Automatic Drilling and Riveting Process Parameter Optimization Method Based on BP Neural Network and Multi-Objective Particle Swarm Optimization Algorithm
LI Chao, WANG Zhongqi, CHANG Zhengping, MA Jianzhi
Northwestern Polytechnical University, Xi’an 710072, China
摘要飞机壁板自动钻铆大量采用干涉连接,干涉量的均匀程度与壁板变形程度的协同控制是目前亟须解决的问题,为此提出一种基于 BP 神经网络(BP neural network,BPNN)的优化方法。以压铆力、压铆过程时间、压铆停留时间和夹紧力为变量,以仿真数据为样本,采用 BP 神经网络,建立干涉量均匀程度和壁板变形程度的预测模型,利用多目标粒子群算法(Multi-objective particle swarm optimization,MOPSO)进行多目标优化。仿真及试验结果表明,优化后的参数能够显著提升干涉量的均匀程度并有效降低板件的变形程度。
Abstract:Interference connection is widely used in automatic drilling and riveting of aircraft panel. The coordinated control of the degree of uniformity of interference and panel deformation is an urgent problem to be solved. In this paper, an optimization method based on BP neural network (BPNN) is proposed. Taking the riveting force, riveting process time, riveting retention time and clamping force as variables, and taking the simulation data as samples, the prediction model of interference uniformity and wall deformation was established by using BP neural network. Multi-objective particle swarm optimization (MOPSO) algorithm was used for multi-objective optimization. The simulation and experimental results show that the optimized parameters can significantly improve the uniformity of interference and effectively reduce the deformation degree of plate.
李超,王仲奇,常正平,马健智. 基于 BP 神经网络和多目标粒子群算法的自动钻铆工艺参数优化方法[J]. 航空制造技术, 2021, 64(23/24): 94-102.
LI Chao, WANG Zhongqi, CHANG Zhengping, MA Jianzhi. Automatic Drilling and Riveting Process Parameter Optimization Method Based on BP Neural Network and Multi-Objective Particle Swarm Optimization Algorithm. Aeronautical Manufacturing Technology, 2021, 64(23/24): 94-102.