Multi-objective process parameter optimization for winding process based on NSGA-II algorithm

Based on the composites winding molding process, the response surface methodology was used to design the wet winding molding test. Taking the interlayer shear strength and porosity of the winding products as the key performance indexes, the multivariate regression prediction model of winding tension, the gap between squeezer rollers and winding speed on the performance of winding products was established based on the experimental results, and the accuracy of the regression model was verified. Combining the regression model and Morris method for ranking the sensitivity of different winding product performance characterization parameters to each process parameter, the relative stability intervals of each process parameter were obtained and the validity of sensitivity analysis through the winding molding test was verified. Taking the large interlayer shear strength and small porosity of the winding products as the target, the dedication rate of interlayer shear strength is 60.9% and the dedication rate of porosity is 39.1% through the principal component analysis, and the optimal solution set of the process parameters is obtained by using the NSGA-II algorithm: The winding tension is 65.1 N, the gap between squeezer rollers is 0.12 mm, and the winding speed is 0.17 m/s. The interlayer shear strength of the winding product is 54.4 MPa, the porosity is 1.24%, and the fiber volume fraction is 74.13vol%.