Study On Seam Tracking System Based On Stripe Type Laser Sensor And Welding Robot

Robots have been widely used in modern industry, but the main application is still using the "Teach- reproduction" mode. In the field of application of welding, due to the spot welding positioning process and welding heat, the deformation may result in robot welding track teaching programming deviating from the trajectory of the reproduction and affecting the quality of welding. Thus, in this paper, described the two-dimensional striped seam tracking system based on laser sensor, which can observe the changes in the welding process in real-time using the laser sensor, calculate the center-point of the weld and then transmit these messages to the robot and then drive the torch to the center-point of the weld. The research project won the national science and technology major special projects(2015ZX04005006) and Guangdong Province emerging industries of strategic core technology research projects(2011A091101001) funding. The main contents of this paper are as follows:(1) Discussed the transformation between the coordinate system of industrial robots and lay a solid theoretical foundation for the rotation transformation calibration feature points and weld laser measurement coordinate system.(2) Proposed the single-point calibration algorithm and multi-point calibration algorithm of the stripe laser sensor to calculate the measurement coordinate system transformation between the coordinate system and the end coordinate system of the robot.(3) Proposed the feature point detection algorithm based on the Wavelet Transform Modulus Maxima Theory and the Least Square Method. Used the wavelet function to position the feature point preliminary, used the result to packet the laser data and then used the least square method to fitting the weld line to position the feature point precisely.(4) Designed and developed seam tracking system based on a striped laser sensor and a welding robot. Developed the PC software based on Microsoft Visual Studio 2010, mainly to complete the two-dimensional stripe laser sensor data acquisition, weld feature point detection algorithms and interactive features. Some experimental studies were carried out on this platform for real-time seam tracking, respectively, feature point extraction experiments and weld seam tracking experiments. Experimental results show the feasibility and effectiveness of the algorithm and theory.The creative results of this paper are as follows:(1) Proposed the single-point calibration algorithm and multi-point calibration algorithm of the stripe laser sensor. The single-point calibration algorithm was proposed according to the mounting of the laser sensor. Ignored the mounting, proposed the multi-point calibration algorithm, including three-points calibration algorithm and five-points calibration algorithm. Experimental results verified the calibration error is less than 0.5mm.(2) Proposed the feature point detection algorithm based on the Wavelet Transform Modulus Maxima Theory and the Least Square Method. Used the wavelet function to position the feature point preliminary, used the result to packet the laser data and then used the least square method to fitting the weld line to position the feature point precisely. Experimental results verified the error is less than 0.4mm.