|
In this paper, instead of using the teaching method by manual operations, anoff line programming of optimal trajectory planning for industrial robots is proposed to improve the accuracy and efficiency. There are two main tasks in optimal trajectory planning: i) the development of the collision avoidance trajectory algorithm, and ii) the minimum time trajectory planning. First, a workcell in CAD system (UARC@) is established. Then, using the developed collision avoidance algorithm, the collision free trajectory can be obtained by searching the simplified configuration space, which describes the status of obstacles in workspace, so that it can avoid prolix process of robot teaching. To find the minimum-time trajectory in joint space, local optimization is Used to optimize the trajectory. Usually a joint path consists of three stages: acceleration, cruise at a constant velocity, and deceleration. By use of the nonlinear and coupled robot dynamic model expressed with Lagrangian formulation, the constraints on torque is converted to compute the maximum acceleration in terms of full use of robot capability. In this manner, the minimum motion time and optimal velocity profile can be procured. Additionally, a path deviation bound is employed to reduce the total traveling time in gross motion. Finally, the position, velocity and acceleration in joint space can be transferred to those in Cartesian space by using forward kinematics and programmed in GSL (Graphical Simulation Language). After animated simulation andcalibration between robot and UARC, the optimal trajectory planning program canbe practically employed to the robot controller.
|