DECENTRALIZED CONTROL OF A GROUP OF AUTONOMOUS MOBILE OBJECTS WHEN FORMING A TRAJECTORY OF MOVEMENT

Abstract

The article considers algorithms for generating unmanned aerial vehicles motion trajectories during search and rescue and liquidation operations. The methods and algorithms for controlling the motion of a unmanned aerial vehicles group in formation, when deployed in a line, when deployed in a rank, when turning, in a column are described. Control is carried out using alternative collective adaptation algorithms based on the ideas of collective behavior. The operating principles of one adaptation machine are considered. The purpose of controlling slave robots is to minimize deviations. To implement the adaptation mechanism, the parameters of the vector are matched with adaptation machines that model the behavior of adaptation objects in the environment. A structure has been developed for the process of alternative collective adaptation of parameters that control the motion of a group of unmanned aerial vehicles in formation. Original rules for controlling parameters have been developed that have a number of advantages over other methods: complete decentralization of control in combination with dynamic correction of robot parameters that set the position and orientation of the robot in an absolute coordinate system, and the linear velocity of the robot, respectively. A structure of a maneuver performed by a robot to correct parameter deviations is proposed. Control is performed using an alternative collective adaptation algorithm based on the ideas of collective behavior of adaptation objects, which allows for efficient processing of emergency situations, such as agent failure, changes in the number of agents due to failure or sudden acquisition of communication with the next agent, as well as in conditions of measurement errors and noise that satisfy certain restrictions.