Open AccessThe use of neural networks in routing tasks that arise when using unmanned aerial vehicles
Author(s) -
Maksym Ogurtsov
Publication year - 2021
Publication title -
fìziko-matematične modelûvannâ ta ìnformacìjnì tehnologìï/fìzìko-matematične modelûvannâ ta ìnformacìjnì tehnologìï
Language(s) - English
Resource type - Journals
eISSN - 2617-5258
pISSN - 1816-1545
DOI - 10.15407/fmmit2021.33.073
Subject(s) - travelling salesman problem , routing (electronic design automation) , computer science , combinatorial optimization , vehicle routing problem , artificial neural network , optimization problem , artificial intelligence , mathematical optimization , cross entropy method , quadratic assignment problem , mathematics , algorithm , computer network
The paper presents an overview of approaches to the neural networks’ usage in combinatorial optimization problems and other problems that arise when using unmanned aircraft vehicles. It has been determined that the neural networks usage (including the deep learning networks) is possible in almost all types of combinatorial optimization problems, in particular, in routing problems (traveling salesman problem, vehicle routing problem in various versions, etc.) and other similar combinatorial optimization problems that arise when using unmanned aerial systems. Recurrent neural networks with nonparametric normalized exponential functions of supervised learning may be used successfully to solve combinatorial optimization problems.