关键词: 加氢站, 截流选址模型, 不确定性, 机会约束, 二阶锥规划

Abstract: Hydrogen is a kind of secondary energy with abundant source, green and low-carbon, and wide application situations. It is gradually becoming one of the most important carriers for global energy decarbonation, and has great application potential in transportation, industry, power generation and energy storage. China has formulated the Medium and Long-term Plan for the Development of Hydrogen Energy Industry (2021—2035), pointing out that the number of hydrogen fuel cell vehicles should reach about 50,000 by the end of 2025, and deploys the construction of hydrogen refueling stations to build a more comprehensive hydrogen energy infrastructure network. In this context, how to maximize the utilization rate and service quality of hydrogen refueling station through scientific and reasonable location method has become a critical problem in the development of hydrogen industry.
Existing works to some extent ignored the uncertainty of traffic flow and customer behavior. To this end, this paper considering both the uncertainty of traffic flow and users driving behavior, introduced the chance constraint programming (CCP) theory and proposed a bi-level optimization model for hydrogen refueling station location problem. In the upper level, the leaders are hydrogen refueling station operators, aiming at capturing most traffic flow. In the lower level, the followers are the drivers, aiming at maximize the possibility of being served. The behavior of driver impacts the distribution of traffic flow and further impacts the location decision of hydrogen refueling station operators, thus forming a bi-level iterative problem. The bi-level decision iterates between each other and jointly determines the location scheme of hydrogen refueling station. Finally, the model was reconstructed as second order cone programming (SOCP) through KKT conditions and deterministic equivalence to achieve efficient solution.
A case study was conducted using the transportation network in the Yangtze River Delta region as an example. The results showed that: (1)Using probability distribution function to tackle traffic flow uncertainty, the proposed model reduced the sensitivity of traffic flow parameters. In the meanwhile, it considers the interactive feedback decision-making process between drivers and operators, making the planning results more reasonable. (2)As the budget of constructing hydrogen refueling station increases, the growth rate of the captured traffic flow will slow down to a plain, thus the construction budget should not be expanded blindly. (3)As the confidence level increases, the hydrogen refueling station operators will behave more conservative, they will only choose the path with higher traffic flow for the construction of hydrogen refueling station, which will cause low utilization rate of constructions, resulting in a reduction in the total captured traffic flow. Therefore, the operator of hydrogen refueling station needs to balance various factors, so as to meet driver’s demand for to the greatest extent.

Key words: hydrogen refueling station, flow capturing location model, uncertainty, chance constraints, second order cone programming