Application of neural network (ANN) to predict XLPE cable in extrusion processes

摘要

This paper presents a new method of modeling the nonlinear parameters of a wire coating systems base on neural Networks with artificial neural network neurons (ANN). A realistic estimation of Crosslinked Polyethylene (XLPE) parameters by means of ANN was calculated. The wire coating thickness and quality depend on the cable take off speed. It could be predicted that rheology of melt viscosity and melt temperature influences crosslinking density, and cable jacketing thickness. In this work, a measurement system consists of an melt pressure sensors, thermocouples, and an ampere meter of the extruder motor drive were used to monitor the extrusion parameter. The artificial neural network-based estimation of the extrusion process parameters prior to plant execution; helps to make the actual extrusion operation more efficient because more realistic parameters may be obtained. Three principal components, most correlated to the state of degree of crosslinking, were extracted from the data set collected by the multiple sensors and fed as inputs to an artificial neural network model designed to predict coating thickness and degree of crosslinking. In this work, it was possible to relate a monitoring system with the principal component analysis and artificial neural networks for determining of changes in the coating layer.