МОДЕЛЮВАННЯ АВТОНОМНОГО ПЕРЕСУВАННЯ РОБОТА НА ПЛОЩИНІ В УМОВАХ НЕВИЗНАЧЕНОСТІ

Анотація

We investigate the issue of autonomous control of the movement of a robot (formal executor) on a plane with obstacles. We model a rectangular plane with a matrix of square sections (cells), obstacles are located in separate sections. The starting and nishing positions (cells) of the robot are given. The placement of obstacles is unknown a priori for the robot, in the current position it can determine obstacles only in adjacent sections

The robot can move in each of eight directions, in a position with obstacles movement is impossible. At the moment of approaching an obstacle, the robot analyzes the state of adjacent cells and determines the direction of movement by making a corresponding turn. The goal of the executor is to move from the starting position to a given nishing position. We believe that the placement of obstacles allows us to construct the desired path. Given the convexity of the obstacles (in terms of cells) and their isolation, a heuristic algorithm for autonomous control of the robot's movement has been constructed, which ensures its movement to a given nishing position. To minimize the number of steps (passed cells), a "least square rule" is proposed. Commands for analyzing the state of the robot and the environment, motion commands, and analytical commands are dened. All commands are presented in the form of methods of the corresponding classes for tasks of autonomous robot motion control. The proposed algorithm is implemented in the form of a software package for modeling and visualization of robot motion.