A solution algorithm to the inverse kinematic problem for redundant manipulators using neural networks

摘要

The aim of this paper is to develop a new method for determining the velocity inverse kinematic problem and to present a fully neural-network-based compuational scheme for its implementation. At the velocity level, the manipulator-velocity-ratio (MVR) is proposed as a relevant performance index. By maximizing the MVR, robot configurations with high flexibility can be established, as obtained from the output of the network. At each time step, the neural network is adopted for both the least-norm joint velocity solution and the infinity norm solution. Simulation results demonstrate that the proposed method is effective.