3D Printing Parameter Optimization Research Based On Gaussian Process Regression EGA Response Surface Method

As a rapid prototyping manufacturing technology,3D printing provides enterprises and individuals with new ideas that are different from traditional manufacturing methods by virtue of its short manufacturing cycle and high material utilization rate.Fused deposition printing technology has become the most widely used 3D printing technology with the highest market share due to its low manufacturing cost,simple process,clean and safe features.However,the current 3D printers on the market have problems such as generally low molding accuracy.Improving the molding accuracy of 3D printing has become an important research direction in the field.The process parameters of fused deposition 3D printing can greatly affect the molding quality.Therefore,by optimizing the printing parameters,the quality of printed parts can be improved at a lower cost.The response surface method is an effective method to solve the parameter optimization problem.The response surface method needs to construct a relationship that can express the implicit function.Gaussian process regression is a regression method with good modeling performance.The model is highly interpretable and easy to implement.For 3D printing,this kind of nonlinear,small sample,high-dimensional system can be handled well.It can make up for the shortcomings of the traditional response surface method using quadratic polynomial modeling.This article first analyzes the influencing factors of 3D printing,determines the side surface forming accuracy of the printed parts as the response variable of this article,and selects the four parameters of layer thickness,printing speed,nozzle temperature,and hot bed temperature as the influencing factors of the experiment.Then,for the parameter optimization problem of 3D printing,the response surface method based on Gaussian process regression was introduced.For the global regression method of Gaussian process regression,Latin hypercube sampling was selected for basic experimental design.The GPR response surface method is optimized by the elite retention genetic algorithm.In addition,for the proposed Gaussian process regression response surface method,simulation experiments are arranged to verify the effectiveness of the method.Finally,an empirical study of 3D printing is carried out based on the Gaussian process regression response surface method,and the same parameters and regions were selected for the traditional response surface method experiments,and the research results of the two methods were compared and analyzed.The results proved the advantages of the response surface method based on Gaussian process regression in 3D printing parameter optimization and complex system modeling.