Multi-objective Optimization Of Process Parameters For Slug Rivet Riveting

With the rapid development of the modern aircraft aviation industry,the safety and reliability of the aircraft flight process are getting more and more attention,and the quality of the connection of the aircraft siding is really important.The headless rivet can form a reliable and uniform interference fit with the nail hole during the riveting process,which can effectively extend the fatigue life of the joint structure and is widely used in various military and civil aircraft assembly.However,due to its unique structure,the headless rivet must use automatic drilling and riveting equipment when riveting.But there are few self-developed automatic drilling and riveting equipment in China,so the research on headless rivets is still in its infancy.In order to meet the urgent needs of the headless rivet in the advanced assembly technology of large airplane,this paper aims to provide a specific scheme to optimize the parameters of the riveting process of the headless rivet,based on the automatic drilling and riveting machine developed by Zhejiang University.Firstly,this paper introduces the background of the project research,and comprehensively compares the research status of common riveting and headless rivet riveting technology at home and abroad.Also,this paper explains the necessity of optimizing the parameters of the headless rivet riveting process,and analyzes the current automatic drilling riveting technology.The development situation points out that the headless rivet interference connection technology still needs to be explored,which leads to the research content of this paper.The main research method is to combine the finite element simulation,fitting and experiment to optimize the parameters of the riveting process parameters.Secondly,this paper combines the idea of parameterization and modularization to construct a FE model,and verifies the accuracy and effectiveness of the FE model through experiments,which lays a foundation for the subsequent research content.At the same time,the simulation method of multi-task distributed computing is proposed,which greatly reduces the time cost of the simulation process.Based on the simulation results and the internal mechanism of the riveting process,this paper analyzes the motion synchronism of the riveting die and the amount of rivet overhang by controlling the variables.It is pointed that the influence of the single variable on the riveting quality can not guide the multivariate situation.The combination of parameters needs to be solved by an optimization algorithm.Then,this paper decomposes the multivariate multi-objective optimization process into two parts:fitting and optimization.The fitting part uses the distributed simulation calculation to obtain the sample data,and inputs the self-designed neural network model to train the fitting model.When fitting,the indirect fitting method combining model fusion is proposed,which is compared with the scheme of directly fitting the target value.The generalization ability of the model has been greatly improved.The optimization part converts the multi-objective problem into a single-objective problem and compares it with the NSGA-? algorithm.It is found that the latter is more efficient and can find a more complete Pareto solution set.After obtaining the optimal solution set,the test method is used again to verify the accuracy of the optimization result,which shows the practical feasibility of the whole solution.Studies have shown that the optimal average interference size in the Pareto solution is about 25.6%higher than the baseline model,and the optimal interference uniformity is improved by about 66.7%.Finally,the main research work and innovation content of this paper are summarized,and the future research direction is proposed based on this.