Intelligent layout optimization of reconfigurable flexible fixture for assembling multiple aircraft panels.

The traditional assembly of aircraft panels mainly adopts special welding fixtures, which cannot meet the needs of timely production and delivery of diversified products. To address this problem, this paper designs a low-cost reconfigurable flexible fixture for preassembling the aircraft panel (short for RFFP), which can achieve the low-cost and flexible preassembly of multiple panels. Besides aiming at the layout optimization problem of the RFFP to ensure the aerodynamic accuracy of the panel, an intelligent layout optimization method of the fixture is proposed, which is combined the convolutional neural network (CNN) with the genetic algorithm (GA) to solve the optimal layout of contour boards as locators for the skin. In this method, the mathematical model of fixture layout optimization is firstly established with the objective of minimizing skin all strain energy. On this basis, the finite element model of the panel-fixture system is built considering the riveting, the self-weight of the panel, and the tension of the elastic bandage. Then, a fixture layout optimization agent model based on CNN is created, and GA is applied to obtain the optimal solution of the agent model, i.e., the optimal fixture layout. Finally, an aircraft fuselage panel is taken as an experimental case for validating the proposed method. The results show that the skin all strain energy can be effectively reduced. The preassembly quality and efficiency of the aircraft panel can be improved, demonstrating the effectiveness and feasibility of this proposed method. [ABSTRACT FROM AUTHOR]