In this work, the approaches of genetic algorithms (GA) and artificial neural networks (ANN) are compared to solve the inverse kinematics applied in a robotic manipulator. The method with the best result in the comparison is then used to act in conjunction with another concept of robotics which is collaborative robotics, responsible for increasing the safety of both the manipulator and the human being when an object and/or person appears in the trajectory of the manipulator. The classic concept of inverse kinematics is related to the relatively new concept of collaborative robotics through trajectory planning, which in this work used the fifth-order polynomial due to its ability to control position, speed, and acceleration. According to the results obtained, the best method in the comparison for the solution of the inverse kinematics was that of artificial neural networks because it has the shortest response time and the most robust results.

This work presents a mathematical model of an aeropendulum system with two sets of motors with propellers and the design and simulation of a loop-shaping ℋ_∞ control for this system. In this plant, the objective is to control the angular position of the pendulum rod through the torque generated by the thrust of the motorized propellers at the end of the rod’s axis. The control design is obtained by first using feedback linearization and then designing the ℋ_∞ controller using the resulting linear system. For the control strategy validation, simulations were conducted in the Matlab/Simulink® environment, and the weighting functions for the ℋ_∞ controller were adjusted to obtain the desired performance and stability of the closed-loop system. The simulation results show the efficiency of the applied methodology.

Qualis indexed Engineering Journal and Science Journal to publish paper with DOI, NAAS Rating and journal has global recognized indexing

Qualis indexed Engineering Journal and Science Journal to publish paper with DOI, NAAS Rating and journal has global recognized indexing Open Access International Journal