In this work, we propose to use a methodology based on an identification and control scheme using recurrent high-order neural networks (RHONNs) trained online with an unscented Kalman filter (UKF) based algorithm. Initial parameters for the UKF are selected employing two bio-inspired algorithms: Bat Algorithm (BA) and Grey Wolf Optimizer (GWO). The controller design includes sliding modes techniques for trajectory tracking of the 7 degrees of freedom (DOF) Mitsubishi PA10-7CE robotic arm.
A bio-inspired optimization approach for trajectory tracking of robotic arms.
2022-12-05
1208208 byte
Conference paper
Electronic Resource
English