Research And Development Of Process Design ES Of Resistance Spot Welding

This paper analyzes the research progress and development of welding expertsystem(ES) in domestic and overseas, then summarizes the problems of developed ESin domestic, which can not be widely applied to all material and structure. Theknowledge and standards of resistance spot welding process design were generalized,by the actual process design procedure of resistance spot welding from a aviationenterprise. This Process Design ES of Resistance Spot Welding was developed withthe technology of ES, object-oriented and database.The system can be used to much material in aviation industry, such as aluminumalloy, structural steel, stainless steel, superalloy, etc. The parameters determinationmethod of each process design tasks: material weldability, procedure qualificationrequirements, structure design, pretreatment method, welding electrodes choose andwelding parameters, were summarised by such expert experience, standard data anddesign rules in resistance spot welding process, which additionally provide theoryand data foundation for Knowledge Base and Inference Engine of expert system.The welding process Database E-R model was established by the E-R model ofmaterial and spot welding process design. Using modeling tool-Power Designer tomodeling Database of spot welding process design. Knowledge is expressed bygenerative knowledge presentation method, thus the Knowledge Base was built. Inrespect of Inference Engine, forward inference, certainty and uncertainty mixedreasoning method are adopted. Based on each spot welding process design tasks, theinference strategies are established, which curve fitting technology is creativelyutilized to improve the reasoning precision.Based on C/S structure, this resistance spot welding process design Expert systemwas developed with Visual C++6.0,combing the Oracle10g database managementsystem. This system process simple interface and convenient operation, it canintelligently match and design the process design parameters according to the users’requirements. In addition, the accuracy of the system inference was validated throughthe case research comparation, which shows that the process parameters results fromexpert system inference are reliable.