A neuro-fuzzy controller for underwater robot manipulators

Autonomous underwater vehicles are increasingly replacing the prevalent remotely operated vehicle-manipulator systems. Most current generation AUVs are not fitted with manipulators and hence are mainly limited to underwater surveying and surveillance tasks because of the difficulty in the coordinated control of the resulting underwater vehicle-manipulator systems. While several researchers have proposed various techniques for control of AUVs, there is still much research to be done on the precise control of underwater manipulators. This paper presents an intelligent control method for underwater manipulators based on the neuro-fuzzy approach. The controller is composed of fuzzy PD control with feedback gain tuning by linguistic rules. A neural network compensator approximates the dynamics of the multiple degrees of freedom manipulator in decentralized form. The proposed controller has advantages of simplicity of implementation due to decentralized design, precision, and robustness to payload variations and hydrodynamic disturbances. It has lower energy consumption compared to the conventional PD control method. The effectiveness of the proposed controller is illustrated by experimental results for a three degrees of freedom underwater manipulator.