Exploring vehicle allocation strategies for restoring balance of dockless bike-sharing systems at night

In China, bike-sharing systems have already expanded to over 360 cities, with an average daily travel distance of 47 million kilometers. Consequently, the traffic congestion delay index has on average decreased by 2.2%, with a more pronounced effect on weekdays compared to weekends.

Nonetheless, making accurate decisions regarding urban dockless bicycle sharing systems (UDBSS) demand in different city locations is crucial, as incorrect choices can worsen transportation problems, causing difficulties in finding bicycles or excessive deployments leading to disorderly accumulation (Figure 1).

“To address this decision-making challenge, it is essential to consider uncertain factors like daily weather, temperature and workdays,” explains Chao Zhang, a professor at Shanxi University, who led a study on this area.

Zhang, together with a team of researchers from China, Mexico and Ireland, outlines a new vehicle allocation approach they have developed in a study published in the KeAi journal International Journal of Cognitive Computing in Engineering.

“Our three-way classification method and behavioral decision theory can individually classify all sites, and further derive the ranking of vehicle demands,” shares Zhang. “Reasonably forecasting the demand for shared bicycles at different stations helps to better leverage the entire system's role in alleviating traffic pressure. Additionally, it reduces the likelihood of resource waste.

Notably, the three-way classification method offers a cognitive attitude to decision-makers, that is, ‘no-commitment’ when they are faced with complex problems and cannot make an attitude of ‘accept’ or ‘reject’. It changes the two classification methods into three classification methods, which is closer to the cognitive model of human beings.

Furthermore, the three-way decision-making method can provide a separate space for stations that cannot be definitively classified at the moment, rather than forcibly categorizing them into a certain class, effectively enhancing the fault tolerance capability of the model.

"Considering the unique advantage of behavioral decision theory in portraying the mental state of decision-makers, integrating it into the three-way decision-making method can yield classification results that are more in line with reality,” adds Zhang.

Contact author: Chao Zhang, School of Computer and Information Technology, Shanxi University, Taiyuan, 030006, China. Email: czhang@sxu.edu.cn.

Funder: This paper was supported in part by the National Natural Science Foundation of China (62272284; 12201518; 62072294; 61972238), the Graduate Education Innovation Programs of Shanxi Province (2022Y147), the Special Fund for Science and Technology Innovation Teams of Shanxi (202204051001015), the Science and Technology Research Program of Chongqing Education Commission (KJQN202100206; KJQN202100205), the Training Program for Young Scientific Researchers of Higher Education Institutions in Shanxi, the Cultivate Scientific Research Excellence Programs of Higher Education Institutions in Shanxi (CSREP) (2019SK036), and the Research Project Supported by Shanxi Scholarship Council of China (2022-007).

Conflict of interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

See the article: Chao Zhang, et al., UDBSS static rebalancing model based on multi-granularity probabilistic rough sets under behavioral decision theory, International Journal of Cognitive Computing in Engineering in Engineering, Volume 5, 2024, Pages 27-43, https://doi.org/10.1016/j.ijcce.2024.01.001.