Special Welding Grinding Robot Control System

Waterpower is distributed all over the world, hydraulic turbines of power stations are subjected to destruction due to serious cavitation and wear, so that they need to be overhauled regularly to ensure normal running of the power stations. At present, the means for repairing hydraulic turbine blades in our country are still manual weld and manual polish. The working condition is bad and labour intensity great for workers and overhaul efficiency low and inferior. Cavitations in foreign hydraulic turbines appear in clear water condition, so its destructive extent is not far more serious than in Yellow River hi our country. Stronger stiffness and more agility are required for Special-purpose robot for hydraulic turbine repair used in our country. Foreign robot for hydraulic turbine repair could not satisfy our working require, so we develop special-purpose robot system that is used to polish and weld repair cavitate flow path surfaces of hydraulic turbines on all location.Based on Special-purpose robot for hydraulic turbine repair, the design of the welding orientation system used the fuzzy sets and fuzzy control principle, accomplish automatic orientation to the concavity for the special robot. The article introduced in detail the design of fuzzy controller for the first three joins of the special robot, the selecting of fuzzy variables and their fuzzification and defuzzification are introduced also .The approach is simple for real-time implementation.Aiming at robot for hydraulic turbine repair, we proposed a fuzzy adaptive PID controller based on the fuzzy sets for the end operating hand of robot to control location and pose. The controller use peoples' experience and knowledge to tune PID parameters in real time for improved control quality and adaptability.Using the means of vector accumulation we can extract J and J-1. Consequently thetrack in right-angle coordination could turn to the track in joint space coordinate. Thereby we can use the information of vision sensor efficaciously.Using Matlab and Fuzzytoolbox, though programming and establishing fuzzy model we accomplish dynamic simulation for the robot control system of location and pose. Simulation result verifies that this method has better adjusting precision and better rejection of disturbance.