Article
Computer Science, Information Systems
Lihui Xie, Junnan Wu, Yanying Li, Qiuye Sun, Lei Xi
Summary: The integrated energy system based on the ubiquitous power Internet of Things (IoT) has the characteristics of ubiquitous connection, complex energy conversion, and unbalanced supply-demand relationship. In order to address the strong random disturbance problem and achieve optimal cooperative control, a novel deep reinforcement learning algorithm, the collaborative learning actor-critic strategy, is proposed. Simulation tests on the two-area and four-area integrated energy systems show that the algorithm efficiently solves the disturbance problem and demonstrates better convergence and generalization performance.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Green & Sustainable Science & Technology
Lei Xi, Lipeng Zhou, Yanchun Xu, Xi Chen
Summary: The increase in new energies and electric vehicles poses new challenges to the power grid, and an algorithm based on DQ(sigma, lambda) is proposed to improve the stability and performance of multi-area interconnected power systems.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2021)
Article
Computer Science, Information Systems
Ruijie Zhu, Shuning Wu, Lulu Li, Ping Lv, Mingliang Xu
Summary: This article introduces a context-aware multiagent control method based on broad reinforcement learning for traffic light control. In comparison with previous methods, it takes into consideration pedestrian waiting states and adjacent agent states, effectively alleviating traffic congestion.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Lei Xi, Junnan Wu, Yanchun Xu, Hongbin Sun
Summary: This article proposes a deep-reinforcement-learning-based three-network double-delay actor-critic control strategy for improving power grid control performance and achieving optimal coordinated control. Simulation studies demonstrate the strategy's excellent stability and learning ability.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Lei Xi, Haokai Li, Jizhong Zhu, Yanying Li, Shouxiang Wang
Summary: This article proposes an improved reinforcement learning algorithm to solve the problem of frequency instability in power systems caused by large-scale electric vehicles and wind power grid connection. The algorithm expands the exploration space using an optimistic initialization principle and integrates double Q-learning to address the over-estimation issue. Simulation results demonstrate that the proposed algorithm obtains the global optimal solution and outperforms other reinforcement learning algorithms in terms of control performance.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yu Wu, Xin Xi, Jieyue He
Summary: Feature engineering relies on domain knowledge and human intervention. To automate this process, this paper proposes a novel Automatic Feature Generation model based on Graph Structure Learning (AFGSL). AFGSL utilizes the adjacency matrix to model feature relationships and employs Q-learning to train the stacking interaction layers, achieving better performance.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Automation & Control Systems
Zhen Zhang, Yew-Soon Ong, Dongqing Wang, Binqiang Xue
Summary: The study introduces a new MARL algorithm, PGP algorithm, which achieves optimal joint strategy learning in games of identical interest. Theoretical analysis and experimental studies demonstrate that the PGP algorithm outperforms other MARL algorithms in cumulative reward and time steps.
IEEE TRANSACTIONS ON CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Monireh Abdoos
Summary: This article models urban traffic networks using fuzzy graphs and implements collective learning of related agent sets using Q-learning and function approximation methods. The relationship and effectiveness of collective learning methods are studied and compared to independent control modes for better results.
IEEE INTELLIGENT SYSTEMS
(2021)
Article
Energy & Fuels
Linfei Yin, Bin Zhang
Summary: This paper proposes a long-term smart generation control framework with a single time-scale to replace the conventional combined generation control framework with two time-scales, and introduces a time series generative adversarial network controller. Numerical simulation results demonstrate that the proposed controller achieves higher control performance and smaller economic cost in the long-term.
Article
Computer Science, Information Systems
Mateusz Orlowski, Pawel Skruch
Summary: This paper presents an approach for defining, solving, and implementing dynamic cooperative maneuver problems in autonomous driving applications. A reinforcement learning technique is applied to find a suboptimal policy. The trained policy has been successful in solving the cooperation problem in all scenarios and the positive effects of applying shared rewards between agents have been presented and studied. The results obtained in this work provide a window of opportunity for various automotive applications.
Article
Engineering, Electrical & Electronic
Weichao Zhang, Wanxing Sheng, Qing Duan, Hanyan Huang, Xiangwu Yan
Summary: As synchronous generators are being replaced by renewable energy sources, the frequency stability of power systems decreases due to the lack of inertial response and frequency regulation. Virtual synchronous generators (VSGs) allow inverter-based renewable sources to emulate traditional generators and participate in power and frequency control. This paper focuses on the automatic generation control (AGC) with virtual synchronous renewables (VSRs). It introduces a strategy for the participation of renewable sources in AGC and presents a solution strategy for a security-constrained economic dispatch (SCED) model considering VSR operation. The results show that VSRs can increase the penetration level of renewables and decrease operating costs compared to maximum power point tracking (MPPT) mode.
JOURNAL OF MODERN POWER SYSTEMS AND CLEAN ENERGY
(2023)
Article
Energy & Fuels
Faizan Hassan Hajam, Mairaj Ud-Din Mufti
Summary: This article proposes a method for AGC control in a deregulated power system using Ultrabattery in coordination with TCPS. Simulation studies demonstrate the effectiveness of the scheme in mitigating low magnitude oscillations and improving system performance index.
INTERNATIONAL JOURNAL OF ENERGY RESEARCH
(2021)
Article
Engineering, Electrical & Electronic
Lei Fan, Chaoyue Zhao, Guangyuan Zhang, Qiuhua Huang
Summary: This paper proposes a robust optimization based framework to measure system flexibility by considering the interaction between Economic Dispatch and Automatic Generation Control. By utilizing a cutting plane algorithm with reformulation technique, seven different indices of the system are obtained and the impacts of several system factors on system flexibility are studied numerically.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2022)
Article
Operations Research & Management Science
Ramin Ahadi, Wolfgang Ketter, John Collins, Nicolo Daina
Summary: This study focuses on the operational problem of shared autonomous electric vehicles, aiming to maximize fleet profit and service quality through advanced decision-making aids. It proposes a distributed approach and uses deep learning to enhance the effectiveness and scalability of the model. The model outperforms central static charging strategies and provides insights into the impacts of strategic decisions on fleet performance and charging policies.
TRANSPORTATION SCIENCE
(2022)
Article
Thermodynamics
Tuoyu Deng, Liang Tian, Bo Hu, Xinping Liu, Jizhen Liu, Guiping Zhou, Yuanzhu Zhao
Summary: A new method of using energy storage in district heating networks for AGC is proposed, which includes quantitative analysis of ES capacity and design of control scheme. The study effectively improves power generation flexibility and AGC performance of CHP plant through reasonable use of ES.
APPLIED THERMAL ENGINEERING
(2021)
Article
Green & Sustainable Science & Technology
Bo Yang, Jingbo Wang, Junting Wang, Hongchun Shu, Danyang Li, Chunyuan Zeng, Yijun Chen, Xiaoshun Zhang, Tao Yu
Summary: This study presents a robust fractional-order PID control approach for SCES system, which effectively handles nonlinearities and unmodeled dynamics using high-gain perturbation observer and fractional-order PID controller to achieve online compensation and enhance robustness.
JOURNAL OF CLEANER PRODUCTION
(2021)
Article
Engineering, Electrical & Electronic
Xiaoshun Zhang, Tian Tan, Bin Zhou, Tao Yu, Bo Yang, Xiaoming Huang
Summary: In a performance-based frequency regulation market, increasing number of controllable renewable energy will participate in automatic generation control (AGC) of interconnected power grids. This paper develops a new optimal mileage based AGC dispatch (OMD) to optimally distribute the real-time overall AGC dispatch command. The use of adaptive distributed auction-based algorithm (ADAA) helps in fast convergence towards high-quality dispatch scheme.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2021)
Article
Engineering, Electrical & Electronic
Cheng Lyu, Youwei Jia, Zhao Xu
Summary: This study introduces a decentralized economic dispatch method for microgrids that eliminates the need for communication, showing potential for achieving global consensus in practical cases while significantly reducing the overall communication burden.
IEEE TRANSACTIONS ON SMART GRID
(2021)
Article
Green & Sustainable Science & Technology
Qianwen Xu, Yan Xu, Zhao Xu, Lihua Xie, Frede Blaabjerg
Summary: This article proposes a hierarchically coordinated control scheme for DC MG clusters under uncertainty, optimizing power set points and droop curve coefficients simultaneously to minimize operating costs and transmission loss. The approach ensures information privacy and plug-and-play feature, featuring decentralized power sharing and distributed optimization.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2021)
Article
Engineering, Electrical & Electronic
Yonghui Liu, Yue Wang, Minghao Wang, Zhao Xu, Yang Peng, Mingxuan Li
Summary: A coordinated VSG control strategy is proposed to achieve both grid supporting and maximum PV power harvesting without increasing battery capacity, by segmenting the DC-link voltage level to differentiate the operations of converters.
IEEE JOURNAL ON EMERGING AND SELECTED TOPICS IN CIRCUITS AND SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Jiapeng Li, Guobing Song, Jifei Yan, Yujun Li, Zhao Xu
Summary: This paper presents a data-driven framework for fast and reliable DC fault detection and classification in MTDC systems. Extensive features are extracted using highly comparative time-series analysis (HCTSA) with clear physical interpretations, and valuable features for fault identification are selected using a greedy forward search. A softmax regression classifier (SRC) is proposed based on the reduced features to calculate the probability of each fault category with a minor online computational burden. Numerical simulations demonstrate the effectiveness of the proposed approach under different fault conditions and its robustness against noise corruptions and abnormal samplings.
IEEE TRANSACTIONS ON POWER DELIVERY
(2022)
Article
Engineering, Electrical & Electronic
Heng Liu, Minghao Wang, Jiayong Li, Xu Xu, Zhao Xu, Jingming Dou
Summary: This paper proposes a centralized control framework for the collaborative operation of multiple ES-1s, addressing the active and reactive power coupling issue in smart loads. By utilizing the power variations of deferrable loads, the framework achieves minimum reactive power flow in microgrids, reducing storage requirement and distribution loss.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Engineering, Multidisciplinary
Huayi Wu, Minghao Wang, Zhao Xu, Youwei Jia
Summary: This article proposes a graph attention enabled convolutional network (GAECN) to approximate probabilistic power flow (PPF) and address the state uncertainties in distribution power systems caused by complex correlations among renewable outputs.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2022)
Article
Green & Sustainable Science & Technology
Huayi Wu, Zhao Xu, Minghao Wang
Summary: To address the challenge of timely perceiving distribution system states in power grids with high renewables, an Unrolled Spatiotemporal Graph Convolutional Network (USGCN) is proposed. The USGCN considers the complex spatiotemporal correlations of renewable energy sources and captures spatial and temporal correlations for enhanced accuracy. It also employs node-embedding technique to reveal hidden nonlinear spatiotemporal correlations of RES outputs. Additionally, the USGCN stacks unrolled spatiotemporal convolutional layers to obtain effective ahead-of-time state forecasting results.
IEEE TRANSACTIONS ON SUSTAINABLE ENERGY
(2023)
Article
Computer Science, Information Systems
Huayi Wu, Zhao Xu, Youwei Jia, Xu Xu
Summary: In this paper, an adaptive distributed graph model is proposed for real-time outage identification in power systems. By utilizing novel Laplacian convolution and breadth walk operations, the model effectively addresses the challenges of limited measurements and noise to achieve accurate outage identification.
IEEE SYSTEMS JOURNAL
(2023)
Article
Automation & Control Systems
Huayi Wu, Zhao Xu, Jian Zhao, Songjian Chai
Summary: Due to limited monitoring and measurement devices, timely identification of distribution grid topology has been a challenge. Therefore, this article proposes a power grid topological generative adversarial network (Gridtopo-GAN) model to identify the distribution grid topology with limited measurements. The model efficiently handles large-scale systems with different topological configurations by leveraging the topology preserved node embedding architecture and the generative capability of GAN. Numerical simulations on various distribution systems demonstrate the effectiveness and efficiency of the proposed topology identification model.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Yufei He, Minghao Wang, Zhao Xu, Youwei Jia
Summary: Electric springs (ES) are suggested as a solution for voltage instability in low-voltage distribution networks (LVDN). However, existing control methods for ES assume predominantly inductive grid networks, which is flawed as LVDN has significantly resistive line impedances. To address this, a novel gamma control method is proposed to enhance the voltage regulation performance of ES in LVDN. A comprehensive steady-state model of the ES-based smart load considering different line impedances is developed and optimal operating regions are derived analytically. The proposed control incorporates a smart load model and enables adaptive control boundaries, avoiding suboptimal or positive-feedback operations in voltage regulation and mitigating voltage flickers with a hysteresis proportional integral (PI) controller.
IEEE TRANSACTIONS ON POWER ELECTRONICS
(2023)
Article
Engineering, Electrical & Electronic
Jiapeng Li, Yujun Li, Zhengchun Du, Zhao Xu, Zhaoyang Dong
Summary: In this article, it is discovered that a grid-connected virtual synchronous generator (VSG) with sufficient damping can achieve global stability irrespective of the fault clearing time. The dissipated energy induced by damping is derived from the energy conservation law, and a simple approach to achieving global stabilization is proposed by tuning the damping of the VSG. The effectiveness of the proposed global stability condition and damping tuning method of VSGs are validated through simulation studies.
IEEE TRANSACTIONS ON POWER DELIVERY
(2023)
Article
Engineering, Electrical & Electronic
Xiangke Li, Minghao Wang, Chaoyu Dong, Wentao Jiang, Zhao Xu, Xiaohua Wu, Hongjie Jia
Summary: The paper introduces paralleled BILCs for HMG, which provide a flexible and reliable power interaction between ac and dc subgrids with high power density. A DUC is proposed to achieve resilience reinforcement and global economic operation. The economic droop controls f(ac) - lambda(ac) and v(dc) - lambda(dc) are employed for ac DGs and dc DGs to decrease generation expenses, while coordinating the normalized ac subgrid's frequency and dc subgrid's voltage for economic power interaction.
IEEE TRANSACTIONS ON POWER ELECTRONICS
(2023)
Article
Engineering, Electrical & Electronic
Yonghui Liu, Yue Wang, Xiaokang Liu, Minghao Wang, Zhao Xu, Hang Liu
Summary: This article presents a power model for parallel grid-forming converters (GFCs) in current saturation mode (CSM) and investigates the steady-state angle stability in this context for the first time. The analysis shows that the stability of equilibrium points (EPs) under inductive load is opposite to that under capacitive load. The experimental results validate the proposed model and analysis, highlighting the significant risk of power supply failure for parallel GFCs in CSM and providing general guidance for their operation.
IEEE TRANSACTIONS ON POWER ELECTRONICS
(2023)
