A5010840650- Smart Grid Energy Management
- Electric Vehicles and Infrastructure
- Advanced Battery Technologies Research
- Electric Power System Optimization
- Integrated Energy Systems Optimization
- Energy Load and Power Forecasting
- Microgrid Control and Optimization
- Optimal Power Flow Distribution
- Building Energy and Comfort Optimization
- Thermal Analysis in Power Transmission
- Advanced MIMO Systems Optimization
- Caching and Content Delivery
- Prosthetics and Rehabilitation Robotics
- Satellite Communication Systems
- Advanced DC-DC Converters
- Advanced Wireless Network Optimization
- Power System Reliability and Maintenance
- Power System Optimization and Stability
- Muscle activation and electromyography studies
- Spacecraft and Cryogenic Technologies
- Vibration and Dynamic Analysis
- Machine Learning and ELM
- Hybrid Renewable Energy Systems
- Advanced Data and IoT Technologies
- Phase Change Materials Research
Shenyang University of Technology
2019-2023
Shanghai Jiao Tong University
2013-2020
With the increasing penetration of electric vehicles (EVs), orderly charging–discharging (C–D) strategy EVs can effectively alleviate impact large-scale grid-connected EVs. However, schedule C–D system EV clusters encounters curse dimensionality problem. In addition, performance control faces great challenges environmental uncertainty user demand and electricity price. This paper proposes an cluster scheduling considering real-time prices based on deep reinforcement learning. Firstly, we...
The rapid development of electric vehicles (EVs) makes the load vehicle charging stations (EVCSs) affect power grid. Aiming at low accuracy station forecasting caused by number EVs, temperature and electricity price, other factors, this paper proposes a method EVCSs based on combination multivariable residual correction grey model (EMGM) long short-term memory (LSTM) network. Firstly, influencing factors are analysed, theory is introduced into forecast EVCSs. role EMGM in taking account...
In the power sector, microgrids play a supportive role in bridging adequacy gap conventional electricity supply. Trading of generated energy has recently been improved by blockchain technology which offers new cheap, secure, and decentralized transaction approach. Its operation is however associated with an undesired inherent delay during transactions initiated prosumers, thus, failure to timely attend incidences urgent demand could end up catastrophe at consumer's side. This article thus...
The increase in the penetration rate of renewable energy (RE) and number electric vehicles (EVs) has created problems for balance supply demand power system. To solve this problem, it is necessary to evaluate flexibility system optimize scheduling flexible resources enhance This paper proposes a optimization strategy considering RE EVs. First, models EVs charging load are established. Second, analyzes evaluation indexes taking into account load. Finally, an optimal model improved particle...
Femtocell has been a valuable solution for cost-effective and high-capacity indoor cellular access. With rapid dense deployment of femtocells, inter-femtocell interference arises as one the major technical challenges. Conventional graph-based frequency reuse can only mitigate from individually strong interfering nodes, but not cumulative multiple nodes (the individual each be weak). In this work, we propose hypergraph-based scheme in femtocell networks. The takes into account by constructing...
Due to increasing load and characteristic stagnation fluctuations of existing generation systems capacity, the reliability assessment is crucial system adequacy. Furthermore, a rapid increase could amount consequent sudden deficit in supply before next scheduled assessment. Hence, conducted at regular close intervals ensure This study simulates establishes relationship between growth capacity using data IEEE test (IEEE RTS ‘96 standard). The states risk model were obtained sequential Monte...
In the collection and transmission of power big data, problem data missing exists. response to this problem, paper proposes a detection repair method based on SOM-LSTM. Firstly, large amount collected is analyzed type determined. Then, SOM used classify data. The LSTM trained according characteristic values different users complete types Finally, analysis actual in some regional loads China. Experimental results show that, compared with extreme learning machine (ELM) LSTM, proposed SOM-LSTM...
In practical application of microgrid cluster, the lack full detailed information cause failure dynamic modeling. Although some data-driven black-box modeling method can tackle this problem, insufficient usage prior known physical may reduce accuracy. To challenge, a hybrid physical-data-driven is proposed for behavior cluster. Motivated by equivalence recurrent neural network (RNN) and differential equations, differential-algebraic equations (DAEs) unknown part are represented gate unit...
This paper proposes a cyber–physical approach to enhance the prediction accuracy of electricity consumption solid electric thermal storage (SETS) system, which integrates physical model and data-based cyber model. In model, error is used as an input further calibrate error. Firstly, customers’ behavior characteristics are extracted by integration K-means one-versus-one support vector machine. Secondly, based on ambient temperature, developed predict daily consumption. Finally, levels...
The volatility and uncertainty of high-penetration renewable energy (RE) challenge the stability power system. To tackle this challenge, an optimal dispatch RE based on flexible resources (FRs) is proposed to enhance ability system cope with uncertain disturbances. Firstly, flexibility a integrated analyzed. margin supply adaptability are then introduced as evaluation indices for operation. Finally, multi-objective model enhancement FRs under constraint proposed. simulation results show that...
The internal parameters and topology of the power converter are unknown in some practical cases. Existing modeling methods based on impedance frequency scanning method can only guarantee that dynamic is effective at a single working point. To make established model wide range, an equivalent for LSTM Neural Network presented. At first, equivalence black-box problem deep loop neural network studied. Then, operating range by using proposed. Finally, simulation results under large disturbance...
In recent years, electric vehicles (EVs) have been widely used. A large number of EVs connected to the power grid will affect economy and stable operation grid. Load prediction is basis solve above problems. practical application, parameters traditional model are difficult obtain accurately calculation speed slow. order this problem, a combination method vehicle charging load based on Monte Carlo neural network proposed in paper. Firstly, built fit (EVCL) according user's behavior...