PLANEJAMENTO DE TRAJETÓRIA COM A POSSIBILIDADE DE PRESENÇA DE OBSTÁCULOS PARA UM MANIPULADOR ROBÓTICO COLABORATIVO CLÁSSICO

• DOI: 10.22533/at.ed.9072322126

• Palavras-chave: Sistemas Computacionais Inteligentes, Robótica Cognitiva, Robótica. Teoria de Agentes

• Keywords: Intelligent Computational Systems, Cognitive Robotics, Robotics. Agent Theory

• Abstract: In this work, the objective is to present artificial neural networks (ANNs) to solve the inverse kinematics of a 2 degrees-of-freedom (DOF) robotic manipulator. In addition, a decision tree was included to increase the safety of the arm in motion when an object appears in its trajectory, with two possible alternatives: adding a point that can deviate from the obstacle or stopping and waiting for the trajectory to be cleared, as in collaborative robots that work side by side with humans. The fifth-degree polynomial was used in trajectory planning, and the analysis led to the fifth-degree polynomial because it presented a plausible solution. According to the promising results obtained, the best method is ANN combined with a strategy, as it presented more robust results. This work ends with a conclusion and suggestions for future work.