Research On Layout Optimization Of Welding Spot For Automobile Sheet Metal Assemble

Along with the rapid development of automobile industry in recent years, China has become the world’s largest automobile production and sales, consumers on car body appearance quality request is getting higher and higher with increasing demand for car. The appearance quality of the body depends on the dimension of body assembly, and welding layout strategy directly affect compliant sheet metal assembly dimension deviation, the optimization of welding layout scheme can effectively reduce the welding distortion of sheet metal assembly, and reduce the assembly dimension variations, in order to improve the assembly quality.Considering the problem which assembly deviation statistical analysis by the simulation of compliant sheet metal assembly process though the finite element software is huge time-consuming, the method of establishing sheet metal assembly variation model is proposed, the model combined with the Monte Carlo simulation, assembly deviation statistical analysis results, it described the relationship between assembly variations and part measure spot variations, welding spot positional variations, welding gun positional variations and fixture locating variations,4 variation make the model more complete and consistent with the actual production.Spot welding layout scheme optimization mostly uses the iterative method which is combined with finite element analysis and optimization algorithm, Considering this method is time-consuming and low efficiency, surrogate models for spot welding layout optimization is proposed, and three commonly used surrogate models are established and analysis, polynomial response surface methodology(RSM), genetic algorithm-BP neural network(GA-BP), particle swarm optimization(PSO)-support vector machine regression(PSO-SVR). In order to improve the efficiency and precision of model, Latin hypercube design(LHD) is proposed for experimental design method.In order to improve the optimization efficiency for spot welding layout, an optimization method of shaping welding spot unequal pitch layout for reducing assemble variation based on PSO-SVR is proposed. Firstly, Considering two pitches of welding spot as design parameters, sample schema of welding spot layout are obtained in design space of layout by optimized LHD, Then sample response is obtained by assembly variation model. Finally, PSO-SVR is established by sample data, the best shaping welding spot unequal pitch layout which minimizes the assembly variation of sheet metal can be found by prediction of PSO-SVR. A simple sheet metal assembly was used to indicate its effectiveness though prediction accuracy, and some comparisons with surrogate models of GA-BP neural network and RSM are conducted in the case. The result shows good prediction and calculation performance of the given method, and the practical of the method is indicated by an example of auto-body sheet metal assembly case.