Abstract: Wing-body fairing composite panels are prone to shape deviations after manufacturing, which can degrade the aerodynamic performance of aircraft and therefore require shape adjustment during assembly. To address the low optimization efficiency and the difficulty in effectively controlling adjustment force during the shape adjustment process, an optimization method for shape adjustment process parameters based on back propagation neural network (BPNN) and non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) was proposed. Deformation and force prediction models for the panel to be assembled were established and embedded into a multi-objective optimization framework. Using the ideal point method, the optimal layout and driving displacements of adjustment units were efficiently obtained by simultaneously considering contour accuracy and adjustment force. Experimental results show that the proposed method improves optimization efficiency while ensuring prediction accuracy, and provides a reliable technical route for the high-quality assembly and shape adjustment of composite panels in practice.

Key words: aircraft assembly, composite panels, shape adjustment process parameters, BPNN, multi-objective optimization