Research on Prediction Method of Mechanical Properties of Aluminum Profiles Based on RBF Neural Network

Abstract

This paper studies on the prediction model of mechanical properties of aluminum profiles based on spatial division of sample process parameters, according to the process parameters of aluminum profiles processing and its corresponding product performance. The direct mapping relationship between process parameter subspace and product performance parameter space is established by using RBF neural network. It provides theoretical basis and solution for the mechanical properties prediction of aluminum profiles, optimization of process parameters and improvement of the quality.