Mathematical modelling of mixture of cement raw material with artificial neural networks

Farm Development in computer techonology cause to conclude problems can not solve or solve difficulty with classical method. Chemical composition form raw material of cement need to regulated most suitable. Key of profitable and adavantageous, a production in factories of cement is caused of obtaining mixture of raw material wanted in chemical composition and blending material. In this study, to be modelled of mixture of cement raw material with using artificial neural networks is made real. In that way forming of values, are wanted, for mixture of oxide in farin, with minimum error is armed. Program inputs, which formed for the aim, are low material, high material and iron ore that are raw material of cement. Program outputs are SİO2, AI2O3, Fe203 and CaO that are four main oxide in farm, however. In this way, artificial neural networks there is three input neurons and four output neurons for the artificial neural networks. Back propagation network model is used the form artificial neural networks. Training of network with input-output data in training set and test of network with input-output data in test set is made real. Crooked is between error value that take form and of running of program and number of iteration that take during the training and test. Besides; resuts which is belonging to program, and output value, was compared and their graphics was obtained. Key Words: Artifical Neural Networks, Mixture of Cement Raw Material,