The motion of a robot manipulator submerged in water is strongly affected by fluid forces, and it is an important technique to avoid their influence on the motion of an aquarobot manipulator to achieve high-speed, precise motion. This paper deals with extension of the technique of air bubble ejection from the manipulator surface, i.e., the mechanisms of reduction of drag force by air bubble ejection and its effects on the control of the aquarobot manipulator. Using a two-degree-of-freedom and two-joint manipulator, experiments were performed and the following major results were obtained: (1) There exists a particular pattern of air bubble ejection for reduction fluid drag force acting on the manipulator and it resulted in reduction of drag force by 25% compared to that for no air bubble ejection. (2) There exists a particular pattern of air bubble ejection that brought a 40% reduction of the control torque required for compensating the fluid drag force. (3) The major mechanisms for drag force reduction were found to be the controlled flow pattern around the manipulator formed by ejecting air bubbles. However, it is noted that these effects of air bubble ejection were dependent on the mode of manipulator motion.
Neural Network Control for Fluid Drag Force Reduction in an Aqua-Robot Arm by Ejecting Air Bubbles.
