钢球冲击薄壁结构数据集
收藏资源简介:
本研究提出了一种利用循环神经网络(RNN)进行冲击定位的方法,并使用真实实验数据进行了训练。数据集包含5000次钢球冲击薄壁结构的事件,每个事件都记录了冲击位置以及对应的传感器信号。数据集的创建通过自动化实验装置,使用机器人将钢球随机投掷到铝板上,并用压电传感器记录冲击产生的弹性波。数据集用于训练神经网络,以实现更精确的冲击定位。
This study proposes a recurrent neural network (RNN)-based approach for impact localization, which is trained on real experimental data. The dataset encompasses 5000 steel ball impact events on thin-walled structures, with each event documenting both the impact location and the corresponding sensor signals. The dataset was developed using an automated experimental setup: a robot randomly launches steel balls onto aluminum plates, and piezoelectric sensors are employed to record the elastic waves induced by these impacts. This dataset is designed for training neural networks to achieve more precise impact localization.

- 1Localization of Impacts on Thin-Walled Structures by Recurrent Neural Networks: End-to-end Learning from Real-World Data奥地利林茨约翰内斯开普勒大学技术力学研究所 · 2025年



