An intelligent parking allocation framework for digital society 5.0

Abstract

In recent smart city innovations, parking lot location has garnered a lot of focus. The issue of where to put cars has been the subject of a lot of literature. However, these efforts rely heavily on algorithms built on centralized servers using historical data as their basis. In this study, we propose a smart parking allocation system by fusing k-NN, decision trees, and random forests with the boosting techniques Adaboost and Catboost. Implementing the recommended intelligent parking distribution technique in Smart Society 5.0 offers promise as a practical means of handling parking in contemporary urban settings. Users will be given parking spots in accordance with their preferences and present locations as recorded in a centralized database using the proposed system’s hybrid algorithms. The evaluation of performance considers the effectiveness of both the ML classifier and the boosting technique, and it finds that the combination of Random Forest and Adaboost achieves 98% accuracy. Users and operators alike can benefit from the suggested method’s optimised parking allocation and pricing structure, which in turn provides more convenient and efficient parking options.