Adaptive RBF neural network-based control of an underactuated control moment gyroscope
- PDF / 1,628,155 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 113 Downloads / 218 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
Adaptive RBF neural network-based control of an underactuated control moment gyroscope Jorge Montoya-Cha´irez1 • Fracisco G. Rossomando2 Javier Moreno-Valenzuela1
•
Ricardo Carelli2
•
Vı´ctor Santiba´n˜ez3
•
Received: 11 May 2020 / Accepted: 26 October 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Radial basis function (RBF) neural networks have the advantages of excellent ability for the learning of the processes and certain immunity to disturbances when using in control systems. The robust trajectory tracking control of complex underactuated mechanical systems is a difficult problem that requires effective approaches. In particular, adaptive RBF neural networks are a good candidate to deal with that type of problems. In this document, a new method to solve the problem of trajectory tracking of an underactuated control moment gyroscope is addressed. This work is focused on the approximation of the unknown function by using an adaptive neural network with RBF fully tuned. The stability of the proposed method is studied by showing that the trajectory tracking error converges to zero while the solutions of the internal dynamics are bounded for all time. Comparisons between the model-based controller, a cascade PID scheme, the adaptive regressor-based controller, and an adaptive neural network-based controller previously studied are performed by experiments with and without two kinds of disturbances in order to validate the proposed method. Keywords Control moment gyroscope Underactuated systems Trajectory tracking control Neural networks Real-time experiments Radial basis functions
1 Introduction
& Jorge Montoya-Cha´irez [email protected] Fracisco G. Rossomando [email protected] Ricardo Carelli [email protected] Vı´ctor Santiba´n˜ez [email protected] Javier Moreno-Valenzuela [email protected] 1
Instituto Polite´cnico Nacional-CITEDI, Av. Instituto Polite´cnico Nacional 1310, Col. Nueva Tijuana, 22435 Tijuana, Baja California, Mexico
2
Instituto de Automa´tica-Conicet-Universidad Nacional de San Juan, Av. Libertador 1109, Oeste 5400, Argentina
3
Tecnolo´gico Nacional de Me´xico/Instituto Tecnolo´gico de La Laguna, Blvd. Revolucio´n and Cuauhte´moc SN, 27000 Torreo´n, Mexico
Neural networks have many properties, and therefore, they have been widely studied in the last decades. The universal function approximation property is one of the most important properties for control purposes [1]. Let us mention some recent works relative to neural networks. In [2], authors combined a sliding mode controller with an adaptive neural network technique to compensate for the dynamics of a unicycle-like mobile robot. For the kinematic part, the feedback linearization technique was used. The computed torque controller fused with an adaptive compensator based on neural networks proposed for robotics systems was developed in [3], where the proposed method was validated in a robot SCARA (Selective
Data Loading...
