A5101658440- Transportation and Mobility Innovations
- Smart Parking Systems Research
- Transportation Planning and Optimization
- Sharing Economy and Platforms
- Human Mobility and Location-Based Analysis
- Traffic Prediction and Management Techniques
- Safety and Risk Management
- Physics of Superconductivity and Magnetism
- Infrastructure Maintenance and Monitoring
- Sustainable Building Design and Assessment
- Electric Vehicles and Infrastructure
- Electrocatalysts for Energy Conversion
- Construction Project Management and Performance
- Risk and Safety Analysis
- Magnetic and transport properties of perovskites and related materials
- Hydrogen Storage and Materials
- Gaze Tracking and Assistive Technology
- Advanced Manufacturing and Logistics Optimization
- Advanced Decision-Making Techniques
- Bayesian Modeling and Causal Inference
- Metal Forming Simulation Techniques
- Maritime Navigation and Safety
- Autonomous Vehicle Technology and Safety
- Advanced Research in Science and Engineering
- Civil and Geotechnical Engineering Research
Didi Chuxing (China)
2017-2024
Tianjin University
2020
Shandong Management University
2017
Shandong University
2017
Tohoku Gakuin University
2015
Peking University
2014
Wuhan University
1988-2012
Huazhong University of Science and Technology
2009-2011
Early Warning (United States)
2011
Zhejiang University of Technology
2011
Taxi-booking apps have been very popular all over the world as they provide convenience such fast response time to users. The key component of a taxi-booking app is dispatch system which aims optimal matches between drivers and riders. Traditional systems sequentially taxis riders aim maximize driver acceptance rate for each individual order. However, traditional may lead low global success rate, degrades rider experience when using app. In this paper, we propose novel that attempts...
A fundamental question in any peer-to-peer ridesharing system is how to, both effectively and efficiently, dispatch user's ride requests to the right driver real time. Traditional rule-based solutions usually work on a simplified problem setting, which requires sophisticated hand-crafted weight design for either centralized authority control or decentralized multi-agent scheduling systems. Although recent approaches have used reinforcement learning provide combinatorial optimization...
How to optimally dispatch orders vehicles and how trade off between immediate future returns are fundamental questions for a typical ride-hailing platform. We model as large-scale parallel ranking problem study the joint decision-making task of order dispatching fleet management in online platforms. This brings unique challenges following four aspects. First, facilitate huge number act learn efficiently robustly, we treat each region cell an agent build multi-agent reinforcement learning...
Learning based order dispatching has witnessed tremendous success in ride hailing. However, the halts within individual hailing platforms because sharing raw data across may leak user privacy and business secrets. Such isolation not only impairs experience but also decreases potential revenues of platforms. In this paper, we advocate federated for cross-platform hailing, where multiple collaboratively make decisions without their local data. Realizing concept calls new learning strategies...
Foundation models have indeed made a profound impact on various fields, emerging as pivotal components that significantly shape the capabilities of intelligent systems. In context vehicles, leveraging power foundation has proven to be transformative, offering notable advancements in visual understanding. Equipped with multi-modal and multi-task learning capabilities, understanding (MM-VUFMs) effectively process fuse data from diverse modalities simultaneously handle driving-related tasks...
Improving the efficiency of dispatching orders to vehicles is a research hotspot in online ride-hailing systems. Most existing solutions for order-dispatching are centralized controlling, which require consider all possible matches between available and vehicles. For large-scale ride-sharing platforms, there thousands be matched at every second very high computational cost. In this paper, we propose decentralized execution method based on multi-agent reinforcement learning address problem....
Brain computer interface (BCI) is to provide a direct communication or control channel between human's brain and external devices. Recently, with the development of BCI, several research institutions organizations have developed some games that are interacted BCI enrich user's experience. However, most these only used in computers just focus on realization direction strategies. To meet potential increasing Android smartphone users' needs explore interacting game processes people's mental...
On-demand taxi-calling platforms often ignore the social engagement of individual drivers. The lack incentives impairs work enthusiasms drivers and will affect quality service. In this paper, we propose to form teams among promote participation. A team consists a leader multiple members, which acts as basis for various group-based such competition. We define Recommendation-based Team Formation (RTF) problem many possible while accounting choices RTF is challenging. It needs both accurate...
A fundamental question in any peer-to-peer ridesharing system is how to, both effectively and efficiently, dispatch user's ride requests to the right driver real time. Traditional rule-based solutions usually work on a simplified problem setting, which requires sophisticated hand-crafted weight design for either centralized authority control or decentralized multi-agent scheduling systems. Although recent approaches have used reinforcement learning provide combinatorial optimization...
Nowadays, ridesharing has become one of the most popular services offered by online ride-hailing platforms (e.g., Uber and Didi Chuxing). Existing adopt strategy that dispatches orders over entire city at a uniform time interval. However, uneven spatio-temporal order distributions in real-world systems indicate such an approach is suboptimal practice. Thus, this paper, we exploit adaptive dispatching intervals to boost platform's profit under guarantee maximum passenger waiting time....
Combining the spot investigation results, this paper draws fault tree of ship misfire emergency lock in Three Gorges to exploit inducing structure and countermeasures emergency; calculates minimum cut sets path then analyzes importance basic event by using calculation results minimal sets. The analysis indicate following: combustibles without isolation are main reason for misfire; such 5 types events as heat caused oxidation, high ambient temperature, unseasonable goods channels mixed...
This paper conducts the fault tree analysis on ship drift emergency of three Georges lock with method safety system engineering, draws a map analysis, representing intuitively basic events which can bring about top as well logic relationships between them, obtains minimum cut set, path set and structure importance through qualitative tree, points out possible ways to prevent occurrence accidents.
The mobility on demand (MoD) system relieves traffic pressure by simultaneously dispatching multiple orders to a vehicle via ridesharing. However, since the supply-demand distribution varies over time, existing methods for minimum fleet failed achieve long-term equilibrium, and thus greatly reduces order completion rate. In this paper, ElasticShare, ride-sharing dispatch method, is proposed maximize rate under dynamic distribution. First, we formalize ridesharing problem as an offline...
Mobility on demand (MoD) systems utilize ridesharing, i.e., multiple orders with high associating utility share a single vehicle, to reduce carbon footprint and alleviate traffic pressure. Existing methods mainly promote ridesharing by flocking the minimum required vehicles. However, supply-demand variations may aggregate undersupply in long run affect order completion rate. Meanwhile, it is difficult accurately estimate among lane-level features, such as flow. To fill this gap, we propose...
Foundation models have indeed made a profound impact on various fields, emerging as pivotal components that significantly shape the capabilities of intelligent systems. In context vehicles, leveraging power foundation has proven to be transformative, offering notable advancements in visual understanding. Equipped with multi-modal and multi-task learning capabilities, understanding (MM-VUFMs) effectively process fuse data from diverse modalities simultaneously handle driving-related tasks...
Achieving optimal order dispatching has been a long-standing challenge for online ride-hailing platforms. Early methods would make shortsighted matchings as they only consider prices alone the edge weights in driver-order bipartite graph, thus harming platform's revenue. To address this problem, recent works evaluate value of order's destination region to be long-term income driver could obtain average such and incorporate it into weight influence matching results. However, often result...
Nowadays, a great portion of researches research and industrial innovation is about the electric vehicles (EV) also EV Supply Equipment (EVSE) that play an important role in this context. EVSE requires standardization via effective communication protocols. In paper, we propose to customize existing Internet standard Routing Protocol for Low Power Lossy Networks (RPL) facilitate among networked EVSEs. RPL flexible protocol has special specifications support many low power lossy nodes, which...