NEURAL NETWORK FOR PREDICTING TOOL WEAR AND CURRENT CONSUMPTION IN TURNING AL10% SiCp

This study considers the performance of multilayered perceptions neural network MLPNN for predicting the flank wear and spindle motor current during a turning process. A cemented carbide cutting tool has been used to machine Al/10%SiCp composite material. The input parameters of the MLP model were the cutting speed, feed rate and depth of cut. The output parameters were, flank wear and spindle motor current. The model consists of a three layered feed forward back propagation neural network BPNN. A very good performance of MLP, in terms of agreement with experimental data was achieved.