Study of back-propagation and self organizing maps for robotic motion control: A survey

Capturing the motion is a relevant source of information for programming the humanoid robots. Since, the kinematics and dynamics of robots differ significantly from those of humans. However, it is difficult to directly use captured motion data. A controller can be useful to allow a robot to maintain balance while tracking a given reference motion. This paper presents a case study on path planning and control method for a humanoid robot. Application of robotics to this operation requires a path planning system that can take data representing from an external sensor, extract connected paths, and link paths together to form the Cartesian motion for the robotic system. Humanoid robot have the ability to step over obstacles in there path allowing the potential of increased versatility and agility as compared to wheeled robots. So the neural network models can be used to planning motion in a humanoid robot. A comparison of various models can be depicted in this paper along with the beneficial threshold and vigilance parameters.