Design and experimentation of acceleration-level drift-free scheme aided by two recurrent neural networks

• Published in: IET control theory & applications Vol. 7; no. 1; pp. 25 - 42
• Main Authors: Zhang, Zhijun, Zhang, Yunong
• Format: Journal Article
• Language: English
• Published: Stevenage The Institution of Engineering and Technology 01.01.2013; John Wiley & Sons, Inc
• Subjects: acceleration level drift free scheme; ALDF; Computer simulation; cyclic motion; Drift; GNN; gradient descent method; gradient methods; gradient neural network; joint angle drift problems; joint velocity drift problems; linear equality constraint; motion control; Neural networks; neural‐dynamic method; neurocontrollers; quadratic program; Quadratic programming; Recurrent neural networks; redundant manipulators; RNN; robot manipulators; Robots; second‐order system; Solvers; Tracking; Zhang neural network; ZNN; joint angle drift problems; acceleration level drift free scheme; robot manipulators; GNN; linear equality constraint; second-order system; redundant manipulators; quadratic programming; joint velocity drift problems; cyclic motion; ALDF; RNN; motion control; recurrent neural networks; neurocontrollers; neural-dynamic method; ZNN; quadratic program; gradient descent method; gradient methods; QP; Zhang neural network; gradient neural network
• ISSN: 1751-8644; 1751-8652; 1751-8652
• DOI: 10.1049/iet-cta.2011.0573

To solve the joint-angle and joint-velocity drift problems in cyclic motion of redundant robot manipulators, an acceleration-level drift-free (ALDF) scheme subject to a linear equality constraint is proposed, of which the effectiveness is analysed and proved via the theory of second-order system. The scheme is then reformulated into a quadratic program (QP). Furthermore, two recurrent neural networks (RNNs) are developed for solving the resultant QP problem. The first RNN solver is based on Zhang et al's neural-dynamic method and called Zhang neural network (ZNN), whereas the other is based on the gradient-descent method and called gradient neural network (GNN). Comparison results based on computer simulations between the ZNN and GNN solvers with a circular-path tracking task demonstrate that the ZNN solver has faster convergence and fewer errors. In addition, the hardware experiments of tracking a straight-line path and a rhombic path based on a six degrees of freedom manipulator validate the physical realisability and efficacy of the proposed ALDF scheme and the two RNN QP-solvers. Moreover, the position, velocity and acceleration error analyses indicate the accuracy of the proposed ALDF scheme and the corresponding RNN QP-solvers.