Motion Planning Algorithm for a 10 DOF Redundant Manipulator
Robotics is the most attractive area for scientists and engineers these days. Robotics has potential application in almost every area in human life such like manufacture automation, unknown area exploration, human search and rescue in catastrophe event and handicapped people assistance. In a robot system, manipulator is one of the most important compositions of a robot which can achieve many humanoid function, for example, object grasping and delivering. There are many kinds of manipulator. Classification on basis of DOF:- 1. Non-redundant manipulator: which has 6DOF. 2. Redundant manipulator: which has more than 6DOF. Because the redundant manipulator has more DOF than necessary for position and orientation computation, for a specific position and orientation of end effector, multiple solution exist. For this reason, the redundant manipulator has much more advantages in joint limit constraint avoidance, obstacle avoidance and singularity avoidance compared to the non-redundant manipulator.
Traditional work on redundant manipulator ether can not use the redundancy of the manipulator or just can make motion planning offline. However some tasks require the robot to work in some unknown place with obstacles such like mission of search and rescue in earthquake area. During this type of task the manipulator on the mobile robot is often tele-controlled by a tele-operator with a joystick. The tele-operator sends six-dimension position and orientation command or velocity command to the manipulator through the joystick. However, this command just controls the posi- tion and orientation of end-effector of the manipulator and consequently it can not decide the position of other part of the manipulator. This sometimes may cause big trouble because in some cases when the tele-operator controls the end-effector to one specific position and orientation however in the same time some links of the manipulator hit the obstacle. This situation may probably lead to a serious damage to the manipulator.However in offline motion planning we will use Generalized Voronoi Graph (GVG) to deduce a path from starting point to end point. And we will try to make others to follow the head like a snake the body will follow th head in the most cases.
The redundent manipulators can be used for various purposes like in making snake robots, rescue robots, spy bots, robotic arms with high DOF (like in Spiderman movie)
In this project I would like to explore the various approches to navigate a redundent manipulator in a given space and will try to implement them.