Parking rank: A novel method of parking lots sorting and recommendation based on public information

Abstract

Nowadays, it's difficult to find available parking spaces in big cities within China due to the rapid growth of the vehicles. Parking guidance system(PGS) would play an important role to reduce the time spent on looking for parking spaces. An intuitional idea is to collect the real-time occupancy information of city-wide parking lots and guide the vehicles to the proper one on demands. But the idea is hardly implemented practically because of the huge financial and time costs on linking all city-wide parking lots in early deployment stage. Unlike the expensive real-time data, the public information of parking lots is easily obtained with little costs. What if we dig these free public data and use them in parking recommendation? Inspired by the classic Page rank algorithm used in web searching, this paper propose a novel method, named Parking rank, to sort the parking lots based on the public information totally, such as price, location, total spaces, etc. And a driving-cost sensitive recommendation method is presented in this paper. The simulation shows, when working with the Parking rank algorithm, the recommendation algorithm can help vehicles find the proper parking lots efficiently and reasonably, even in the urban district lack of any real-time information. In fact, Parking rank algorithm can be used as pre-processing of static ranking of city-wide parking lots, and can work with other recommendation algorithm related to real-time occupancy information, or work alone if there is only the public information about parking lots.