| The degree of automation of welding plays an important role in the development of intelligent manufacturing,using welding seam tracking technology to carry out arc welding robot to complete welding automation is a hot research topic in recent years.Weld seam tracking trajectory detection process uses laser vision sensor and arc sensor most widely,laser vision sensor has large measuring range,simple structure,easy to install,but it is Complex when image processing,that reduce it`s real-time,besides the surface morphology of the sample to be tested has serious interference with image acquisition.The arc sensor has strong real-time performance,it has good accessibility and strong anti-arc ability,but its sampling range is small and it may largely be affected by weld bevel type.The welding seam tracking actuators are mostly three-axis robots,three-axis robots are inexpensive,it is easy to use and has good accessibility,but it is constrained by degrees of freedom,the automatic welding of complex weld trajectories in multiple positions cannot be achieved.This paper in order to solve the problems that laser vision sensor image processing difficulties and the single welding path of the three-axis welding robot,it is based on the laser displacement sensor equipped with a six-axis arc welding robot to complete the welding seam tracking,the specific research contents are as follows: It designs and selects the hardware composition of the intelligent robot welding system.The system uses Hampel algorithm to identify outliers.It performs the replacement of the outliers based on the Ljung L method.It uses MATLAB's integrated function smooth for smoothing the sampled data set after filtering.It is based on rigid body kinematics.It replaces the rigid body pose transformation process with a homogeneous array transformation process.It establishes a mathematical model of the arc length of the inflection point of the corrugated plate.It uses the D-H method to represent the 6-axis robot joint coordinate system while modeling a 6-axis robot model.This article builds on the above.It analyzes the forward and backward kinematics of a 6-axis robot.It has performed a solution operation.This paper establishes a mathematical model of the welding gun position correction after the laser displacement sensor scans the inflection point of the corrugated board on the basis of a 6-axis robot kinematics model.It deduces the matrix transformation process of the 6-axis robot welding gun coordinate system relative to the base coordinate system after tracking correction.It combines the characteristics of the inflection point of the corrugated board.It proposes a new Globally-oriented backward adaptive bat optimization algorithm(GBABOA)based on the traditional BA algorithm.It solves the model through LABVIEW combined with a new global-oriented backward adaptive algorithm.It produced a simulation diagram of the overall deviation of the weld tracking.It obtains the left and right deviation simulation diagrams of welding seam tracking and the high and low deviation simulation diagrams respectively,and forwardly projects the coordinate system plane through the overall deviation simulation diagram.The simulation results in this paper show that the lateral deviation of the adjacent two wave bands of the corrugated plate is ± 0.34 mm,and the longitudinal deviation is ± 0.16 mm.It also calculated the deflection angle of the corrugated plate from-1.15 ° to ? 0.92 °.For this system,the tracking experiments of the adjacent wave band welding track and linear welding track of corrugated board are designed,using the same test conditions to compare whether the system is operating normally,the test results show that the system can realize the automatic tracking welding test of the adjacent wave band of the corrugated board and the straight line weld,system operate stably and has high precision,besides the welding seam is well formed to meet the expected effect of the test. |
PDF Full Text Request