Resumen

TACORA, Sandro Alcántara; ZAPATA, Erwin López; TORIBIO, Jesús Peralta  y  BUSTINZA, Ricardo Rodriguez. Path planning using potential field algorithms with optimized parameters through neural networks applied to a 6-dof robotic manipulator. Tecnia [online]. 2021, vol.31, n.2, pp.39-47.  Epub 01-Jun-2021. ISSN 0375-7765.  http://dx.doi.org/10.21754/tecnia.v21i2.848.

This paper presents a new algorithm for obstacle avoidance tasks applied to a robotic manipulator of 6 degrees of freedom based on potential field algorithms. Firstly, this manipulator was designed using CAD software and it was intended to be used as our base model to develop the proposed algorithm. Then, the inverse kinematics model was developed using a multivariate iterative control process. Afterwards, the mathematical model was modified by adding a rotational vector. This vector was obtained by the repulsive forces between the obstacle and the six joints of the robot manipulator. Therefore, the manipulator was able to find possible routes that reached the final desired coordinate and avoided any possible obstacle along its path. To optimize these trajectories, a database was created of different trajectories. This database contains trajectories that depend on the initial, final and the obstacle coordinates, with the trajectories hyperparameters optimized. Finally, the simulations have shown that the manipulator was able to complete the task of reaching a point without falling in the obstacle points.

Palabras clave : Path planning; Potential field; Supervised Neural Network; Inverse Kinematics; Robotic Manipulator.

        · resumen en Español     · texto en Español     · Español ( pdf )