CN112222622A

CN112222622A - A real-time zoom laser welding method

Info

Publication number: CN112222622A
Application number: CN202011065391.7A
Authority: CN
Inventors: 史建平; 晏刘武
Original assignee: Changzhou Institute of Technology
Current assignee: Changzhou Institute of Technology
Priority date: 2020-09-30
Filing date: 2020-09-30
Publication date: 2021-01-15

Abstract

The invention discloses a real-time zooming laser welding method, including: the mechanical mechanism design of the welding head; the design of the radar sensor welding surface measurement system; , the image is processed, the image is thresholded, the external factor interference during laser welding is removed, only the data value of the plane is adopted, the connectivity analysis is performed on the thresholded area, and the area with a small area is removed, and the final result is obtained. The average value is processed, the differential comparison is performed, and the optimal height value is output. The invention ensures the focal length of welding, detects the welding surface in real time, adjusts the focusing lens inside the laser head, has fast response speed, reduces the vibration of the machine, improves the consistency of welding, and reduces the scrap rate of products.

Description

Real-time zooming laser welding method

Technical Field

The invention belongs to the field of laser welding, and particularly relates to a real-time zooming laser welding method.

Background

When a product is welded by laser at present, if the welding focal length is changed under the condition of a certain welding process, poor phenomena such as insufficient solder can be directly caused. The traditional method for solving the problem is to adopt point laser ranging, measure the height before welding, move the Z-axis height of a laser welding head and realize height deviation compensation. Since only one point is measured, if the bonding surface is not flat, poor bonding may also be caused due to the non-uniform height. Meanwhile, the measurement system and the welding system are mutually independent, so that the rapid integration of an integrator is not facilitated, and the periodicity and the uncontrollable property of the project are increased. Limited by the influence of the laser focal length, and easily causes the false welding of the product and poor welding when the height of the product plane changes. The equipment board that the laser head removal caused rocks, has increased the defective rate of welding equally.

Disclosure of Invention

The invention mainly solves the problems that when the welding surface is not flat, the internal system of the welding head is adjusted under the condition that the laser welding head is not moved, the zoom welding is realized, the real-time response is fast, and the vibration of a machine table is reduced. Meanwhile, the welding and height measuring system is combined to stably operate, so that the research and development period of an integrator is reduced, and the integration level of the system is increased.

In order to achieve the purpose, the invention provides the following technical scheme:

a real-time zoom laser welding method, comprising:

step 1, designing a welding head mechanical mechanism: a focusing lens with adjustable vertical height is arranged on the welding head;

step 2, designing a radar sensor welding surface measuring system: adding an ultrasonic measuring device outside the laser welding head to measure the lower welding surface, calibrating and zeroing with the servo position, calibrating the radar sensor measurement value and the servo up-down adjustment value to obtain the linear relation of the radar sensor measurement value and the servo up-down adjustment value;

step 3, designing a radar sensor measuring system: processing the acquired three-dimensional point cloud data by adopting an upper computer design mode, processing an image, segmenting an image threshold, removing external factor interference during laser welding, only adopting a planar data value, performing connectivity analysis on the region after threshold processing, removing a region with a smaller area, performing mean value processing, differential comparison and outputting an optimal height value on the finally obtained region;

step 4, the height value is substituted into a linear relation obtained in calibration to obtain the absolute position of movement, and the absolute position is sent to a servo driver to be executed;

and 5, circularly executing the steps at a high speed, comparing the height difference of the movement, and if the height difference is smaller or larger, abandoning the movement.

Furthermore, in step 2, the calibration is performed by a five-point method, that is, the radar sensor measures five sets of data, and the servo linkage device moves five times to obtain a linear relationship between the two.

Further, in step 3, the processing step of the collected three-dimensional point cloud data is as follows: converting XYZ numerical values of the three-dimensional point cloud data into a height map, wherein the conversion principle is that Z-axis data is linearly converted into 16-bit data and corresponds to the brightness value of a 16-bit map; the XY direction point cloud data correspond to pixel coordinates, and the spare pixel positions adopt an interpolation algorithm and add the brightness value of the field, so that the three-dimensional data are converted into a two-dimensional space for processing.

Furthermore, in step 3, the processing of the image includes preprocessing the image and removing the interference points by using gaussian filtering.

Furthermore, in step 3, the height value information of the plane is automatically segmented by adopting a dynamic threshold processing method for threshold segmentation.

Compared with the prior art, the invention has the beneficial effects that:

the invention ensures the focal length of welding, detects the welding surface in real time, adjusts the focusing lens in the laser head, has high response speed, reduces the vibration of a machine table, improves the welding consistency and reduces the rejection rate of products.

Drawings

FIG. 1 is a flow chart of a method of the present invention;

fig. 2 is a schematic view of the bond head mechanical mechanism.

The labels in the figure are: 1. a servo motor; 2. a radar 3D sensor; 3. a linkage device; 4. a focusing lens; 5. the plane was measured.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings.

As shown in fig. 1, a real-time variable-focus laser welding method according to the present invention includes:

firstly, designing a mechanical mechanism of a welding head. As shown in fig. 2, the bonding head mechanical mechanism includes a servo motor 1, a radar 3D sensor 2, a linkage device 3, a focusing lens 4, and a measuring surface 5. Compared with the traditional welding head, the mechanism with the adjustable height of the focusing lens is added, the mechanism adopts a high-precision linkage mechanism, and the servo motor drives the focusing lens to adjust up and down. Meanwhile, a limiting device and an original point device are added, and the position controllability in the system operation process is guaranteed.

And secondly, a radar sensor welding surface measuring system. And adding an ultrasonic measuring device outside the laser welding head to measure the lower welding surface. While at the same time aligning zero with the servo position. And calibrating the measured value of the radar sensor and the servo upper and lower adjustment values by adopting a five-point method, namely measuring five groups of data by the radar sensor, and moving the servo linkage device five times to obtain the linear relation between the measured value and the servo upper and lower adjustment values.

Thirdly, designing a radar sensor measuring system, processing the collected three-dimensional point cloud data in an upper computer design mode, and comprising the following steps: the XYZ values of the three-dimensional point cloud data are converted into a height map by linear conversion of the Z-axis data into 16-bit data, for example, the height range bits 0 to 100, and correspondingly linear conversion into 0 to 65535, which corresponds exactly to the brightness value of the 16-bit bitmap. And the XY direction point cloud data corresponds to pixel coordinates, and the spare pixel positions adopt an interpolation algorithm to add the brightness values of the fields. Thus, three-dimensional data is converted into two-dimensional space for processing. The image is processed. Firstly, preprocessing an image, and removing interference points by adopting Gaussian filtering. And (3) segmenting an image threshold, removing external factor interference such as splashing during laser welding and the like, and only adopting a plane data value. The threshold segmentation algorithm can adopt a dynamic threshold processing method, similar to auto _ threshold operator in Halcon, and can automatically segment the height value information of the plane. And performing connectivity analysis on the region subjected to threshold processing, removing the region with a smaller area, performing mean processing, differential comparison on the finally obtained region, and outputting an optimal height value.

And fourthly, the height value is brought into a linear relation obtained in calibration to obtain the absolute position of movement, and the absolute position is directly sent to a servo driver through a network port to be executed, so that the real-time performance of the system is improved.

And fifthly, the process is circularly executed at high speed, the height difference of the movement is compared, if the height difference is smaller or larger, the movement is abandoned, and the jitter of the system is reduced.

In conclusion, the invention adopts the real-time zoom welding head to adjust the focal length in real time and weld, and the integrator can only use the real-time zoom welding head without knowing the implementation principle, thereby accelerating the research and development period of the project and the stability of the equipment. The welding head is zoomed in real time, the welding surface at the lower part of the laser head is measured through the radar sensor, the collection of point cloud data is realized, and a source is provided for data analysis. Meanwhile, the three-dimensional data measured by the radar sensor is stable and reliable, and the welding surface of the lower part of the laser head can be measured and analyzed during laser welding. After data analysis, the height of the focusing lens is adjusted through servo to realize focal length zooming, the mode has high response speed, vibration of a machine table is reduced, and welding effect and yield are guaranteed. The laser welding head with real-time zooming continuously adjusts the focal length, and ensures the welding consistency.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the present invention. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. a laser welding method of real-time zoom, is characterized in that, comprises:

Step 1. Design of the mechanical mechanism of the welding head: install a focusing lens with adjustable height up and down on the welding head;

Step 2. Design of the radar sensor welding surface measurement system: add an ultrasonic measuring device to the outside of the laser welding head to measure the lower welding surface, at the same time calibrate and zero with the servo position, and calibrate the radar sensor measurement value and the servo upper and lower adjustment values to obtain two the linear relationship of

Step 3. Design of radar sensor measurement system: adopt the design method of the upper computer, process the collected 3D point cloud data, process the image, segment the image by threshold, remove the interference of external factors during laser welding, and only adopt the data value of the plane , perform connectivity analysis on the thresholded area, remove the area with a smaller area, perform mean processing, differential comparison on the final obtained area, and output the optimal height value;

Step 4. Bring the height value into the linear relationship obtained during calibration, obtain the absolute position of the movement, and send it to the servo driver for execution;

Step 5. Perform the above steps in a high-speed loop, compare the height difference of the movement, if it is smaller or larger, then give up the movement.

2. the laser welding method of a kind of real-time zooming according to claim 1, is characterized in that, in step 2, described calibration adopts five-point method to demarcate, namely radar sensor measures five groups of data, and servo linkage device moves five times , to obtain a linear relationship between the two.

3. a kind of laser welding method of real-time zooming according to claim 1, is characterized in that, in step 3, the processing step to the collected three-dimensional point cloud data is: three-dimensional point cloud data XYZ numerical value is converted into height map, The principle of conversion is that the Z-axis data is linearly converted into 16-bit data, corresponding to the brightness value of the 16-bit image; the XY direction point cloud data corresponds to the pixel coordinates, and the remaining pixel positions use an interpolation algorithm to add the brightness value of the field, so that the three-dimensional The data is transformed into a two-dimensional space for processing.

4 . The laser welding method with real-time zooming according to claim 1 , wherein, in step 3, the processing of the image includes preprocessing the image, and using Gaussian filtering to remove interference points. 5 .

5 . The real-time zoom laser welding method according to claim 1 , wherein, in step 3, the threshold segmentation adopts a dynamic threshold processing method to automatically segment the height value information of the plane. 6 .