Optimal Allocation of Public Parking Slots Using Evolutionary Algorithms

Abstract

This article presents an innovative approach based on an evolutionary algorithm to calculate the best allocation of available parking slots in a city according to the driver's preferences. We have worked with an urban scenario created with the SUMO traffic simulator, in which cars follow a pattern of real movements to go from a start position to the parking slot assigned by the algorithm. The results of the SUMO analysis of a potential solution are used for calculating its fitness value. Additionally, we have used different amounts of cars and parking slots to consider diverse loads of the system and therefore diverse algorithm behaviors. As a sanity check, we have compared the results versus other techniques, like random search and simulated annealing, obtaining a significant improvement in the results.