Station-level short-term demand forecast of carsharing system via station-embedding-based hybrid neural network

Zhao, Feifei; Wang, Weiping; Sun, Huijun; Yang, Huinan; Wu, Jianjun

doi:10.1080/21680566.2021.1951885

Cited by 8 publications

(3 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [16], the authors have optimized vehicle relocation based on car pickup forecasting. However, some other research works have focused on forecasting car pickup and car return demand, as in the work of [59]. The authors proposed a stationembedding-based hybrid neural network, and based on a case study from China, the hourly demand forecasting error was reduced by about 56.5% compared to conventional forecasting methods.…”

Section: Future Research Directionsmentioning

confidence: 99%

Vehicle Relocation in One-Way Carsharing: A Review

Eliyan,

Kerbache

2024

Sustainability

View full text Add to dashboard Cite

Carsharing has become increasingly popular in recent years as a sustainable transportation solution, offering individuals access to shared vehicles on a short-term basis. One-way carsharing, in particular, presents unique challenges due to its flexible nature, allowing users to pick up and drop off vehicles at different locations within a designated service area. This flexibility increases the service ridership but comes at the expense of vehicle imbalance among the stations, as some stations may have excess vehicles while other stations have vehicle shortages. Therefore, carsharing companies need to decide on strategies to ensure a balanced distribution of vehicles among the stations. This is essential as unbalanced vehicle distribution can lead to an unavailability of vehicles when needed or, conversely, result in an increased number of unnecessary rebalancing trips, thereby exacerbating traffic congestion and environmental pollution. Such issues can potentially undermine the overall contribution of carsharing to urban sustainability. To this end, this paper reviews the vehicle imbalance problem that arises in this field and the solution algorithms that solve them.

show abstract

Section: Future Research Directionsmentioning

confidence: 99%

Vehicle Relocation in One-Way Carsharing: A Review

Eliyan,

Kerbache

2024

Sustainability

View full text Add to dashboard Cite

show abstract

“…The elbow method and the silhouette coefficient method are two commonly utilized methods for this purpose. The clustering effect is evaluated using two metrics, namely the sum of squared errors (SSE) and the silhouette coefficient (SC) (29). As shown in Figure 7, the SC exhibits a fluctuating pattern of decline, with the highest value achieved at a cluster number of ''2''.…”

Section: Metro Exit Numbermentioning

confidence: 99%

Medical Supplies Delivery Route Optimization under Public Health Emergencies Incorporating Metro-based Logistics System

Sun,

Gu,

Wan Fei Ma

et al. 2024

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

To effectively mitigate the spread of infections during public health crises, precise and timely distribution of medical supplies is crucial. This paper proposes the integration of the metro-based underground logistics system (M-ULS) into the delivery process to address the vehicle routing problem (VRP). Considering various factors related to the metro-based VRP in public health emergencies, we formulate a mixed-integer nonlinear function model that aims to minimize total delivery costs while maximizing the load factor of medical vans and the average demand index concerning demand urgency and satisfaction rate of medical centers. To achieve this, we develop an improved adaptive genetic algorithm (IAGA) that incorporates adaptive crossover, adaptive mutation, and an elitist strategy. A case study is conducted to verify and analyze the performance of the optimized model and solving algorithm. Extensive numerical experiments are carried out to assess the economical and efficient advantages of M-ULS and the computational efficiency and solution quality of IAGA. Sensitivity analysis is conducted to evaluate the impact of medical van capacity on the routes. Lastly, we consider inter-medical-center transfer (IMCT) to further enhance emergency health response.

show abstract

“…Of these, the level of simplification of data from resource clusters is the highest, followed by data from economic clusters and data from social clusters. The reason for the above phenomenon is that the data of resource agglomeration are more repetitive and have a lower impact on the agglomeration of specialized villages [ 32 ]; therefore, the level of simplification is greater. Data on economic agglomerations are closely related to specialized village agglomerations but are also simplified by fixed interest rates and long-term economic policies.…”

Section: Case Analysis Based On Neural Array and System Dynamics Modelmentioning

confidence: 99%

An Optimization Analysis Model of Tourism Specialized Villages Based on Neural Network and System Dynamics

Wang

Sui-ying

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

With the rapid development of tourism, professional tourism villages emerge one after another, which has become the focus of the tourism industry. At present, there are some problems in tourism professional villages, such as imperfect management and inaccurate prediction of future development, which affect the rational allocation of tourism resources. In order to improve the distribution of tourism resources and better predict the development of tourism professional villages, it is necessary to make comprehensive judgment and analysis, especially the analysis of indicators such as the prediction and development judgment of tourism professional villages. This paper discusses the optimization analysis of the agglomeration of tourism specialized villages by backpropagation (BP) neural network and system dynamics model, analyzes the system structure of the agglomeration factors of tourism specialized villages, and promotes the intelligent integration of the agglomeration factors. The development of clusters of professional villages promotes data integration among resources, economy, society, and other elements and presents the characteristics of big data. As the level of concentration of professional villages increases, the complexity of the associated factors also increases, which increases the difficulty and effectiveness of tourism analysis. In view of this situation, taking mountain tourism as the research object, this paper proposes an improved system dynamics model based on BP, extracts features from cross factor (resource, economic, and social) data, and optimizes the relationship between professional village agglomeration and various factors. The MATLAB simulation results show that based on the improved system dynamics analysis, the simplification rate of (resources, economy, and society) data can be controlled at more than 24%, the degree of agglomeration is more than 95%, and the construction time of the relationship map of related factors is less than 40 s. Therefore, the analysis method proposed in this paper is suitable for the calculation of the agglomeration of tourism professional villages in the mountain area and can meet the needs of the development of tourism professional villages in the mountain area.

show abstract

Station-level short-term demand forecast of carsharing system via station-embedding-based hybrid neural network

Cited by 8 publications

References 50 publications

Vehicle Relocation in One-Way Carsharing: A Review

Vehicle Relocation in One-Way Carsharing: A Review

Medical Supplies Delivery Route Optimization under Public Health Emergencies Incorporating Metro-based Logistics System

An Optimization Analysis Model of Tourism Specialized Villages Based on Neural Network and System Dynamics

Contact Info

Product

Resources

About