Design And Motion Control Of Intelligent Welding Robot Mechanism For Pipeline Welding Of Boiler Inner Wall

At present,there are a large number of boiler tube inner tube pipe welding in the domestic large waste heat boiler industry.In the welding process,there are many problems such as complicated weld position,variable welding gun posture,narrow welding working space,low welding frequency and so on.As a result,the degree of automation of pipe inner tube welding is low.Therefore,it is particularly urgent to improve the automation level of pipe inner tube welding.Difficulties in tube wall tube welding technology are as follows: 1.The formation of the relevant line weld between the tube and the cylinder is a complex space curve.It is difficult to ensure the welding speed and effect by hand welding.Second,it belongs to the full position with the largest difficulty coefficient in the welding field.Welding,the annular fillet weld formed between the convection pipe and the cylinder is closely arranged on the inner wall of the boiler barrel,especially in the annular fillet weld on the same workpiece,there are flat welding,vertical welding,overhead welding,and welding process.It needs to be adjusted in real time according to the actual weld space position and angle.3.At this stage,the robot welding is difficult to reach the level of the existing skilled welding workers.Based on the above-mentioned problems in the automation technology of tube inner tube tube welding,the main research contents of this paper are as follows:According to the distribution characteristics of the fillet weld,the design difficulties and key points of the inner ring angle weld robot of the drum are discussed in detail,and a set of mechanical structure schemes suitable for the ring fillet welded in this paper is designed.The kinematics model of the ring weld robot was established in the motion attitude and direction angle adjustment scheme,and kinematics analysis was carried out to provide a reliable theoretical model for subsequent welding gun identification and trajectory planning.In-depth study of the existing arc sensing method,proposed a new arc sensor for the fillet weld-electronically controlled swing arc sensor.A set of electronically controlled oscillating arc sensing system for ring weld seam tracking is established.The working principle of electronically controlled oscillating arc sensing is carefully analyzed.Based on adaptive filtering theory,a Mallat wavelet decomposition and reconstruction are selected.The linear Kalman noise reduction method combines the extraction and utilization of the arc sensing eigen-signal,which effectively solves the intrinsic high-dimensional problem in the process of welding gun attitude recognition.Finally,the purpose of identifying the torch posture is achieved.Aiming at the difficult problem of weld seam welding in the inner wall of the cylinder,the mathematical model of weld seam and the welding attitude model of welding torch are analyzed.A trajectory planning based on angle approximation algorithm is proposed.The simulation of the relevant line trajectory model is verified by Matlab software.Reliability,and proposed a B-DPPO trajectory algorithm based on reinforcement learning to optimize it.Experiments show that the above method can significantly improve the tracking accuracy of the fillet weld trajectory of the drum.The weld seam tracking system based on the electronically controlled oscillating electric sensing system is built.The experimental results show that the method can effectively improve the welding seam tracking accuracy of the inner wall of the boiler tube and provide a new space curve weld seam tracking.method.