Radial Basis Function Network Compensators for Uncertainties of Robotic Manipulators

This report proposes a decentralised compensation scheme for uncertainties and modelling errors of robotic manipulators. The scheme employs a central decoupler and independent joint neural network controllers. Recursive Newton Euler formulas are used to decouple robot dynamics to obtain a set of equations in terms of each joint's input and output. To identify and suppress the effects of uncertainties associated with the model, each joint is controlled separately by Gaussian radial basis function network controllers using direct adaptive techniques. The effectiveness of the proposed adaptive control scheme is demonstrated by controlling the three primary joints of PUMA 560. Simulation results show that this control scheme can achieve fast and precise robot motion control under substantial payload variations.