James F. Whidborne

Gabriele Dessena, Alessandro Pontillo, Dmitry I. Ignatyev et al.

In general, there is a mismatch between a finite element model {(FEM)} of a structure and its real behaviour. In aeronautics, this mismatch must be small because {FEM}s are a fundamental part of the development of an aircraft and of increasing importance with the trend to more flexible wings in modern designs. Iterative finite element model updating can be computationally expensive for complex structures, and surrogate models can be employed to reduce the computational burden. A novel approach for FEM updating, namely assembly-like, is proposed and validated using real experimental data from a flexible wing. The assembly-like model updating framework implies that the model is updated as parts are assembled. Benchmarking against the classical global, or one-shot, approach demonstrates that the proposed method is more computationally efficient, since a normalised workload proxy based on solver-reported model size and memory footprint indicates about 28\% lower overall effort. Aapproximately 95\% of the required solves are performed on lower-fidelity subassembly models with smaller equation counts and memory requirements. Despite the reduced reliance on full-wing evaluations, the new approach retains the fidelity, within 1\% of a joint natural frequencies and modal shapes index, of the global approach.

Lei He, Nabil Aouf, James F. Whidborne et al.

In this paper, a deep reinforcement learning (DRL) method is proposed to address the problem of UAV navigation in an unknown environment. However, DRL algorithms are limited by the data efficiency problem as they typically require a huge amount of data before they reach a reasonable performance. To speed up the DRL training process, we developed a novel learning framework which combines imitation learning and reinforcement learning and building upon Twin Delayed DDPG (TD3) algorithm. We newly introduced both policy and Q-value network are learned using the expert demonstration during the imitation phase. To tackle the distribution mismatch problem transfer from imitation to reinforcement learning, both TD-error and decayed imitation loss are used to update the pre-trained network when start interacting with the environment. The performances of the proposed algorithm are demonstrated on the challenging 3D UAV navigation problem using depth cameras and sketched in a variety of simulation environments.