Research And Development Of Robotic Welding Expert System For Backing TIG Of Pipeline

Welded pipelines have the advantages of good sealing and high safety,and are widely used in petrochemical and other fields.For thick-walled pipelines,backing welding seams are the basis for filler and cover welding,and are also the core of ensuring joint quality and improving welding efficiency.Compared to manual welding,the use of robots for backing TIG results in more consistent welding quality.The design of the welding process based on understanding of the base material and the weld material is the key to ensuring the quality of the backing welding seams.Therefore,this paper develops a robotic welding expert system for backing TIG of pipeline for the actual needs of the company.Firstly,the structure of the system was designed,divided into four modules: welding process design,welding database,knowledgebase management,and user management.The development was completed by combining the Java language,Java Swing tools,and My SQL database management system.The system knowledge was divided into factual,rule-based and case-based categories,for which the design of knowledge acquisition methods of semi-automatic,combined manual and semi-automatic,combined automatic and semi-automatic,and knowledge representation methods of generative representation combined with predicative verb representation,respectively,completed the design of the knowledgebase.Based on the welding process design for backing welding,the inference method was a combination of case-based and rule-based,the inference direction was positive,and the conflict resolution strategy was a combination of the matching strategy and the firstness strategy,completed the design of the reasoning machines.Secondly,in accordance with welding industry standards,using the My SQL database management system which is based on the Data Grip,a database of base materials and welding materials was established for backing TIG of pipeline,storing 620 data records involving stainless steel,carbon steel,low alloy steel,aluminium alloy and other materials.JDBC was used to complete the interaction between the Java and the database to develop the welding database module of the expert system,which operates the base material and weld material data records through the expert system program.Then,according to the experience of welding experts and supplementary tests,fitting analysis based on MATLAB was carried out on welding current,welding speed and wire feed speed for backing TIG of pipeline,to obtain its optimal function model and calculation equations,to lay the foundation for the development of the welding process design module.Based on the thickness of backing welding fitting,the thickness is the independent variable,ranging from 0.8 to 4 mm;welding current and welding speed are the dependent variables.Based on the welding current equation selected by the user,the system automatically gives the welding speed equation to match the welding current and welding speed.The optimal function model for welding current of stainless steel is a third-order polynomial,for welding speed is a sine function and a power function respectively;the optimal model for welding current of aluminium alloy is a second-order polynomial.Based on the bevel size fitting,the bevel gap is the independent variable,wire feed speed is the dependent variable.The optimal functional model for wire feed speed of stainless steel is a first order polynomial and a second order polynomial,for wire feed speed of low carbon steel is a second order polynomial.Finally,the process parameters design principles and process parameter equations were combined to form a complete welding process design flow.The complete welding process design can be obtained by multi-level reasoning and can be exported as a welding process card.At the same time,the knowledgebase management module and user management module have been developed with the aim of ensuring the safety and scalability of the welding expert system,resulting in a complete welding expert system that meets the requirements for backing TIG of pipeline using robotics.