AUTHOR=Huang Wenhui , Lin Yunhan , Liu Mingxin , Min Huasong TITLE=Velocity-aware spatial-temporal attention LSTM model for inverse dynamic model learning of manipulators JOURNAL=Frontiers in Neurorobotics VOLUME=18 YEAR=2024 URL=https://www.frontiersin.org/journals/neurorobotics/articles/10.3389/fnbot.2024.1353879 DOI=10.3389/fnbot.2024.1353879 ISSN=1662-5218 ABSTRACT=Introduction

An accurate inverse dynamics model of manipulators can be effectively learned using neural networks. However, further research is required to investigate the impact of spatiotemporal variations in manipulator motion sequences on network learning. In this work, the Velocity Aware Spatial-Temporal Attention Residual LSTM neural network (VA-STA-ResLSTM) is proposed to learn a more accurate inverse dynamics model, which uses a velocity-aware spatial-temporal attention mechanism to extract dynamic spatiotemporal features selectively from the motion sequence of the serial manipulator.

Methods

The multi-layer perception (MLP) attention mechanism is adopted to capture the correlation between joint position and velocity in the motion sequence, and the state correlation between hidden units in the LSTM network to reduce the weight of invalid features. A velocity-aware state fusion approach of LSTM network hidden units' states is proposed, which utilizes variation in joint velocity to adapt to the temporal characteristics of the manipulator dynamic motion, improving the generalization and accuracy of the neural network.

Results

Comparative experiments have been conducted on two open datasets and a self-built dataset. Specifically, the proposed method achieved an average accuracy improvement of 61.88% and 43.93% on the two different open datasets and 71.13% on the self-built dataset compared to the LSTM network. These results demonstrate a significant advancement in accuracy for the proposed method.

Discussion

Compared with the state-of-the-art inverse dynamics model learning methods of manipulators, the modeling accuracy of the proposed method in this paper is higher by an average of 10%. Finally, by visualizing attention weights to explain the training procedure, it was found that dynamic modeling only relies on partial features, which is meaningful for future optimization of inverse dynamic model learning methods.