ДОСЛІДЖЕННЯ ВИЗНАЧЕННЯ МАРШРУТУ ТРАНСПОРТНОГО ЗАСОБУ У «РОЗУМНОМУ МІСТІ»

The article is devoted to the development of a multicriteria approach to vehicle routing in a smart city using a genetic algorithm. The proposed solution combines the principles of evolutionary optimisation while taking into account distance and traffic load indicators, which enables the formation of efficient routes under dynamic changes in the transport environment. The algorithm implements a combination of tournament selection, ordered crossover, and shuffle-based mutation, ensuring a balanced trade-off between convergence speed and solution diversity. A fitness function is presented that determines the quality (optimality) of each potential route by considering the Euclidean distance between nodes and four levels of road congestion. A scenario for optimal route search based on such criteria as route length and road segment congestion level is analysed, which makes it possible to evaluate not only the distance but also the capacity of the route. For simulation modelling, it is assumed that the smart city nodes are randomly distributed within the considered area, subject to a constraint on the maximum communication range. In addition to random placement, nodes are installed at each intersection and road turn, which improves network connectivity and ensures communication at points where the direction of movement changes. The graph of the investigated smart city network is presented. The simulation results, demonstrated using the Sykhiv district of Lviv as a case study, confirm the algorithm’s ability to provide adaptive regulation of traffic flows and to determine a set of alternative routes, thereby offering greater flexibility in route selection under congested traffic conditions. The practical value of the study lies in the possibility of integrating the proposed approach into intelligent transport management systems.