Kinematic control for a quadruped robot using the Newton-Raphson method

Abstract

The present work consisted of the implementation of a method for the solution of the inverse kinematic problem of a quadruped robot in four limbs with 3 degrees of freedom each. The proposed solution was supported by the application of an iterative method (Newton-Raphson) and special emphasis was placed on fine-tuning the algorithm parameters, that is, the number of iterations and the threshold distance, in order to guarantee the convergence of the solution in a few iterations. The work proposed to use the same method both for the function of walking by following the trajectory of a closed Bézier curve of five points, as well as for the control of body posture through the development of pitch and roll movements, and yaw from the base. The results of this solution were simulated in the PyBullet module to which the numerical values of the joints of the quadruped robot obtained from the inverse kinematics were exported to validate its behavior and movements in a simulated physical environment.

Keywords

• cinemática directa
• cinemática inversa
• método de newton-raphson
• robot cuadrúpedo