Comparison of the Technological Time Prediction Models

The paper sets out to describe the results obtained by investigating the prediction of technological parameters and, indirectly, of technological time needed for seam tube polishing. The results of experimental investigations were used to define, analyse and compare two models. One is a mathematical i.e. statistical model obtained by the application of the least squares method and the least absolute deviation method. The other is a model based on the application of neural networks. To define the model based on the application of neural networks various structures of the back-propagation neural network with one hidden layer were analysed and the optimal one with the minimum RMS error was selected. The more precise predictions of technological time provided by the models supplement the previously defined manufacturing operations, replace the predictions based on the technologists' experience and form the basis on which to plan production and control delivery times. The work of technologists is thus made easier and the production preparation technological time shorter.