-
Emerson V. A. Dias
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
-
Josias G. Batista
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
-
Catarina G. B. P. Silva
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
-
Geraldo L. B. Ramalho
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
-
Darielson A. Souza
Research Group on Automation, Control and Robotics - Department of Electrical Engineering, Federal University of Ceará, Fortaleza, CE
-
José Leonardo N. Silva
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
-
André P. Moreira
Mobile Robotics Laboratory - Department of Industry, Federal Institute of Education, Science and Technology of Ceará - IFCE, Campus Fortaleza, Fortaleza, CE
Keywords:
Radial basis function, probabilistic roadmap, collision avoidance, path planning, SCARA robot
Abstract
This paper proposes a Path Planning with Collision Avoidance based on a Radial Basis Function (RBF) trained with random points generated by a Probabilistic Roadmap algorithm. Experiments were performed on a computational model of a SCARA manipulator. The trajectory achieved was evaluated using computational cost, R2 (multiple correlation coefficient) and root mean square error (RMSE). The results of the trajectory generated by the algorithms in the Cartesian space and also the trajectories of each joint of the manipulator, calculated from the inverse kinematics, show that RBF proved to be an efficient path estimator. The result was compared with Artificial Neural Networks Multilayer Perceptron (MLP) algorithm, where the RBF proved to be more efficient.
