Abstract:A neural network based response surface method was proposed in order to overcome the limitation of quadratic response surface in solving non-linear problems.A 2D two-step axisymmetric forging problem as an example was studied with the proposed method.Relationship between preform shape and filling ratio was established and optimum was achieved by using PSO optimization method.Meanwhile,with respect to the complexity of the optimum solution,the knowledge based redesign idea was proposed to obtain a simple one.Comparison was done between the original optimum shape and the redesign one,which shows better result is achieved by the using new proposed method.

The Optimum Preform Design and Redesign Based on Neural Network Response Surface Methodology