Nguyen Trang Hoang Thuy 1, 2
1 National Research Tomsk State University
2 Viet Nam Maritime University
The key issue of mobile robot development is locating and navigating the robot in an unknown environment. Navigation and location techniques often have to use a mainframe system to process information received from the robot about the environment in which it traveled. In fact, mobile robots are usually very small, so robot also needs optimal, compact and appropriate technical problem-solving. There are many proposed methods, but these methods only solve algorithmically and the way to process information received from robots on hardware devices is not mentioned. Using the right hardware techniques for mobile robots is also a challenge for scientific researchers. This article describes the way to build a high-performance reconfigurable model based on the principles of parallel architecture combined with the SLAM algorithm for mapping mobile robots in indoor environments. Mobile robots will be built on the ROS platform in the Gazebo environment, putting obstacles that robots are not known, combining in parallel with the SLAM algorithm and the Octomap method. The reconfigurable model is designed in a Matlab environment consisting of many identical computational components, being controlled automatically based on the robot’s location and distance parameter. The results show that the built model will receive high-precision map, quickly and simultaneously with robot movement. In addition, this model can be installed on an IC, PC or FPGA, so this model has high practical applicability
Reconfigurable computing environment
high-performance reconfigurable model
high-performance computing system
SLAM
binary logic model
Нгуен Чанг Хоанг Тхуи ПОСТРОЕНИЕ НАВИГАЦИОННЫХ РЕКОНФИГУРИРУЕМЫХ МОДЕЛЕЙ МОБИЛЬНЫХ РОБОТОВ // Современные наукоемкие технологии. 2020. № 11-1.
С. 49-55;