Multiobjective Optimization System of Extrusion Process Parameters for Targeted Microtubes Based on RSM and NSGA-II

Abstract

An optimization system of extrusion process parameters was designed to rapidly extrude a microtube with a target diameter and wall thickness for forming medical balloons with personalized size characteristics. Firstly, the double convected pom-pom (DCPP) model was selected for the numerical simulation of the tube extrusion process, and different parameter combinations and their corresponding simulation results were extracted based on orthogonal experiments. Secondly, the nonlinear functional relationships between microtube diameter and wall thickness with extrusion parameters were established based on the response surface model (RSM). Finally, the nondominated sorting genetic algorithm II (NSGA-II) and response surface model (RSM) were mixed to find the optimal parameters of the tube extrusion. The extrusion experiment results, taking three microtubes with different target size characteristics (1, 0.15), (1, 0.2), and (1.5, 0.2) as examples, show that the average error between the actual and predicted values of diameter and wall thickness of three microtubes is 6.53%, where the average diameter error is 7.3%, the wall thickness error is 5.77%, the maximum value is 9%, and the minimum value is only 1.3%.