运用人工神经网络，对板料基本成形性（单向拉伸、平面应变等基本试验的参数）与模拟成形性（拉深、杯突、扩孔、福井等模拟试验指标）二者的相关性进行了研究．在大量试验数据和反向传播算法的基础上，建立了描述相关性的ＢＰ网络模型．通过该模型，对已知基本成形性参数的板料的模拟成形性指标进行了计算机预测，预测结果与试验结果比较接近．本文的研究方法和结果表明，人工神经网络是研究板料的模拟成形性的一条有效途径 Using artificial neural network, the correlation between the basic formability (uniaxial tensile, plane strain and other basic test parameters) and the simulation formability (such as drawing, cup projection, reaming and Fukui simulation test) Were studied. Based on a large amount of experimental data and back propagation algorithm, a BP network model describing the correlation is established. Through the model, computer simulation of the predictive index of the formability of the sheet with the known basic formability parameters is carried out. The predicted result is close to the experimental result. The research methods and results of this paper show that artificial neural network is an effective way to study the formability of sheet metal