Optimization Method Of Redundant Manipulators With Multiple Targets

Since redundant manipulators are used in more complex environment, redundant manipulators with the end-effector task have to avoid obstacles. Obstacle avoidance means the manipulator only needs to avoid the obstacle in time, which doesn’t mean that the manipulator cannot get close to the obstacle. Joint motion for the end-effector task is called task motion. Joint motion which doesn’t affect the pose of the end-effector and changes the manipulator’s configuration is called self motion. Task motion and self motion both affect the distance between the manipulator and the obstacle. The method about task motion and self motion is proposed to make the manipulator successfully fulfill the end-effector task and obstacle avoidance. Such task motion and self motion are considered to be coordinated and the method is called coordination method.Firstly, task motion and self motion is discussed. Two kinds of criterion are set up for different applications. One is the minimum distance between the manipulator and the obstacle. The other is the weighted sum of distances between each link and the obstacle. A planar four degrees of freedom manipulator and its simulation model in MATLAB is given to prepare for study below.Secondly, the definition of avoidance performance is proposed to measure the effect on obstacle avoidance from task motion and self motion. After the avoidance performance of self motion is studied when the gradient projection method is used, weighted general inverse is introduced in self motion. The self motion control is discussed by the parameterization of the weighted general inverse. Features of self motion are demonstrated.Thirdly, self motion with negative avoidance performance is discussed. Coordination method proposed here is that task motion is conducted to fulfill the end-effector task and suitable self motion is conducted when needed in order to avoid the obstacle in time. Two kinds of control strategies which are the strategies of the coordination method are proposed.Fourthly, the planar four degrees of freedom manipulator tries to avoid different obstacles with the gradient projection method and the coordination method here. With the coordination method, the manipulator successfully avoids an obstacle which cannot be avoided with the gradient projection method. It shows that, for planar four degrees of freedom manipulators to avoid the obstacle, the coordination method is more applicable than the gradient projection method.Finally, a spatial manipulator can successfully avoid different obstacles with the gradient projection method and the coordination method here. Since joint velocities are not continuous with the coordination, it is concluded that the gradient projection method is superior to the coordination method for spatial seven degrees of freedom manipulators to avoid the obstacle.