Latest Papers

ASME Journal of Mechanisms and Robotics

  • Stable Inverse Dynamics for Feedforward Control of Nonminimum-Phase Underactuated Systems
    on January 25, 2023 at 12:00 am

    AbstractAn enhanced inverse dynamics approach is here presented for feedforward control of underactuated multibody systems, such as mechanisms or robots where the number of independent actuators is smaller than the number of degrees of freedom. The method exploits the concept of partitioning the independent coordinates into actuated and unactuated ones (through a QR-decomposition) and of linearly combined output, to obtain the internal dynamics of the nonminimum-phase system and then to stabilize it through proper output redefinition. Then, the exact algebraic model of the actuated sub-system is inverted, leading to the desired control forces with just minor approximations and no need for pre-actuation. The effectiveness of the proposed approach is assessed by three numerical test cases, by comparing it with some meaningful benchmarks taken from the literature. Finally, experimental verification through an underactuated robotic arm with two degrees of freedom is performed.

Neural Network-Based Transfer Learning of Manipulator Inverse Displacement Analysis

Abstract

In this paper, the feasibility of applying transfer learning for modeling robot manipulators is examined. A neural network-based transfer learning approach of inverse displacement analysis of robot manipulators is studied. Neural networks with different structures are applied utilizing data from different configurations of a manipulator for training purposes. Then, the transfer learning was conducted between manipulators with different geometric layouts. The training is performed on both the neural networks with pretrained initial parameters and the neural networks with random initialization. To investigate the rate of convergence of data fitting comprehensively, different values of performance targets are defined. The computing epochs and performance measures are compared. It is presented that, depending on the structure of the neural network, the proposed transfer learning can accelerate the training process and achieve higher accuracy. For different datasets, the transfer learning approach improves their performance differently.
Read More
Journal of Mechanisms and Robotics Open Issues

Generated by Feedzy