| 1. |
- Lu, Yizhou, et al.
(author)
-
Data- driven decentralized volt/var control for smart PV inverters in distribution systems
- 2022
-
In: 2022 24Th European Conference On Power Electronics And Applications (EPE'22 ECCE EUROPE). - : IEEE.
-
Conference paper (peer-reviewed)abstract
- The growing penetration of renewable energy sources (RES) in modern grids may result in severe voltage violation problems due to high stochastic features. Conventional centralized approaches could provide optimal solutions for voltage regulation while with great communication burdens. Control methods based on local information usually have non-optimal results and cannot always guarantee voltage security. This paper proposes a neural network-based decentralized strategy for volt/var control using inverter reactive power capacity. Learning from optimal power flow (OPF) results of historical data, the developed controller can provide optimal results approximate to centralized solutions and outperform local control methods in minimizing the power loss. The proposed method is tested on the IEEE 33-bus system and simulation results illustrate the effectiveness in voltage regulation and loss minimization.
|
|
| 2. |
- Mao, Jia, et al.
(author)
-
CNN and LSTM based Data-driven Cyberattack Detection for Grid-connected PV Inverter
- 2022
-
In: IEEE International Conference on Control and Automation, ICCA. - : Institute of Electrical and Electronics Engineers (IEEE). ; , s. 704-709
-
Conference paper (peer-reviewed)abstract
- Growing penetration of renewables comes with increased cyber security threat due to inherent low inertia characteristic and sophisticated control and communication networks of power electronics. This paper proposes a data-driven cyberattack detection strategy for grid-connected photovoltaic (PV) inverters. Ideas of long short term memory (LSTM) and convolutional neural network (CNN) as the core of detection achieve time series classification to diagnose the target and mode of cyberattack. Input de-redundancy and hyperparameter selection are conducted to optimize the detection. Meanwhile, well-designed cyberattack toolboxes of false data injection (FDI), denial-of-service (DoS) and delay are applied upon the communication of both sampled signals and issued commands in a grid-connected inverter model. By observing system performance via electrical measurements, this case study evaluates the LSTM, CNN-LSTM and convolutional LSTM based detection and obtains stable high quality of classification.
|
|
| 3. |
- Weiss, Xavier, et al.
(author)
-
Energy Management of Smart Homes with Electric Vehicles Using Deep Reinforcement Learning
- 2022
-
In: 2022 24th european conference on power electronics and applications (EPE'22 ECCE europe). - : IEEE.
-
Conference paper (peer-reviewed)abstract
- The proliferation of electric vehicles (EVs) has resulted in new charging infrastructure at all levels, including domestically. These new domestic EVs can potentially provide vehicle to home (V2H) services where EVs are used as energy storage systems (ESSs) for the home when they are not in use. Energy management systems (EMSs) can control these EVs to minimize the electricity cost to the owner but must satisfy constraints. Uncertainty in EV availability and the microgrid environment is also a challenge and can be addressed through real-time operation. Hence this paper formulates the EV charge/discharge scheduling problem as a Markov Decision Process (MDP). A safe implementation of Proximal Policy Optimization (PPO) is proposed for real-time optimization and compared to a day-ahead Mixed Integer Linear Programming (MILP) benchmark. The resulting PPO agent is able to minimize RA and SD costs for a typical EV user 3% better than the MILP solution. It obtains a 39% higher electricity cost than MILP, but unlike MILP does not require accurate forecasting data and operates in real-time.
|
|
| 4. |
- Yang, Jianfei, et al.
(author)
-
EfficientFi : Towards Large-Scale Lightweight WiFi Sensing via CSI Compression
- 2022
-
In: IEEE Internet of Things Journal. - : Institute of Electrical and Electronics Engineers (IEEE). - 2327-4662. ; 9:15, s. 13086-13095
-
Journal article (peer-reviewed)abstract
- WiFi technology has been applied to various places due to the increasing requirement of high-speed Internet access. Recently, besides network services, WiFi sensing is appealing in smart homes since it is device-free, cost-effective and privacy-preserving. Though numerous WiFi sensing methods have been developed, most of them only consider single smart home scenario. Without the connection of powerful cloud server and massive users, large-scale WiFi sensing is still difficult. In this paper, we firstly analyze and summarize these obstacles, and propose an efficient large-scale WiFi sensing framework, namely EfficientFi. The EfficientFi works with edge computing at WiFi APs and cloud computing at center servers. It consists of a novel deep neural network that can compress fine-grained WiFi Channel State Information (CSI) at edge, restore CSI at cloud, and perform sensing tasks simultaneously. A quantized auto-encoder and a joint classifier are designed to achieve these goals in an end-to-end fashion. To the best of our knowledge, the EfficientFi is the first IoT-cloud-enabled WiFi sensing framework that significantly reduces communication overhead while realizing sensing tasks accurately. We utilized human activity recognition and identification via WiFi sensing as two case studies, and conduct extensive experiments to evaluate the EfficientFi. The results show that it compresses CSI data from 1.368Mb/s to 0.768Kb/s with extremely low error of data reconstruction and achieves over 98% accuracy for human activity recognition.
|
|
| 5. |
- Zhang, Mengfan, et al.
(author)
-
Decentralized Coordination and Stabilization of Hybrid Energy Storage Systems in DC Microgrids
- 2022
-
In: IEEE Transactions on Smart Grid. - : Institute of Electrical and Electronics Engineers (IEEE). - 1949-3053 .- 1949-3061. ; 13:3, s. 1751-1761
-
Journal article (peer-reviewed)abstract
- Hybrid energy storage system (HESS) is an attractive solution to compensate power balance issues caused by intermittent renewable generations and pulsed power load in DC microgrids. The purpose of HESS is to ensure optimal usage of heterogeneous storage systems with different characteristics. In this context, power allocation for different energy storage units is a major concern. At the same time, the wide integration of power electronic converters in DC microgrids would possibly cause the constant power load instability issue. This paper proposes a composite model predictive control based decentralized dynamic power sharing strategy for HESS. First, a composite model predictive controller (MPC) is proposed for a system with a single ESS and constant power loads (CPLs). It consists of a baseline MPC for optimized transient performance and a sliding mode observer to estimate system disturbances. Then, a coordinated scheme is developed for HESS by using the proposed composite MPC with a virtual resistance droop controller for the battery system and with a virtual capacitance droop controller for the supercapacitor (SC) system. With the proposed scheme, the battery only supplies smooth power at steady state, while the SC compensates all the fast fluctuations. The proposed scheme achieves a decentralized dynamic power sharing and optimized transient performance under large variation of sources and loads. The proposed approach is verified by simulations and experiments.
|
|
| 6. |
- Zhang, Mengfan, et al.
(author)
-
Multi-Agent Deep Reinforcement Learning for Decentralized Voltage-Var Control in Distribution Power System
- 2022
-
In: 2022 IEEE Energy Conversion Congress and Exposition, ECCE 2022. - : Institute of Electrical and Electronics Engineers (IEEE). - 9781728193885
-
Conference paper (peer-reviewed)abstract
- With the large integration of renewables, the traditional power system becomes more sustainable and effective. Yet, the fluctuation and uncertainties of renewables have led to large challenges to the voltage stability in distribution power systems. This paper proposes a multi-agent deep reinforcement learning method to address the issue. The voltage control issue of the distribution system is modeled as the Markov Decision Process, while each grid-connected interface inverter of renewables is modeled as a deep neural network (DNN) based agent. With the designed reward function, the agents will interact with and seek for the optimal coordinated voltage-var control strategy. The offline-trained agents will execute online in a decentralized way to guarantee the voltage stability of the distribution without any extra communication. The proposed method can effectively achieve a communication-free and accurate voltage-var control of the distribution system under the uncertainties of renewables. The case study based on IEEE 33-bus system is demonstrated to validate the method.
|
|
| 7. |
- Zhang, Yuanzhi, et al.
(author)
-
Twin delayed deep deterministic policy gradient-based deep reinforcement learning for energy management of fuel cell vehicle integrating durability information of powertrain
- 2022
-
In: Energy Conversion and Management. - : Elsevier BV. - 0196-8904 .- 1879-2227. ; 274, s. 116454-
-
Journal article (peer-reviewed)abstract
- Deep reinforcement learning (DRL)-based energy management strategy (EMS) is attractive for fuel cell vehicle (FCV). Nevertheless, the fuel economy and lifespan durability of proton exchange membrane fuel cell (PEMFC) stack and lithium-ion battery (LIB) may not be synchronously optimized since transient degradation variations of PEMFC stack and LIB are not generally regarded for DRL-based EMSs. Furthermore, the inappropriate action space and the overestimated value function of DRL can lead to suboptimal EMS for on-line control. To this end, the objective of this research endeavors to formulate a twin delayed deep deterministic policy gradient (TD3)based EMS integrating durability information of PEMFC stack and LIB, which can interact with the vehicle operating states to continuously control the hybrid powertrain and limit the overestimation of DRL value function for ensuring maximum multi-objective reward at each moment. Unlike traditional DRL-based EMSs, the multi-objective reward function for this study is enlarged to incorporate the hydrogen consumption, state of charge (SOC)-sustaining penalty and transient lifespan degradation information of PEMFC stack and LIB in offline training and on-line control. The results demonstrate that the proposed EMS can drastically lessen the training time and computational burden. Meanwhile, in contrast with deep Q-network (DQN)-based and deep deterministic policy gradient (DDPG)-based EMSs in the various real-world urban and standard driving cycles, the proposed EMS can achieve hydrogen abatement at least 9.76% and 1.07%, and slow down total powertrain degradation at least 9.11% and 2.62%, respectively.
|
|
