A5083061893- Electric Vehicles and Infrastructure
- Domain Adaptation and Few-Shot Learning
- Smart Grid Energy Management
- Transportation and Mobility Innovations
- Topic Modeling
- Advanced Neural Network Applications
- Advanced Graph Neural Networks
- Digital Imaging for Blood Diseases
- Optimal Power Flow Distribution
- Data Quality and Management
- Vehicle emissions and performance
- Enterobacteriaceae and Cronobacter Research
- Microgrid Control and Optimization
- Iron and Steelmaking Processes
- Green IT and Sustainability
- Transportation Safety and Impact Analysis
- Age of Information Optimization
- Integrated Energy Systems Optimization
- Astronomical Observations and Instrumentation
- Structural Response to Dynamic Loads
- Sensor Technology and Measurement Systems
- High-Velocity Impact and Material Behavior
- Advanced Battery Technologies Research
- Electric Power System Optimization
- Advanced Image and Video Retrieval Techniques
Northeastern University
2024
Beijing Jiaotong University
2020-2024
Ocean University of China
2024
Energy Research Institute
2024
Renewable Energy Systems (United States)
2023
PLA Army Engineering University
2016
Effectively utilizing large-scale Private Electric Vehicles (PREVs) in load restoration can enhance the capability of Distribution System Operators (DSOs) to respond accidental power outages. However, challenge remains incentivizing and guiding PREVs participate restoration. This paper proposes a Stackelberg game-based incentive mechanism, with V2G Station Operator (VSO) discharge guidance model based on Huff attraction embedded address this challenge. The VSO is introduced manage issue...
Deep Reinforcement Learning (DRL) is effective in solving complex, non-linear optimization problems, which particularly relevant energy management within Integrated Energy Systems (IESs). However, DRL approaches conventionally focus on single-objective policy learning, inadequate for the multi-objective tasks commonly encountered IESs management. To improve this, these typically combine multi-objectives, such as operating cost objective and safety into a single reward function using...
With the rapid development of information technology, electricity consumption Internet Data Centers (IDCs) increases drastically, resulting in considerable carbon emissions that need to be reduced urgently. In addition introduction Renewable Energy Sources (RES), joint use spatial migration capacity IDC workload and temporal flexibility demand Electric Vehicle Charging Stations (EVCSs) provides an important means change footprint IDC. this paper, a sustainability improvement strategy for...
With the increasing penetrations of electric vehicles (EV) and distributed renewable generations in power distribution systems, peak load EVs are overlapped with conventional systems creates curve, which challenges system operations. In this work, a deep reinforcement learning (DRL)-based EV charging optimization strategy is proposed. Firstly, problem formulated as Markov Decision Process (MDP) for DRL algorithms. Secondly, Deep Deterministic Policy Gradient (DDPG) algorithm implemented to...
After extreme events, the uninterrupted power supply to critical loads in distribution network faces significant challenges. Electric vehicles (EVs), as mobile energy storage units, can through Vehicle-to-Grid (V2G) technology, assisting restoration of after outages. A collaborative fault strategy is proposed considering spatiotemporal EVs supplying network, while also reconfiguration network. The probability selecting V2G stations and are characterized. Based on this, a load constructed,...
Entity alignment (EA) refers to the task of linking entities in different knowledge graphs (KGs). Existing EA methods rely heavily on structural isomorphism. However, real-world KGs, aligned usually have non-isomorphic neighborhood structures, which paralyses application these structure-dependent methods. In this paper, we investigate and tackle problem entity between heterogeneous KGs. First, propose two new benchmarks closely simulate scenarios heterogeneity. Then conduct extensive...
Abstract The rapid development of aircraft has created a strong demand for structural health monitoring, but current methods that rely on multiple sensor fusion suffer from complex hardware systems. Computational sensing with metastructures provides promising approach to reduce cost, the lack calibrated information makes it challenging identify impact regions. In this study, we propose concept spatial coding metastructure region recognition single sensor. Owing multi-order local resonance...
Entity alignment (EA) refers to the task of linking entities in different knowledge graphs (KGs). Existing EA methods rely heavily on structural isomorphism. However, real-world KGs, aligned usually have non-isomorphic neighborhood structures, which paralyses application these structure-dependent methods. In this paper, we investigate and tackle problem entity between heterogeneous KGs. First, propose two new benchmarks closely simulate scenarios heterogeneity. Then conduct extensive...
Wildlife is an important biological resource in China. Classifying images of wildlife through computer technology can help people identify wildlife, which great significance to understand and protect wildlife. Therefore, this issue worth studying. Traditional methods mostly use standard Convolutional Neural Networks (CNN) classify wild animal images, but these have disadvantages such as slow computing speed, long time consumption low accuracy. With attempt address issues, paper proposes a...
With the development of self-attention mechanism, Transformer model has demonstrated its outstanding performance in computer vision domain. However, massive computation brought from full attention mechanism became a heavy burden for memory consumption. Sequentially, limitation reduces possibility improving model. To remedy this problem, we propose novel economy named Couplformer, which decouples map into two sub-matrices and generates alignment scores spatial information. A series different...
Continual Learning aims to learn multiple incoming new tasks continually, and keep the performance of learned at a consistent level. However, existing research on continual learning assumes pose object is pre-defined well-aligned. For practical application, this work focuses pose-agnostic tasks, where object's changes dynamically unpredictably. The point cloud augmentation adopted from past approaches would sharply rise with task increment in process. To address problem, we inject...
Compared with other electric vehicles, taxis charge more frequently due to public service. The charging behavior and mobility characteristics of taxi clusters will make them one the flexible resources available for power grid operation scheduling, bring scheduling possibilities. is analyzed based on empty-loading ratio, traffic energy system model considering interaction vehicle, road, established fully reflect potential between grid. Considering mobile participation renewable a cross-areal...
The effective dispatch of the power system can meet severe challenges electric vehicles connecting to grid, for which mining demand-side scheduling potential provides reliable data support. At present, research on dispatching only focuses that have been connected has great limitations. We expand object road network area near charging station, introduces traffic flow parameters and establishes a vehicle prediction model based LSTM. same time, considering capacity limitation actual this paper...