Optimal Pathfinding Based on Artificial Intelligence Tools

Nowadays, artificial intelligence tools are developing at a fast pace, improving their efficiency and productivity. Large volumes of data generated in various spheres of human activity have not become a problem for modern information technologies. The artificial intelligence tools use statistical theory, methods of machine and deep learning. These tools are the basis for obtaining new knowledge, decision-making, optimization, and are the main element of the well-known Industry 4.0 approach. Transport and logistics problems can also be solved with the help of artificial intelligence. Therefore, this paper focuses on researching the use of reinforcement learning to find the best vehicle route. The developed method consists of three elements: a nonparametric clustering procedure, Dijkstra’s algorithm and a method of accelerating calculations. In general, the proposed method is reduced to: 1) transformation of the original image of the map into a maze with the designation of permitted and prohibited passages; 2) determination of the fastest path from the starting point to the exit from the maze. The paper also contains the program realization for implementing the proposed method.