To forecast the mechanical property of the hot rolled Q235B strip with artificial neural network

The data of this test were randomly extracted from the administrative machine of the thin strip hot-rolled in an iron and steel company.The purpose is to survey the mechanical properties between the productions by the relationship of the original chemical compositions and the main processing parameters of the hot-rolled with the mechanical properties of the productions established by using the BP network of the artificial neural network.The offline emulation indicates that the relative errors between the forecasting nmubers are very similar to the actual ones.There are 88 percent of the tensile strength forecasting numbers,with only 2 percent errors being more or less than the actual ones,and 86 percent of the yield strength forecasting numbers with only 4 percent errors more or less than the actual ones;78 percent of the elongation forecas-ting numbers have only 4 percent errors being more or less than the actual ones.