Advanced Parking Space Management Strategy Design: An Agent-Based Simulation Optimization Approach

Abstract

In this paper, we present a bi-level approach to determine optimal parking management in large parking lots with multiple user classes and different types of parking spaces. The upper level of the model aims to maximize the social benefits of the local parking system, such as the system searching time and walking time by drivers. The lower level is a simulation system constructed using the multi-agent modeling environment NetLogo. However, it is essential to understand how drivers choose their parking location, then we can describe drivers’ movements and responses to different parking management strategies in the micro-simulation system. This study presents the results from a stated preference experiment on the choice behavior of drivers in Shanghai, China. The model attributes are walking distance, traveling time and uncertainties towards finding a vacant parking spot, which were selected based on a literature review. A simulated annealing algorithm is adapted to solve this bi-level simulation optimization model. The approach is applied to optimize parking management in the campus of Tongji University, China. The results show that the implementation of different duration limits is useful to adapt the time-varying demand with different user classes. The approach provides a powerful tool for strategically designing parking management, and for assessing parking policies and parking information systems.