MMPD-Net: An Aircraft Deformation Prediction Network Based on Multimodal Fusion

Abstract

In the assembly process of aircraft manufacturing, the aircraft structural is flexible, it is easy to deform under the influence of material factors and working condition factors. Therefore, it is necessary to propose a deformation prediction model to timely find deformation problems. In recent years, with the development of artificial intelligence technology, deep neural network has been widely used in intelligent manufacturing. This paper combines deep neural network with aircraft deformation prediction, and proposes an aircraft deformation prediction network that based on multimodal fusion (MMPD-Net). We consider both the aircraft structure data and the working condition data on aircraft deformation prediction. MMPD-Net extracts the features of aircraft structure mode and working condition mode respectively. We propose a multimodal fusion network to fuse features, which include the algorithm of average pooling and Bayesian decision. Compared with the mainstream point cloud classification network, MMPD-Net achieves higher classification accuracy on aircraft deformation dataset.