Parking Spaces Repurchase Strategy Design via Simulation Optimization

Abstract

Because of the rapid increase of vehicle numbers and land values, there is an increasing shortage in the supply of parking facilities in many cities. In many parking lots, residential drivers bought private parking spaces with high prices to ensure that they have spaces to park at any time. Recently, some parking lot operation companies plan to temporally repurchase a few private parking spaces back during certain time periods in a day and sell these places to public users to fully utilize the limited parking resources. How to choose the repurchase amounts and stopping time so as to maximize the profit then becomes an important problem. To solve this problem, a Gaussian mixture model is first proposed in this article to describe the time-varying arriving/departing behaviors of drivers and meanwhile the stochastic constraints of the profit maximization problem. Then, the expected optimal repurchase amounts and stopping time are estimated via simulation optimization. This new approach not only provides a useful statistical tool for parking spaces modeling but also overcomes several key limitations of current queuing based methodologies. Particularly, it emphasizes how to model drivers' behaviors on a small time-scale and explains the resulting benefits.