Optimization technology of sheet metal deep drawing with variable blank holder force

For the sake of higher quality in sheet metal deep drawing, a non-axial symmetry part-automobile headlight reflector is taken as the research object. The objective function is obtained by approaching function of artificial neural network (ANN). Numerical simulation and multi-object genetic algorithms are adequately linked, whereafter, optimization model of variable blank holder force (VBHF) is established. At last, the reasonable VBHF curve is obtained. The result indicates that the VBHF curve optimized tallies with the tendency of plastic deformation in deep-draw better than traditional constant BHF and makes for even deformation and high quality product.