Study Of Intelligent System Of E5015, Welding Rod With Alkaline Coating

Combined with the actual production of electrodes, in accordance with practical requirements, the expert system has been built for a welding material design, management of software. The relations of the chemical composition of the electrode and the mechanical properties are established, which can predict the mechanical properties through the chemical composition in raw materials.The establishment of non-linear relationship is based on artificial neural network. E5015 electrode is the object for research. The chemical composition of raw materials(C,Si, Mn, S, P) as input, mechanical properties as the output, using MATLAB and BP artificial neural network, nonlinear relationship is established.VB is used to establish prediction system of mechanical properties. Entering chemical composition of raw materials interface by VB, through ActivX components using MATLAB, the mechanical properties of prediction can go back to the interface.