Smart Parking Reservation System Based on Distributed Multicriteria Approach

摘要

In metropolitan area, finding a parking space is a difficult task for drivers especially in rush hours. This causes waste of time and fuel and results in traffic congestion. We propose a smart parking system based on multi-agent approach to provide a real-time decision-aid for drivers by handling their preferences. The system ensures an online space allocation based on real-time information by optimizing drivers' preferences with respect to operational constraints. The online allocation problem is considered as a multicriteria decision-aid problem for which we present a mathematical formulation: multicriteria parking reservation problem. The solution is an optimal compromise from the set of efficient solutions which is determined by means of a multicriteria ranking method ELECTRE III. An update of resource allocation is performed to avoid reservation conflict and to ensure constraints satisfaction. After the reservation process, the driver is assisted via a guidance module to reach the reserved place through the shortest path. Simulation results show the wide applicability of the approach in real cases.