Multilayer feedforward networks are universal approximators
Neural Networks
Stable adaptive control of robot manipulators using “neural” networks
Neural Computation
Discrete and Combinatorial Mathematics: An Applied Introduction
Discrete and Combinatorial Mathematics: An Applied Introduction
Robot Dynamics and Control
Introduction to Robotics: Mechanics and Control
Introduction to Robotics: Mechanics and Control
Adaptive Control
Adaptive Neural Network Control of Robotic Manipulators
Adaptive Neural Network Control of Robotic Manipulators
Modelling and Control of Robot Manipulators
Modelling and Control of Robot Manipulators
Control of Robot Manipulators
Convergence Rates of Approximation by Translates
Convergence Rates of Approximation by Translates
Universal approximation bounds for superpositions of a sigmoidal function
IEEE Transactions on Information Theory
An Improved Dynamic Neurocontroller Based on Christoffel Symbols
IEEE Transactions on Neural Networks
Gaussian networks for direct adaptive control
IEEE Transactions on Neural Networks
| Hi-index | 0.01 |
Dynamic modelling plays an important role in the design of controllers for robot manipulators. The presence of uncertainties such as those from unknown parameters makes the structured network modelling a powerful tool. In this sense, static neural networks based on the Kronecker product are introduced and some interesting properties connected to the dynamics are exploited. The main result of this work is the mathematical determination of the bandwidth regarding with nonlinear terms of the dynamics equations. In order to follow a methodological approach that is consistent with the sampling theorem, it is necessary to treat the dynamics in the context of multivariate Fourier analysis. Finally, the selection of design parameters is crucial in the performance of the system, making a tremendous influence on both the boundedness of the weights and the approximation errors so that explicit formulae are provided to treat with this topic.