Abstracto
- Off-line analysis of path tracking of functionally redundant robot manipulators is essential for effective use of industrial robots. The kinematic redundancy of the robot with respect to specific task can be exploited to optimize a desired criterion. Recently, industrial serial robots of 6 DOF are being widely used for a variety of tasks that only require 5 DOF or less, which is causes functional redundancy. For the case of functional redundancy, where the redundant parameters are within the operational space, once the optimal parameters have been selected, the inverse mapping between the operational space and the joint space often corresponds to a closed solution, which means that there are a finite number of joints configurations. This is the case for the most commonly used non-redundant 6 DoF serial robot manipulators. Moreover, each configuration lies on a specific solution branch into joint space. Due to this characteristic, during the optimization process by means of the Pareto Front technique, a sufficiently diverse set of Pareto Optimum points is not obtained to ensure that a solution close to the best combination of the multiple objectives can be found for a global optimization of the robot path. To overcome this difficulty, in this work, the Pareto Front is used to determine the topology of the selfmotions generated by redundancy and from these to identify the invertible subregions between the operational space and the configuration space. Multi-Objective Genetic Algorithms, MOGA, is used to solve the Inverse Kinematics (IK) for global optimization of functionally redundant robots, for incomplete orientation constrained task.