SEAM TRACKING METHOD OF WELDING ROBOT BASED ON PROBABILITY MODEL CONSIDERED THE PROPERTY OF 'THROUGH-ARC' SENSOR

摘要

A method for tracking of welding seam through a probability model with reflected characteristics of an arc sensor is provided to improve reliability, welding quality and productivity accordingly by calculating a compensation distance using a welding current value having reflected characteristics of the arc sensor, thereby setting optimal welding conditions and automatically tracking a welding line. A method for tracking of welding seam through a probability model with reflected characteristics of an arc sensor comprises: a step(S1) in which a welding robot controller drives an arc sensor to move the arc sensor to a welding starting point; a step(S3) in which the welding robot controller performs a welding operation for obtaining test data to abstract current values per each measurement time; a step(S5) in which the welding robot controller calculates an average value and a deviation of data calculated using the abstracted current values; a step(S7) in which the welding robot controller calculates a normal distribution function using the calculated average value and the calculated deviation; a step(S9) in which the welding robot controller calculates a compensation distance through probability data, arc sensor distance values and current values by calculating a probability model; and a step(S11) in which the welding robot controller controls a welding robot according to the compensation distance.