Abstract
The Vehicle Routing Problem is about optimizing the routes of vehicles tomeet the needs of customers at specific locations. The route graph consists ofdepots on several levels and customer positions. Several optimization methodshave been developed over the years, most of which are based on some type ofclassic heuristic: genetic algorithm, simulated annealing, tabu search, antcolony optimization, firefly algorithm. Recent developments in machine learningprovide a new toolset, the rich family of neural networks, for tackling complexproblems. The main area of application of neural networks is the area ofclassification and regression. Route optimization can be viewed as a newchallenge for neural networks. The article first presents an analysis of theapplicability of neural network tools, then a novel graphical neural networkmodel is presented in detail. The efficiency analysis based on test experimentsshows the applicability of the proposed NN architecture.