The Allocation and Relocation of Rental Vehicles by the Rolling-Horizon Approach

• Main Authors: Yu-An Yuan, 袁雨安
• Other Authors: Yat-Wah Wan
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/784x3s

碩士 === 國立東華大學 === 運籌管理研究所 === 106 === The statistics from the Ministry of Transportation and Communications shows that self-owned cars and motorcycles are the main means of transportation in Taiwan. However, on average a car owner only uses his car for eight hours per week, i.e., while cars are on demand, most of the time they are idling. Thus, if there is a Short-Term Vehicle Rental Service, then there is no need for citizens to buy cars, which can save citizens much money on buying and maintaining cars. The Short-Term Vehicle Rental Service comes from the Carsharing concept. Many signals show that Carsharing will be the reality in the near future. This thesis discusses how to solve the vehicle imbalanced problem for one-way can rental for the Short-Term Vehicle Rental Service, ideally for self-driving vehicles. For one-way car rental, a customer can rent a car from one place and return it to another place. Also, for Short-Term Vehicle Rental Service, the car rental duration of customers can be measured in hour or even in minute. As cars are moved by customers to various places, the distribution of cars at rental spots may not match with the actual demand of cars. It is necessary to allocate and relocate cars to balance the supply and demand of cars so that customers can have cars to use. This thesis proposes four allocation and relocation methods (FIFO, RFIFO, MINRE, and RMINRE) to solve the problem. The profit and various types of cost of these methods are compared to those of the NMORE method, the naïve scheme that allocates but does not relocate cars. The results show that all four proposed allocation and relocation methods are better than NMORE. Among the four proposed methods, MINRE provides the best performance in relocation cost, total cost, and total profit.