A four-layer neural network with 49 neurons trained on analytical dynamic model data predicts joint torques for 7-DOF exoskeleton trajectory tracking, augmented by a PD controller, achieving comparable performance to traditional controllers with reduced computation in simulations.
The choice of loss function depends on the application, for instance, mean squared error (MSE) for regression problems or cross-entropy loss for classification problems
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2022 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Development of a Deep Learning-Driven Control Framework for Exoskeleton Robots
A four-layer neural network with 49 neurons trained on analytical dynamic model data predicts joint torques for 7-DOF exoskeleton trajectory tracking, augmented by a PD controller, achieving comparable performance to traditional controllers with reduced computation in simulations.