Trajectory Control of Flexible Manipulators Using Forward and Inverse Models with Neural Networks

This study explores trajectory control in flexible manipulators using neural-network-based forward and inverse modeling. Unlike traditional approaches that enhance precision by increasing structural rigidity—often at the cost of added weight and energy consumption—this work focuses on lightweight flexible manipulators, which are more suitable for aerospace and other weight-sensitive applications but introduce control complexities due to elastic deformations. To address these challenges, neural-network-based models are proposed for a two-link, three-degree-of-freedom (3-DOF) flexible manipulator. Simulation and experimental results show that incorporating system delay compensation into the training data significantly improves tracking accuracy. Nonetheless, difficulties remain in achieving smooth trajectory generation. The findings highlight the potential of neural networks in adaptive control and point to future opportunities for refining input–output modeling to better align theoretical developments with practical implementation.