Auto Body Tolerance Allocation Optimization Based On Improved NSGA-â…¡ Algorithm

The body is one of the four main parts of vehicle, which quality influences the performance, appearance and the vehicle quality directly. Also, the performance and cost are influenced by the part manufacturing variation, assembly variation and accumulated deviation. Though tight tolerance can improve the assembly quality, the manufacturers are stressed by much more investment simultaneously. On the contrary, loose tolerance means manufacturing cost decline, but it may cause a series of problems like repairing rate increasing and customers’ complaints. Therefore, it is meaningful to develop a new reasonable tolerance allocation that considering the conflict between quality and cost. Traditional tolerance methods for the body may not be good enough to evaluate the interactive inluences effectively. The main work is as follows:(1) Sensitivity matrix of flexible sheet metal assembly variation modelingTo sovle the above problems, as the basis of variation analysis, the sheet metal assembly variation model is built. First, this paper studies the welding small plane contact model, and builds the sheet metal assembly variation analysis model, which is based on power balance method. Therefore, the sensitive martrix between input variations of parts and output variations of assenly is built.(2) Proposing the improved NSGA-Ⅱ algorithmThe optimization result of NSGA-Ⅱ can meet the needs of each target at the same time. However, the Pareto results are easy to get together while the parts tolerance allocting. Therefore, this thesis proposes the orthogonal design to produce the initial population and the accumulated sorting stategy. The improved algorithm shows that it has a very good convergence and it is easier to choose the excellent individuals by using the improving strategy. The standard test functions are taken to verify the effectiveness of the improved algorithm.(3) Application of improved NSGA-Ⅱ on the optimization of tolerance allocationAfter analyzing the main factors that influence the tolerance allocation, the “tolerance-deviation” relationship is built that based on considering manufacturing deviation and transfering deviation. Then, the improved NSGA-Ⅱ is applied in the body tolerance allocation. Tail lamp bracket is taken to examine it. Multi-oobjective problems are solved by steps of the binary coding, tournament selection, crossover and mutation, and the Pareto set of tolerance allocation is got. At last, TOPSIS, which is a multi-objective decision making method, is used to select optimal solution from the Pareto set.