprocating swing welding trajectory of the weld seam according to the position data of the large weld seam gap; and generating the weld seam trajectory of the welding robot according to the planned reciprocating swing welding trajectory and the point cloud segments. The present invention can solve the problems of large recognition error of the positions of the wave fins of the radiator, improper handling of the positions of large weld seam gaps, and unreasonable planning of the swing welding trajectory, effectively ensuring the welding quality of the transformer corrugated oil tank and the airtightness of the weld seam.

Description

Technical Field

[0001] The present invention relates to the field of manufacturing oil-immersed transformers, and particularly to a method for identifying the weld seams and trajectory planning of the corrugated oil tank of an oil-immersed transformer. Background Art

[0002] The oil-immersed transformer is an important device in the power system, and its stable operation is crucial for the safety of the power system. As a key component of the transformer, the welding quality of the corrugated oil tank directly affects the airtightness and service life of the transformer. After prefabrication, the corrugated oil tank needs to be welded to the transformer oil tank frame. The weld form between the corrugated oil tank and the frame is lap joint. At the lower opening positions at both ends of the corrugated fins, there are herringbone gaps formed by bending, and generally weaving welding is required to ensure filling. In the past, when manual welding was used, the welding quality was restricted by the experience and skills of the welders, and it was difficult to improve both the production efficiency and quality. With the development of technology, the weld visual recognition technology has been widely used, and it has become possible to automatically recognize and weld the weld seams between the corrugated oil tank and the frame. One of the technical keys is how to accurately identify the positions of the corrugated fins of the corrugated oil tank and plan a reasonable welding path at the fin positions to achieve stable and full fusion filling of the large-gap weld seams.

[0003] The prior art patent CN 112122775B identifies the corrugated fins by judging the laser line angle threshold. However, in practical applications, it has the following deficiencies: 1. It has insufficient adaptability to the placement deviation of the workpiece. The laser weld tracking technology mainly solves the problem that the traditional teaching reproduction trajectory does not match the actual position of the weld seam under the condition of rough positioning of non-standard product parts. In the actual application scenario, the scanning tracking path is preset by teaching. When the workpiece is placed with rough positioning and there is an obvious deviation from the original teaching position, the laser incident angle changes. When the laser incident angle changes, the laser line angles at the fin positions are different, and it is difficult to set the angle threshold, which is likely to cause misjudgment of the fin positions. 2. It cannot identify and process the large-gap positions of the weld seams at non-fin positions. For non-fin positions, when the local deformation causes a large gap in the lap joint, this solution cannot make an identification and processing, and it is easy to have incomplete welding. 3. It is easily interfered. The implementation method is to identify and track in real time during the welding process. The recognition accuracy is affected by the welding arc light, spatter and the movement of the robot, and there will be cases of missed recognition. 4. The weaving welding trajectory planning is unreasonable. Because the lower opening of the corrugated fin plate is a herringbone large gap, simple weaving welding is likely to result in incomplete filling or poor fusion. Therefore, a method for identifying the weld seams and trajectory planning of the corrugated oil tank of an oil-immersed transformer based on laser scanning is needed. This method can solve the problems of large recognition errors at the fin positions of the radiator fins, improper treatment of the large-gap positions of the weld seams and unreasonable weaving welding trajectory planning in the prior art, and can improve the manufacturing quality and welding automation level of the oil-immersed transformer. Summary of the Invention

[0004] To solve the technical problems existing in the prior art, the present invention provides a method for identifying and trajectory planning of the welds of the corrugated oil tank of an oil-immersed transformer. By combining laser weld identification, point cloud data processing, and swing welding trajectory planning, it can solve the problems of large error in identifying the position of the wave fins of the radiator, improper handling of the large-gap position of the weld, and unreasonable swing welding trajectory planning in the prior art.

[0005] The object of the present invention can be achieved by adopting the following technical solutions:

[0006] A method for identifying and trajectory planning of the welds of the corrugated oil tank of an oil-immersed transformer, comprising the following steps:

[0007] S1. Calibrate the line laser sensor, install the line laser sensor on the welding torch of the welding robot, and adjust the installation position and height of the line laser sensor;

[0008] S2. Teach the scanning path of the line laser sensor and the torch posture of the welding robot;

[0009] S3. Scan the weld features between the corrugated oil tank and the frame through the line laser sensor, set the weld height threshold, output the point cloud data of the weld according to the weld height threshold, and transmit the point cloud data of the weld to the welding robot in real time;

[0010] S4. Divide the point cloud data of the weld into multiple point cloud segments, set the interval threshold of the point cloud segments, and define the gap of the point cloud data greater than the interval threshold as the large gap of the weld;

[0011] S5. Obtain the position data of the large gap of the weld according to the position data of the weld point cloud segments, and plan the round-trip swing welding trajectory of the weld according to the position data of the large gap of the weld;

[0012] S6. Generate the weld trajectory of the welding robot according to the planned round-trip swing welding trajectory and the point cloud segments, and the welding robot welds the weld of the corrugated oil tank of the oil-immersed transformer according to the weld trajectory.

[0013] Specifically, teaching the scanning path of the line laser sensor and the torch posture of the welding robot includes: setting the user coordinate system with a certain point on the corrugated oil tank frame as the origin, setting the base coordinate system with the center of the body base of the welding robot as the origin, teaching the robot scanning path in the user coordinate system, with the scanning start position in front of the weld start end and the scanning end position behind the weld end.

[0014] Specifically, the step S3 includes the steps:

[0015] Adopt the line laser weld identification technology to scan the weld features between the corrugated oil tank and the frame through the line laser sensor;

[0016] Obtain the lap weld height h at different positions by laser sensor scanning, and preset the weld height threshold as h max , when h < h max , the sensor outputs position data to the welding robot; when h > h max , the laser sensor does not output position data.

[0017] Specifically, the step S4 includes: further segmenting each weld point cloud segment by using the random sample consensus algorithm to remove the outlier points on the weld point cloud segment.

[0018] Specifically, the step S5 of planning the reciprocating weaving trajectory of the weld according to the position data of the large gap of the weld specifically includes:

[0019] S51. Calculate the coordinates of the gap center position point according to the position coordinates of the two adjacent front and rear point cloud segments of the large gap of the weld;

[0020] S52. Search forward a preset distance from the gap center position point to obtain the feature points on the front point cloud segment, and use the feature points on the front point cloud segment as the weaving start point.

[0021] S53. Search backward a preset distance from the gap center position point to obtain the feature points on the rear point cloud segment, and use the feature points on the rear point cloud segment as the last weaving point;

[0022] S54. Calculate the weaving midpoint coordinates according to the coordinates of the weaving start point and the coordinates of the last weaving point;

[0023] S55. Set the weaving height according to the welding requirements so that the molten pool can fill the large gap, set the weaving offset in the X-axis direction of the user coordinate system, and respectively represent the coordinates of the weaving start point, the highest weaving point, and the end weaving point after calculating and compensating the offset according to the weaving height and the weaving offset. Plan and generate the reciprocating weaving trajectory of the weld according to the coordinates of the weaving start point, the highest weaving point, and the end weaving point after compensating the offset.

[0024] Specifically, the step S55 includes:

[0025] Establish a user coordinate system and a base coordinate system. The X-axis direction of the user coordinate system is the welding direction, and the Z-axis direction of the user coordinate system is perpendicular to the plane of the welded workpiece. Calculate the coordinates of the weaving start point after compensating the offset according to the coordinates of the weaving start point and the weaving offset in the user coordinate system;

[0026] Calculate the coordinates of the highest weaving point after compensating the offset according to the coordinates of the weaving midpoint, the weaving height, and the weaving offset in the user coordinate system;

[0027] According to the coordinate representation of the last point of the weaving welding in the user coordinate system and the weaving offset, calculate the coordinate representation of the last point of the weaving welding after compensating the offset;

[0028] Respectively convert the coordinate representations of the starting point, the highest point, and the ending point of the weaving welding after calculating and compensating the offset back to the coordinate representations of the starting point, the highest point, and the ending point of the weaving welding in the base coordinate system, and plan and generate the reciprocating weaving welding trajectory of the weld seam according to the coordinate representations of the starting point, the highest point, and the ending point of the weaving welding in the base coordinate system.

[0029] Specifically, calculating the coordinate representation of the highest point of the weaving welding after compensating the offset according to the coordinate representation of the midpoint of the weaving welding, the weaving height, and the weaving offset offset, including:

[0030] Calculating the coordinate representation of the highest point of the weaving welding after compensating the offset according to the coordinate representation of the midpoint of the weaving welding, the weaving height, and the weaving offset offset, including:

[0031] Suppose the coordinate representation of the midpoint of the weaving welding in the user coordinate system is B user (x Bu , y Bu , z Bu ), and the coordinate representation of the highest point of the weaving welding is B' user (x' Bu , y Bu , z' Bu ), then z' Bu = z Bu + w, x' Bu = x Bu + offset; where w is the weaving height and offset is the weaving offset.

[0032] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0033] The present invention provides a method for identifying and trajectory planning of the weld seam of the corrugated oil tank of an oil-immersed transformer. By scanning the weld seam features between the corrugated oil tank and the frame with a line laser sensor, setting a weld seam height threshold, outputting the point cloud data of the weld seam according to the weld seam height threshold, dividing the point cloud data of the weld seam into multiple point cloud segments, setting an interval threshold for the point cloud segments, defining the interval of the point cloud data greater than the interval threshold as the large gap of the weld seam, obtaining the position data of the large gap of the weld seam, and planning the reciprocating weaving welding trajectory of the weld seam according to the position data of the large gap of the weld seam. It can solve the problems of large recognition error of the wave fin position of the radiator, improper treatment of the large gap position of the weld seam, and unreasonable planning of the weaving welding trajectory in the prior art, effectively ensuring the welding quality and weld seam airtightness of the transformer corrugated oil tank, and having important significance for improving the manufacturing quality and welding automation level of the oil-immersed transformer. Description of the Drawings

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0035] Figure 1 is a schematic structural diagram of the corrugated oil tank of the oil-immersed transformer in the embodiment of the present invention;

[0036] Figure 2 is a step diagram of the method for identifying and trajectory planning of the weld seam of the corrugated oil tank of the oil-immersed transformer in the embodiment of the present invention;

[0037] Figure 3 is a schematic diagram of the placement relationship between the corrugated oil tank of the transformer and the welding robot in the embodiment of the present invention;

[0038] Figure 4 is a schematic diagram of the typical image features when the line laser irradiates on the weld seam in the embodiment of the present invention;

[0039] Figure 5 is a schematic cross-sectional diagram of the lap joint in the embodiment of the present invention;

[0040] Figure 6 is a schematic diagram of the point cloud data in the embodiment of the present invention;

[0041] Figure 7 is a schematic diagram of the swing welding trajectory planning in the embodiment of the present invention;

[0042] Figure 8 is a schematic diagram of the swing welding trajectory planning points in the embodiment of the present invention;

[0043] Figure 9 is a schematic diagram of the reciprocating swing welding trajectory in the embodiment of the present invention;

[0044] The labels in the figure are: 1 - corrugated oil tank, 2 - corrugated fins of the oil tank, 3 - transformer oil tank frame, 4 - lap joint, 5 - laser sensor, 6 - six-axis multi-joint robot, 7 - user coordinate system, 8 - base coordinate system. Detailed implementation manners

[0045] The following will further describe the technical solutions of the present invention in detail in combination with the drawings and embodiments. Obviously, the described embodiments are some but not all of the embodiments of the present invention, and the implementation manners of the present invention are not limited thereto. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0046] Example 1:

[0047] As Figure 1 shown, it is a schematic diagram of the corrugated tank structure of an oil-immersed transformer. The corrugated tank 1 has a corrugated shape structure prefabricated by a special machine through die pressing and folding. The two ends of the corrugated fins 2 of the corrugated tank are sealed by welding, and the inside of the corrugated fins 2 of the corrugated tank is hollow and can store oil. After prefabrication, the corrugated tank 1 needs to be welded and installed on the transformer tank frame 3. The weld form between the corrugated tank and the frame is lap joint. There is a herringbone lap joint 4 formed by bending at the lower opening position at both ends of the corrugated fins. It is necessary to accurately identify the position of the corrugated fin sheets of the corrugated tank and plan a reasonable welding path at the corrugated fin position to achieve stable and full deposition filling of the large-gap weld.

[0048] As Figure 2 shown, it is a step diagram of a method for identifying and trajectory planning of the weld of the corrugated tank of an oil-immersed transformer. The method for identifying and trajectory planning of the radiator fin weld of an oil-immersed transformer based on laser scanning according to the present invention includes the following steps:

[0049] S1. Calibrate the line laser sensor. Install the line laser sensor on the welding torch of the welding robot, and adjust the installation position and height of the line laser sensor.

[0050] Specifically, install the line laser sensor on the calibrated welding torch, adjust the installation position and height of the line laser sensor so that the welding torch does not interfere with the workpiece during scanning, and the distance between the sensor lens and the workpiece surface is within the effective viewing distance of the sensor to calibrate the line laser sensor to obtain accurate laser scanning data.

[0051] S2. Teach the scanning path of the line laser sensor and the welding torch posture of the welding robot.

[0052] Specifically, teach the scanning path of the line laser sensor. Set the user coordinate system with a certain point on the corrugated tank frame as the origin, and set the base coordinate system with the center of the body base of the welding robot as the origin. Teach the robot scanning path in the user coordinate system. The scanning start position is placed in front of the start end of the weld, and the scanning end position is placed behind the end of the weld to ensure the integrity of the weld trajectory. Teach the welding torch posture of the welding robot, and the welding posture should ensure that it does not interfere with the workpiece.

[0053] As Figure 3As shown, a schematic diagram of the placement relationship between the transformer corrugated oil tank and the welding robot is shown. The welding robot is a common six-axis multi-joint robot 6. The laser sensor 5 is set on the welding gun on the six-axis multi-joint robot 6. The user coordinate system 7 is set with a certain point on the corrugated oil tank frame as the origin. The X-axis direction of the user coordinate system is the welding direction, and the Z-axis direction is perpendicular to the welding workpiece plane. The base coordinate system 8 is set with the center of the body base of the six-axis multi-joint robot 6 as the origin. Teach the robot scanning path SCAN1 in the user coordinate system. When running automatically, the robot will carry the laser sensor 5 to scan the weld along the taught scanning path SCAN1. The laser sensor transmits data to the robot in real time during the scanning process.

[0054] S3. Scan the weld features between the corrugated fuel tank and the frame through a line laser sensor, set a weld height threshold, output the weld point cloud data according to the weld height threshold, and transmit the weld point cloud data to the welding robot in real time.

[0055] The line laser weld recognition technology is used to scan the weld features between the corrugated fuel tank and the frame through a line laser sensor. The line laser weld recognition technology is a high-precision and high-efficiency weld position recognition method, which is a mature technology that comes with various brands of sensors on the market. This technology uses a line laser to emit a thin laser beam and project it onto the weld to be identified. The laser beam interacts with the weld to form a laser line that reflects the outline of the weld. Using image processing algorithms, the characteristic points of the laser line, such as edges, inflection points, etc., are accurately extracted. These characteristic points contain the precise position information of the weld. According to the extracted characteristic points, the system can identify the specific position of the weld and output the corresponding coordinate information. This information can be used to guide the motion trajectory of the welding robot to achieve precise welding. The point cloud data of the entire weld between the corrugated fuel tank and the frame is a discrete data set composed of a series of three-dimensional coordinate points. Each coordinate point includes the weld coordinates in the three directions of X, Y, and Z. The weld coordinates can be used to describe the geometric features of the weld, such as the surface shape and spatial position. The line laser sensor is used to scan the surface of the object, and a large number of coordinate points on the surface of the object are obtained to form point cloud data.

[0056] Specifically, during the laser line scanning process, the lap weld height h at different positions is obtained by scanning with a laser sensor. When the laser line is irradiated at weld positions with different gaps, the lap weld height h value will change. The weld height threshold is preset to h. max , the lap weld height h at different positions and the weld height threshold h max Compare, when h<h max When h>h max When the laser sensor does not output position data, there will be large intervals in the point cloud data obtained by the welding robot. The large gaps in the weld will be judged based on the large intervals in the point cloud data.

[0057] As shown Figure 4-5 in the figure, as Figure 4 shown, a schematic diagram of typical image features when a line laser irradiates on a lap joint seam, as Figure 5 shown, a schematic diagram of the cross-section of the lap joint seam. In the practical application of this embodiment, when the laser irradiates at the lower opening position of the corrugated fin of the corrugated fuel tank, the light beam is not strictly parallel to the plate surface of the corrugated fin of the corrugated fuel tank and has a certain incident angle. When the laser scans to different positions, L1 is not a single straight line segment, and different spot images will be obtained. Figure 4 The spot image in (1) in Figure 4 is the spot image obtained when the laser scans to the spot position 1, Figure 4 The spot image in (3) in

[0058] is the spot image obtained when the laser scans to the spot position 3. It is set that the weld feature of the laser sensor is lap joint, L1 is the upper surface spot of the lap weld, L2 is the lower surface spot, L1 may be a broken line, and the identified weld feature point is the foot of the perpendicular drawn from the end point of the straight line or broken line L1 to the straight line L2. h represents the height of the lap weld. Figure 5 Specifically, the height h of the lap weld is equal to the sum of the thickness t1 of the corrugated fuel tank plate and the joint gap b, that is, the height h of the lap weld = t1 + b,

[0059] S4. Divide the point cloud data of the weld into multiple segments of weld point cloud fragments, set the interval threshold of the weld point cloud fragments, and define the large interval between the weld point cloud fragments larger than the interval threshold as the large weld gap.

[0060] As Figure 6 shown, a schematic diagram of the point cloud data, and the set interval threshold D min should be slightly smaller than the span of the herringbone opening at the lower end of the corrugated fin. Compare the gaps D1, D2, D3... of the point cloud data with the point cloud fragment interval threshold D min to obtain the position of the large weld gap. The large weld gap includes the large gap at the lower opening of the corrugated fin and the large gap in the flat area between the two corrugated fins. The herringbone large gap at the lower opening of the corrugated fin is inherent and inevitable, while the large gap in the flat area between the two corrugated fins is caused by accidental incoming material errors.

[0061] Furthermore, the Random Sample Consensus (RANSAC) algorithm is used to further segment each piece of weld seam point cloud, removing the outlier points on the point cloud segment. Each piece of weld seam point cloud segment is processed independently to remove outlier points, so as to improve the accuracy and stability of weld seam recognition. The Random Sample Consensus (RANSAC) algorithm is an iterative method for estimating the parameters of a mathematical model from a set of data containing outliers. In point cloud processing, the RANSAC algorithm is used to further segment the point cloud segment and remove outlier points. The specific steps are as follows:

[0062] S01. Random sampling: Randomly select a set of data points with the minimum number from the weld seam point cloud data as samples;

[0063] S02. Model estimation: Use the selected samples to estimate the weld seam model;

[0064] S03. Consensus set calculation: Calculate the error between all data points and the weld seam model estimated by the samples, and add the data points within the set threshold of the error to the consensus set;

[0065] S04. Iterative optimization: Repeat steps S01 - S03 multiple times, and each time try to find a larger consensus set;

[0066] S05. Best model selection: Select the model with the largest consensus set as the best weld seam model;

[0067] S06. Outlier removal: Use the best weld seam model to regard all points that do not conform to the model as outlier points and remove them from the data.

[0068] After removing the outlier points, a relatively pure weld seam point cloud data set is obtained. This method is important for improving the accuracy of the weld seam point cloud data and the reliability of subsequent processing steps.

[0069] S5. Obtain the position data of the large gap of the weld seam according to the position data of the weld seam point cloud segment, and plan the reciprocating weaving welding trajectory of the weld seam according to the position data of the large gap of the weld seam.

[0070] Specifically, according to the position data of the lower opening of the corrugated fin of the corrugated fuel tank and the position data of the large gap of the weld seam, for the large gap area where the weld seam height exceeds h max and the length exceeds D min carry out weaving welding planning, which can ensure that the large gap at the lower opening of the corrugated fin can be recognized necessarily. Plan the reciprocating weaving welding trajectory of the weld seam at the large gap of the point cloud segment. The height of the weld bead is increased by reciprocating welding, and the bridging ability of the molten droplet molten pool is increased by raising, so as to fully adapt to large gap welding and ensure the welding quality.

[0071] Such as Figure 7As shown in the figure, it is a schematic diagram of the swing welding trajectory planning. The swing welding trajectory includes the actual trajectory, which includes the swing welding starting point, the swing welding highest point, the swing welding last point, and the swing welding ending point. The swing welding starting point and the swing welding ending point coincide. The trajectory between the swing welding starting point and the swing welding highest point is trajectory L1, the trajectory between the swing welding highest point and the swing welding last point is trajectory L2, and the trajectory between the swing welding last point and the swing welding ending point is trajectory L3. The order of the swing welding trajectory is from L1 → L2 → L3. The swing welding trajectory is an isosceles triangle, and the actual trajectories L1 and L2 are of the same length.

[0072] As Figure 8 shown in the figure, it is a schematic diagram of the swing welding trajectory planning points. According to the position data of the weld seam point cloud segment, the position data of the large gap in the weld seam is obtained. According to the position data of the large gap in the weld seam, the reciprocating swing welding trajectory of the weld seam is planned, specifically including:

[0073] S51. Calculate the coordinates of the gap center position point according to the position coordinates of the two adjacent front and rear point cloud segments of the large gap in the weld seam.

[0074] Specifically, calculate the coordinates of the gap center position point M(x M , y M , z M ) according to the end point position coordinates P1(x1, y1, z1) in the point cloud segment 1 and the start point position coordinates P2(x2, y2, z2) in the point cloud segment 2, where:

[0075] x M = (x1 + x2) / 2, y M = (y1 + y2) / 2, z M = (z1 + z2) / 2;

[0076] S52. Search forward a preset distance from the gap center position point to obtain the feature point on the front point cloud segment, and use the feature point on the front point cloud segment as the swing welding starting point. The preset distance is the swing amplitude.

[0077] Specifically, search forward a distance d2 from the gap center point M to obtain the feature point A'(x' A , y' A , z' A ) of the point cloud segment 2 as the swing welding starting point, and d2 is the swing amplitude. The significance of setting d2 is to ensure that the swing amplitude completely covers the length of the gap at the lower mouth of the corrugated fins of the corrugated fuel tank.

[0078] S53. Search backward a preset distance from the gap center position point to obtain the feature point on the rear point cloud segment, and use the feature point on the rear point cloud segment as the swing welding last point. The preset distance is the swing amplitude.

[0079] Search backward a distance d2 from the gap center point M to obtain the feature point C'(x' of the point cloud segment 1C , y' C , z' C ), as the last point of the weaving weld, d2 is the weaving amplitude; since the point cloud data fluctuates, especially P1 and P2 near the gap. Therefore, M(x M , y M , z M ) cannot be used as the midpoint position of the weaving weld, and it is necessary to take the more stable A' and C' of the point cloud data to calculate the midpoint of the weaving weld. B'(x' B , y' B , z' B ) has a certain distance from M(x M , y M , z M ).

[0080] S54. Calculate the coordinates of the midpoint of the weaving weld according to the coordinates of the starting point and the last point of the weaving weld.

[0081] According to the starting point A' and the ending point C' of the weaving weld, calculate the midpoint B'(x' B , y' B , z' B ) of the weaving weld:[[]]

[0082] x' B = (x' A + x' C ) / 2, y' B = (y' A + y' C ) / 2, z' B = (z' A + z' C ) / 2.

[0083] S55. Set the weaving height according to the welding requirements so that the molten pool can fill the large gap, set the weaving offset in the X-axis direction of the user coordinate system, represent the starting point coordinates, the highest point coordinates, and the ending point coordinates of the weaving weld after calculating and compensating the offset respectively according to the weaving height and the weaving offset, and plan and generate the reciprocating weaving trajectory of the weld according to the starting point coordinates, the highest point coordinates, and the ending point coordinates of the weaving weld after compensating the offset. The weaving offset offse is used to compensate for the position offset caused by the system error.

[0084] Specifically, step S55 specifically includes the following steps:

[0085] S551. Establish a user coordinate system and a base coordinate system. The X-axis direction of the user coordinate system is the welding direction, and the Z-axis direction is perpendicular to the plane of the welded workpiece (the placement plane of the corrugated oil tank of the oil-immersed transformer). The representation of the starting point coordinates of the weaving weld of the starting point A' in the user coordinate system is A user, calculate the coordinates of the starting point of the swing welding after compensating for the offset according to the swing offset offset, denoted as A′ user . The user coordinate system and the base coordinate system are as Figure 3 shown.

[0086] The starting point A’(x′ A ,y′ A ,z′ A ) in the user coordinate system is represented as A user (x Au ,y Au ,z Au ). Calculate A′ user (x′ Au ,y Au ,z Au ) after compensating for the offset according to the swing offset offset, where x′ Au = x Au + offset.

[0087] S552. Calculate the coordinates of the highest point of the swing welding after compensating for the offset according to the swing height w and the swing offset offset.

[0088] Specifically, the midpoint B’(x′ B ,y′ B ,z′ B ) in the user coordinate system is represented as: B user (x Bu ,y Bu ,z Bu ). Calculate the coordinates of the highest point of the swing welding after compensating for the offset according to the swing height w and the swing offset offset, denoted as B′ user (x′ Bu ,y Bu ,z′ Bu ), where z′ Bu = z Bu + w, x′ Bu = x Bu + offset.

[0089] S553. Calculate the coordinates of the last point of the swing welding after compensating for the offset according to the swing offset offset, denoted as C′ user .

[0090] The last point C’(x′ C ,y′ C ,z′ C ) in the user coordinate system is represented as C user (x Cu ,y Cu ,z Cu ). Calculate C′user (x′ Cu , y Cu , z Cu ), where x′ Cu = x Cu + offset;

[0091] S554. Respectively convert the converted representations of the starting point coordinates of the weaving weld, the highest point coordinates of the weaving weld, and the ending point coordinates of the weaving weld after calculating the compensation offset back to the representations of the starting point coordinates of the weaving weld, the highest point coordinates of the weaving weld, and the last point coordinates of the weaving weld in the base coordinate system. Plan and generate the reciprocating weaving weld trajectory of the weld seam according to the representations of the starting point coordinates of the weaving weld, the highest point coordinates of the weaving weld, and the last point coordinates of the weaving weld in the base coordinate system.

[0092] Specifically, respectively convert the converted representation A′ user of the starting point coordinates of the weaving weld, the converted representation B′ user of the highest point coordinates of the weaving weld, and the converted representation C′ user of the last point coordinates of the weaving weld back to the representation A of the starting point coordinates of the weaving weld, the representation of the highest point coordinates of the weaving weld, and the representation C of the last point coordinates of the weaving weld in the base coordinate system. Convert A′ user (x′ Au , y Au , z Au ), B′ user (x′ Bu , y Bu , z′ Bu ) and C′ user (x′ Cu , y Cu , z Cu ) back to the coordinate representations A(x A , y A , z A ), B(x B , y B , z B ) and C(x C , y C , z C ) in the base coordinate system. Plan and generate the reciprocating weaving weld trajectory of the weld seam according to the coordinate representations A(x A , y A , z A ), B(x B , y B , z B ) and C(x C , y C , z C ).

[0093] Step 6: Generate the weld path of the welding robot according to the planned reciprocating swing welding path and the point cloud segment data, and the welding robot welds the welds of the corrugated tank of the oil-immersed transformer according to the weld path.

[0094] As Figure 9 shown, in the schematic diagram of the reciprocating swing welding path, insert the swing planning points A, B, and C into the welding path to generate the final welding path. The order of generating the final swing welding path is

[0095] P1→A→B→C→P2;.

[0096] The welding robot welds according to the path generated by the point cloud processing. At the position without large gaps, it welds along the straight-line path according to the point cloud segment data. At the position with large gaps or the lower mouth of the corrugated fins, it performs swing welding according to the planned swing welding path.

[0097] A method for identifying and path planning of the welds of the corrugated tank of an oil-immersed transformer provided by the present invention combines laser weld identification, point cloud data processing, and swing welding path planning. It scans the weld features between the corrugated tank and the frame through a line laser sensor, sets a weld height threshold, outputs the point cloud data of the weld according to the weld height threshold, divides the point cloud data of the weld into multiple point cloud segments, sets an interval threshold for the point cloud segments, defines the gap in the point cloud data greater than the interval threshold as a large weld gap, obtains the position data of the large weld gap, and plans the reciprocating swing welding path of the weld according to the position data of the large weld gap. It can solve the problems of large recognition errors at the positions of the corrugated fins of the radiator, improper handling of the positions of large weld gaps, and unreasonable swing welding path planning in the prior art, effectively ensuring the welding quality and weld airtightness of the transformer corrugated tank, and is of great significance for improving the manufacturing quality and welding automation level of oil-immersed transformers.

[0098] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A method for identifying and trajectories of weld seams in a corrugated oil tank of an oil-immersed transformer, characterized in that: The following steps are involved: S1. Calibrate the line laser sensor, install the line laser sensor on the welding gun of the welding robot, and adjust the installation position and height of the line laser sensor; S2, the scanning path of the teaching line laser sensor and the welding gun posture of the welding robot; S3, scanning the weld features between the corrugated fuel tank and the frame through a line laser sensor, setting a weld height threshold, outputting the weld point cloud data according to the weld height threshold, and transmitting the weld point cloud data to the welding robot in real time; S4, dividing the point cloud data of the weld into multiple point cloud segments, setting an interval threshold of the point cloud segments, and defining the interval of the point cloud data greater than the interval threshold as a large gap of the weld; S5. Acquire the position data of the large gap of the weld according to the position data of the weld point cloud fragment, and plan the reciprocating welding trajectory of the weld according to the position data of the large gap of the weld; The planning of the reciprocating welding trajectory of the weld according to the position data of the large gap of the weld specifically includes: S51, calculating the coordinates of the center position point of the gap according to the position coordinates of the two adjacent point cloud fragments in front and behind the large gap of the weld; S52, starting from the gap center point and searching forward for a preset distance to obtain a feature point on the previous point cloud segment, and using the feature point on the previous point cloud segment as the starting point of the swing welding; S53, starting from the gap center position point and searching backward for a preset distance to obtain feature points of the rear point cloud fragment, and taking the feature points of the rear point cloud fragment as the last point of the pendulum welding; S54, calculating the coordinates of the midpoint of the weaving welding according to the coordinates of the starting point of the weaving welding and the coordinates of the last point of the weaving welding; S55, according to the welding requirements, the swing height is set so that the molten pool can fill the large gap, the swing offset in the X-axis direction of the user coordinate system is set, and the swing welding starting point coordinate representation, the swing welding highest point coordinate representation, and the swing welding end point coordinate representation after the offset compensation are calculated according to the swing height and the swing offset, and the round-trip swing welding trajectory of the weld is planned and generated according to the swing welding starting point coordinate representation, the swing welding highest point coordinate representation, and the swing welding end point coordinate representation after the offset compensation; S6. Generate a weld trajectory of the welding robot according to the planned reciprocating welding trajectory and the point cloud fragment data, and the welding robot welds the weld of the corrugated oil tank of the oil-immersed transformer according to the weld trajectory.

2. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 1 is characterized in that: Teaching the scanning path of the line laser sensor and the welding gun posture of the welding robot includes: setting the user coordinate system with a certain point on the corrugated oil tank frame as the origin, setting the base coordinate system with the center of the welding robot's body base as the origin, teaching the robot scanning path in the user coordinate system, placing the scanning start position in front of the weld start end, and the scanning end position in back of the weld end.

3. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 1 is characterized in that: The step S3 comprises the steps of: Line laser weld recognition technology is used to scan the weld features between the corrugated fuel tank and the frame through a line laser sensor; The lap weld height h at different positions is obtained by scanning with a laser sensor, and the weld height threshold is preset as h max , when h<h max When h>h max , the laser sensor does not output position data.

4. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 3 is characterized in that: The lap weld height h is equal to the sum of the thickness of the corrugated oil tank plate and the joint gap.

5. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 1 is characterized in that: The step S4 includes: using a random sampling consensus algorithm to further segment each weld point cloud segment and remove outliers on the weld point cloud segment.

6. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 5 is characterized in that: The method of further segmenting each weld point cloud segment using a random sampling consensus algorithm to remove outliers on the weld point cloud segment includes the following steps: S01, random sampling, randomly select a minimum number of data points from the weld point cloud data as samples; S02, model estimation, estimating the weld model using the selected sample; S03, consensus set calculation, calculating the errors between all data points and the sample estimated weld model, and adding the data points whose errors are within the set threshold to the consensus set; S04, iterative optimization, repeating steps S01-S03 multiple times, each time trying to find a larger consensus set; S05, optimal model selection, selecting the model with the largest consensus set as the optimal weld model; S06, outlier removal, using the optimal weld model, all points that do not conform to the model are considered as outliers and removed from the data.

7. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 1 is characterized in that: The step S55 comprises: Establish a user coordinate system and a base coordinate system. The X-axis direction of the user coordinate system is the welding direction, and the Z-axis direction of the user coordinate system is perpendicular to the plane of the welding workpiece. The coordinate representation of the starting point of the welding in the user coordinate system and the coordinate representation of the starting point of the welding after the swing offset calculation and compensation offset are used; The coordinate representation of the highest point of the weaving welding after the offset compensation is calculated based on the coordinate representation of the midpoint of the weaving welding in the user coordinate system, the weaving height and the weaving offset; The coordinate representation of the last point of the oscillating welding is based on the coordinate representation of the last point of the oscillating welding in the user coordinate system and the coordinate representation of the last point of the oscillating welding after the oscillating offset is calculated and compensated; The coordinate representation of the swing welding starting point, the coordinate representation of the highest point and the coordinate representation of the swing welding end point after calculating the compensation offset are converted back to the coordinate representation of the swing welding starting point, the coordinate representation of the highest point and the coordinate representation of the swing welding end point in the base coordinate system, and the round-trip swing welding trajectory of the weld is planned and generated according to the coordinate representation of the swing welding starting point, the coordinate representation of the highest point and the coordinate representation of the swing welding end point in the base coordinate system.

8. The method for identifying and trajectories of weld seams of corrugated oil tanks of oil-immersed transformers according to claim 1 is characterized in that: According to the coordinate representation of the midpoint of the weaving welding in the user coordinate system, the weaving height and the weaving offset, the coordinate representation of the highest point of the weaving welding after compensating the offset is calculated, including: The coordinate of the midpoint of the weaving welding in the user coordinate system is expressed as B user (x Bu ,y Bu ,z Bu ), the coordinate of the highest point of the weaving welding is expressed as B′ user (x′ Bu ,y Bu ,z′ Bu ), then z′ Bu =z Bu +w,x′ Bu =x Bu +offset; where w is the swing height and offset is the swing offset.

Images