Parking recommendation with meta-heuristic algorithms

Abstract

Due to the exponential growth of cars in urban areas, parking problems have become a significant concern. Addressing this issue requires efficient methods for locating available parking spaces, enhancing the overall experience for drivers. This paper introduces a parking lot recommendation model leveraging meta-heuristic algorithms to generate a list of potential parking locations based on the users travel destinations. The primary objectives of these algorithms include minimizing travel distance, reducing total parking fees, and selecting parking lots with ample available spaces. The proposed model incorporates bio-inspired algorithms, including simulated annealing, genetic algorithms, and their adaptive variants. Our evaluation compares the performance of these algorithms, highlighting the adaptive simulated annealings superior quality of solutions and robustness against local minima. However, it is important to note that this approach comes with a trade-off, requiring longer execution times. In summary, this research contributes a novel parking lot recommendation model that effectively addresses the challenges posed by urban parking. The performance evaluation underscores the efficacy of the adaptive simulated annealing approach, showcasing its potential for practical implementation despite its relatively longer execution time.