Feedback attitude control of space robot using neural motion generator with oscillator and modulator

Description

This paper presents two methods to generate rhythmic and cyclic motions observed in locomotion of animals or insects by using oscillators and modulators. The proposed method enables to specify the approximation accuracy of the generated trajectory to the target trajectory though recurrent neural networks cannot easy to specify it. One named Fourier series method is based on a nonlinear oscillator generating sinusoidal motion and the Fourier series. The other named modulation method is composed of a RNN oscillator and a layered NN modulator. The realized dynamics has the desired trajectory as a limit cycle. Effectiveness of the proposed methods is examined by numerical simulations where a space robot changes its orientation by the cyclic motion of the manipulator.

Journal

Details 詳細情報について

Report a problem

Back to top