Study On U-shaped Pieces Of Springback Forecast Based On ANN

Generally speaking, springback problems have always existed in sheet metal forming process. Especially during the course of sheet metal bending and low-drawing, this phenomenon is obviously serious which greatly affect precision of parts and manufacture efficiency. So it is necessary to do further research to control the spingback. There is not much research work of spingback control formerly, however, and experience as well as trial-and-error method is used to reduce or eliminate the spingback in engineering practice. But since the 90th of last century, with the gradual solution of wrinkling and cracking in the course of drawing, spingback problem has daily risen to an important research project. In addition, ceaseless perfection of CAE simulation technology of sheet-metal pressing also provides a necessary foundation for research of springback control.Leading artificial intelligence (AI) technology and method into pressing process is a study hotspot on pressing process field at present. Artificial neural network (ANN) is one kind of AI method which is stood up by imitating person's brain nerve delivering information. It is one kind of distributed parallel processing system, the acquired results save in the matrix with weight values distributed. By the network, an optional nonlinear input-output mapping relationship can be realized. Concrete mapping relationship materialize at the distributed linking weight values between neurons that build up the ANN. Due to the strong self-adaptability and self-learning-ability as well as excellent and robustness and tolerance ability, it can not only replace many traditional algorithm which is very complicated and time consumption, but also, because the processing to information is more close to person's thought activity habit, It provides a new way for solving the prediction of nonlinear system and unknown model.In this paper, the characteristic is that the U-shaped springback problems is made more comprehensive impact analysis, It is beneficial to lucubrate the springback problems. Put neural network technology, orthogonal test and numerical simulation together into the optimization parameters for the stamping process, guarantee the accuracy of analysis on the premise, obviously, save the time of the making technique, and improve the efficiency of the designing process.