Roughness Prediction Model Based on Neural Network

Abstract

In order to meet the needs of people's consumption quality upgrade, major manufacturers have put forward higher requirements for the machining accuracy, surface roughness and material properties of the workpiece. By building a prediction model (PM) to predict the surface of the formed parts with different process parameters Roughness, using NN to build a roughness prediction model (PM), adjust the process parameters through the model, and ensure the surface quality of the parts. Therefore, this paper discusses and studies the roughness PM based on neural network (NN). This paper firstly introduces the basic situation of the surface topography simulation system and the workflow of the PM in detail, and then builds the PM from the surface roughness measurement, data processing and overall parameter design, and finally compares the model results. Analysis and evaluation index analysis.