Real-Time Prediction of Off-Street Parking Spaces Based on Dynamic Resource Allocation and Pricing

Due to rapid economic development, people’s living standards continue to improve, causing the number of urban motor vehicles to rise. Nevertheless, as the number of motor vehicles continues to grow, it is becoming increasingly difficult to find a vacant parking space. This study proposes a new intelligent parking system based on a multi-agent approach and dynamic pricing in order to achieve more effective, convenient, and accurate parking space prediction effect. By selecting a path, the driver is guided to a parking lot with unoccupied spaces. The system assigns and reserves a vacant parking space based on the driver’s utility that combines travel time, cruising time, walking to destination and parking cost. MATSim transport simulation platform is used to simulate drivers from off-street parking in Tunis city center. The numerical analysis, based on real data from Open Data Tunisia, demonstrates that the developed intelligent off-street parking system reduces traffic congestion, minimizes travel time, and utilizes parking space more efficiently during peak hours.