Multiple self-organizing maps for control of a redundant manipulator in an environment with obstacles

Abstract

This paper proposes multiple self-organizing maps (SOMs) for control of a visuo-motor system that consists of a redundant manipulator and multiple cameras in an unstructured environment. The maps control the manipulator so that it reaches its end-effector to targets given in the camera images. Also the maps make the manipulator take obstacle free poses. Multiple cameras are introduced to avoid occlusions and multiple SOMs are introduced to deal with multiple camera images. Using two or more SOMs may cause inconsistency among them. We, therefore, developed a new learning method of SOMs to keep the consistency. We also developed a simple collision avoidance approach by using the multiple SOMs and a simple path planning technique. Since the collision free pose is accomplished by the multiple SOMs, the path planning system only plans the end-effectorpsilas path. Simulation results will be shown. This paper proposes multiple self-organizing maps (SOMs) for control of a visuo-motor system that consists of a redundant manipulator and multiple cameras in an unstructured environment. The maps control the manipulator so that it reaches its end-effector to targets given in the camera images. Also the maps make the manipulator take obstacle free poses. Multiple cameras are introduced to avoid occlusions and multiple SOMs are introduced to deal with multiple camera images. Using two or more SOMs may cause inconsistency among them. We, therefore, developed a new learning method of SOMs to keep the consistency. We also developed a simple collision avoidance approach by using the multiple SOMs and a simple path planning technique. Since the collision free pose is accomplished by the multiple SOMs, the path planning system only plans the end-effectorpsilas path. Simulation results will be shown.