Design of a Strategy for Planning of Trajectories With Avoidance of Obstacles of a Quadrotor Drone Used in the Characterization of Geological

Abstract

This article focuses on the development of an algorithm for a Quadrotor-type Drone that allows it to recognize the environment and move around in it by avoiding obstacles. This article covers the simulation phase that takes place in the MATLAB software, where algorithms for planting trajectories in known static environments were created in two stages: an artificial force field (APF) model is created for each element in the environment using sigmoid functions, the trajectories are subsequently calculated using gradient descent based strategies. Two strategies were developed, the first based on mobile points that interconnect the drone with the goal, where later each point moves towards areas free from the influence of obstacles, following the potential field, which makes paths free of obstacles. The second is based on the use of the concept of safe zones, which is used as a criterion to update the position of the points.