Mivar-based route planning simulation model for obstacle-aware autonomous agricultural machinery

Bauman Moscow State Technical University, 2-ya Baumanskaya Street, 5/1, Moscow, 105005, Russian Federation

^* Corresponding author: ovarlamov@gmail.com

Abstract

Autonomous robot navigation is increasingly becoming an important task that requires solutions. This paper explores the practical application of logical artificial intelligence to address the problem of route planning, using the example of a computer game. Within the scope of this work, a knowledge base model was created, and a new version of the mivar reasoner was developed for integration with Unity. This reasoner allows the activation of logical rules with linear complexity. As a result, a system was developed that processes user input for position and moves an autonomous agent in a virtual environment according to the rules. This work confirms the feasibility of using mivar technologies to improve control systems of autonomous robots in the area of agriculture. The study also emphasizes the adaptability of mivar networks in dynamic environments, demonstrating their ability to effectively process changes in real-time. This research shows enhanced decision-making capabilities and reliable navigation strategies for autonomous agents, setting a precedent for future developments in autonomous digital solutions for agriculture. The results obtained open new prospects for the advancement of technologies in the field of autonomous navigation.