An nonlinear model predictive controller (NMPC) is proposed in this paper for compensations of single line-to-ground (SLG) faults in resonant grounded power distribution networks (RGPDNs), which reduces the likelihood of power line bushfire due to electric faults. Residual current compensation (RCC) inverters with arc suppression coils (ASCs) in RGPDNs are controlled using the proposed NMPC to provide appropriate compensations during SLG faults. The proposed NMPC is incorporated with the estimation of ASC inductance, where the estimation is carried out based on voltage and current measurements from the neutral point of the distribution network. The compensation scheme is developed in the discrete time using the equivalent circuit of RGPDNs. The proposed NMPC for RCC inverters ensures that the desired current is injected into the neutral point during SLG faults, which is verified through both simulations and control hardware-in-the-loop (CHIL) validations. Comparative results are also presented against an integral sliding mode controller (ISMC) by demonstrating the capability of power line bushfire mitigation.
{"title":"Nonlinear Model Predictive Controller for Compensations of Single Line-to-Ground Fault in Resonant Grounded Power Distribution Networks","authors":"Warnakulasuriya Sonal Prashenajith Fernando;Mostafa Barzegar-Kalashani;Md Apel Mahmud;Shama Naz Islam;Nasser Hosseinzadeh","doi":"10.35833/MPCE.2023.000065","DOIUrl":"10.35833/MPCE.2023.000065","url":null,"abstract":"An nonlinear model predictive controller (NMPC) is proposed in this paper for compensations of single line-to-ground (SLG) faults in resonant grounded power distribution networks (RGPDNs), which reduces the likelihood of power line bushfire due to electric faults. Residual current compensation (RCC) inverters with arc suppression coils (ASCs) in RGPDNs are controlled using the proposed NMPC to provide appropriate compensations during SLG faults. The proposed NMPC is incorporated with the estimation of ASC inductance, where the estimation is carried out based on voltage and current measurements from the neutral point of the distribution network. The compensation scheme is developed in the discrete time using the equivalent circuit of RGPDNs. The proposed NMPC for RCC inverters ensures that the desired current is injected into the neutral point during SLG faults, which is verified through both simulations and control hardware-in-the-loop (CHIL) validations. Comparative results are also presented against an integral sliding mode controller (ISMC) by demonstrating the capability of power line bushfire mitigation.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 4","pages":"1113-1125"},"PeriodicalIF":5.7,"publicationDate":"2024-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10505131","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141769404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-18DOI: 10.35833/MPCE.2023.000565
Zifeng Zhang;Yuntao Ju
Networked microgrids (NMGs) are critical in the accommodation of distributed renewable energy. However, the existing centralized state estimation (SE) cannot meet the demands of NMGs in distributed energy management. The current estimator is also not robust against bad data. This study introduces the concepts of relative error to construct an improved robust SE (IRSE) optimization model with mixed-integer nonlinear programming (MINLP) that overcomes the disadvantage of inaccurate results derived from different measurements when the same tolerance range is considered in the robust SE (RSE). To improve the computation efficiency of the IRSE optimization model, the number of binary variables is reduced based on the projection statistics and normalized residual methods, which effectively avoid the problem of slow convergence or divergence of the algorithm caused by too many integer variables. Finally, an embedded consensus alternating direction of multiplier method (ADMM) distribution algorithm based on outer approximation (OA) is proposed to solve the IRSE optimization model. This algorithm can accurately detect bad data and obtain SE results that communicate only the boundary coupling information with neighbors. Numerical tests show that the proposed algorithm effectively detects bad data, obtains more accurate SE results, and ensures the protection of private information in all microgrids.
联网微电网（NMGs）对于适应分布式可再生能源至关重要。然而，现有的集中式状态估计（SE）无法满足分布式能源管理中的 NMGs 需求。目前的估计器对坏数据也不具有鲁棒性。本研究引入了相对误差的概念，利用混合整数非线性编程（MINLP）构建了改进的鲁棒状态估计（IRSE）优化模型，克服了鲁棒状态估计（RSE）在考虑相同容差范围时不同测量结果不准确的缺点。为了提高 IRSE 优化模型的计算效率，基于投影统计和归一化残差方法减少了二进制变量的数量，有效避免了因整数变量过多而导致的算法收敛慢或发散的问题。最后，提出了一种基于外近似（OA）的嵌入式共识交替乘法（ADMM）分布算法来求解 IRSE 优化模型。该算法能准确检测出不良数据，并获得只与邻域传递边界耦合信息的 SE 结果。数值测试表明，所提出的算法能有效检测坏数据，获得更准确的 SE 结果，并确保所有微电网中私人信息的保护。
{"title":"An Embedded Consensus ADMM Distribution Algorithm Based on Outer Approximation for Improved Robust State Estimation of Networked Microgrids","authors":"Zifeng Zhang;Yuntao Ju","doi":"10.35833/MPCE.2023.000565","DOIUrl":"10.35833/MPCE.2023.000565","url":null,"abstract":"Networked microgrids (NMGs) are critical in the accommodation of distributed renewable energy. However, the existing centralized state estimation (SE) cannot meet the demands of NMGs in distributed energy management. The current estimator is also not robust against bad data. This study introduces the concepts of relative error to construct an improved robust SE (IRSE) optimization model with mixed-integer nonlinear programming (MINLP) that overcomes the disadvantage of inaccurate results derived from different measurements when the same tolerance range is considered in the robust SE (RSE). To improve the computation efficiency of the IRSE optimization model, the number of binary variables is reduced based on the projection statistics and normalized residual methods, which effectively avoid the problem of slow convergence or divergence of the algorithm caused by too many integer variables. Finally, an embedded consensus alternating direction of multiplier method (ADMM) distribution algorithm based on outer approximation (OA) is proposed to solve the IRSE optimization model. This algorithm can accurately detect bad data and obtain SE results that communicate only the boundary coupling information with neighbors. Numerical tests show that the proposed algorithm effectively detects bad data, obtains more accurate SE results, and ensures the protection of private information in all microgrids.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 4","pages":"1217-1226"},"PeriodicalIF":5.7,"publicationDate":"2024-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10505132","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141769405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-18DOI: 10.35833/MPCE.2023.000893
Xu Yang;Haotian Liu;Wenchuan Wu;Qi Wang;Peng Yu;Jiawei Xing;Yuejiao Wang
As numerous distributed energy resources (DERs) are integrated into the distribution networks, the optimal dispatch of DERs is more and more imperative to achieve transition to active distribution networks (ADNs). Since accurate models are usually unavailable in ADNs, an increasing number of reinforcement learning (RL) based methods have been proposed for the optimal dispatch problem. However, these RL based methods are typically formulated without safety guarantees, which hinders their application in real world. In this paper, we propose an RL based method called supervisor-projector-enhanced safe soft actor-critic (S3AC) for the optimal dispatch of DERs in ADNs, which not only minimizes the operational cost but also satisfies safety constraints during online execution. In the proposed S3AC, the data-driven supervisor and projector are pre-trained based on the historical data from supervisory control and data acquisition (SCADA) system, effectively providing enhanced safety for executed actions. Numerical studies on several IEEE test systems demonstrate the effectiveness and safety of the proposed S3AC.
{"title":"Reinforcement Learning with Enhanced Safety for Optimal Dispatch of Distributed Energy Resources in Active Distribution Networks","authors":"Xu Yang;Haotian Liu;Wenchuan Wu;Qi Wang;Peng Yu;Jiawei Xing;Yuejiao Wang","doi":"10.35833/MPCE.2023.000893","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000893","url":null,"abstract":"As numerous distributed energy resources (DERs) are integrated into the distribution networks, the optimal dispatch of DERs is more and more imperative to achieve transition to active distribution networks (ADNs). Since accurate models are usually unavailable in ADNs, an increasing number of reinforcement learning (RL) based methods have been proposed for the optimal dispatch problem. However, these RL based methods are typically formulated without safety guarantees, which hinders their application in real world. In this paper, we propose an RL based method called supervisor-projector-enhanced safe soft actor-critic (S3AC) for the optimal dispatch of DERs in ADNs, which not only minimizes the operational cost but also satisfies safety constraints during online execution. In the proposed S3AC, the data-driven supervisor and projector are pre-trained based on the historical data from supervisory control and data acquisition (SCADA) system, effectively providing enhanced safety for executed actions. Numerical studies on several IEEE test systems demonstrate the effectiveness and safety of the proposed S3AC.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 5","pages":"1484-1494"},"PeriodicalIF":5.7,"publicationDate":"2024-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10505133","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-18DOI: 10.35833/MPCE.2023.000674
B. Vinod Kumar;Aneesa Farhan M A
The popularity of electric vehicles (EVs) has sparked a greater awareness of carbon emissions and climate impact. Urban mobility expansion and EV adoption have led to an increased infrastructure for electric vehicle charging stations (EVCSs), impacting radial distribution networks (RDNs). To reduce the impact of voltage drop, the increased power loss (PL), lower system interruption costs, and proper allocation and positioning of the EVCSs and capacitors are necessary. This paper focuses on the allocation of EVCS and capacitor installations in RDN by maximizing net present value (NPV), considering the reduction in energy losses and interruption costs. As a part of the analysis considering reliability, several compensation coefficients are used to evaluate failure rates and pinpoint those that will improve NPV. To locate the best nodes for EVCSs and capacitors, the hybrid of grey wolf optimization (GWO) and particle swarm optimization (PSO) (HGWO_PSO) and the hybrid of PSO and Cuckoo search (CS) (HPSO_CS) algorithms are proposed, forming a combination of GWO, PSO, and CS optimizations. The impact of EVCSs on NPV is also investigated in this paper. The effectiveness of the proposed optimization algorithms is validated on an IEEE 33-bus RDN.
{"title":"Optimal Simultaneous Allocation of Electric Vehicle Charging Stations and Capacitors in Radial Distribution Network Considering Reliability","authors":"B. Vinod Kumar;Aneesa Farhan M A","doi":"10.35833/MPCE.2023.000674","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000674","url":null,"abstract":"The popularity of electric vehicles (EVs) has sparked a greater awareness of carbon emissions and climate impact. Urban mobility expansion and EV adoption have led to an increased infrastructure for electric vehicle charging stations (EVCSs), impacting radial distribution networks (RDNs). To reduce the impact of voltage drop, the increased power loss (PL), lower system interruption costs, and proper allocation and positioning of the EVCSs and capacitors are necessary. This paper focuses on the allocation of EVCS and capacitor installations in RDN by maximizing net present value (NPV), considering the reduction in energy losses and interruption costs. As a part of the analysis considering reliability, several compensation coefficients are used to evaluate failure rates and pinpoint those that will improve NPV. To locate the best nodes for EVCSs and capacitors, the hybrid of grey wolf optimization (GWO) and particle swarm optimization (PSO) (HGWO_PSO) and the hybrid of PSO and Cuckoo search (CS) (HPSO_CS) algorithms are proposed, forming a combination of GWO, PSO, and CS optimizations. The impact of EVCSs on NPV is also investigated in this paper. The effectiveness of the proposed optimization algorithms is validated on an IEEE 33-bus RDN.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 5","pages":"1584-1595"},"PeriodicalIF":5.7,"publicationDate":"2024-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10505134","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142324270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-09DOI: 10.35833/MPCE.2023.000901
Lingxia Lu;Ju-Song Kang;Miao Yu
Non-intrusive load monitoring (NILM) can provide appliance-level power consumption information without deploying submeters for each load, in which load event detection is one of the crucial steps. However, the existing event detection methods do not efficiently detect both the starting time of an event (STE) and the ending time of an event (ETE), and their adaptability to scenarios with different sampling rates is limited. To address these problems, in this paper, an event detection method based on robust random cut forest (RRCF) algorithm, which is an unsupervised learning method for detecting anomalous data points within a dataset, is proposed. First, the mean-pooling preprocessing is applied to the aggregated load power series with a high sampling rate to minimize fluctuations. Then, the power differential series is obtained, and the anomaly score of each data point is calculated using the RRCF algorithm for preliminary detection. If an event has been preliminarily detected, misidentification caused by fluctuation will be further eliminated by using an adaptive power difference threshold approach. Finally, linear fitting is used to finely and accurately adjust the STE and ETE. The proposed method does not require any pretraining of the detection model and has been validated with both the BLUED dataset (with high and low sampling rates) and the REDD dataset (with low sampling rate). The experimental results demonstrate that the proposed method not only meets real-time requirements, but also exhibits strong adaptability across multiple scenarios. The precision is greater than 92% in distinct sampling rate scenarios, and the F1 score of phase B on the BLUED dataset reaches 94% in the scenario with a high sampling rate. These results indicate that the proposed method outperforms other state-of-the-art methods.
{"title":"Event Detection Based on Robust Random Cut Forest Algorithm for Non-Intrusive Load Monitoring","authors":"Lingxia Lu;Ju-Song Kang;Miao Yu","doi":"10.35833/MPCE.2023.000901","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000901","url":null,"abstract":"Non-intrusive load monitoring (NILM) can provide appliance-level power consumption information without deploying submeters for each load, in which load event detection is one of the crucial steps. However, the existing event detection methods do not efficiently detect both the starting time of an event (STE) and the ending time of an event (ETE), and their adaptability to scenarios with different sampling rates is limited. To address these problems, in this paper, an event detection method based on robust random cut forest (RRCF) algorithm, which is an unsupervised learning method for detecting anomalous data points within a dataset, is proposed. First, the mean-pooling preprocessing is applied to the aggregated load power series with a high sampling rate to minimize fluctuations. Then, the power differential series is obtained, and the anomaly score of each data point is calculated using the RRCF algorithm for preliminary detection. If an event has been preliminarily detected, misidentification caused by fluctuation will be further eliminated by using an adaptive power difference threshold approach. Finally, linear fitting is used to finely and accurately adjust the STE and ETE. The proposed method does not require any pretraining of the detection model and has been validated with both the BLUED dataset (with high and low sampling rates) and the REDD dataset (with low sampling rate). The experimental results demonstrate that the proposed method not only meets real-time requirements, but also exhibits strong adaptability across multiple scenarios. The precision is greater than 92% in distinct sampling rate scenarios, and the F1 score of phase B on the BLUED dataset reaches 94% in the scenario with a high sampling rate. These results indicate that the proposed method outperforms other state-of-the-art methods.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 6","pages":"2019-2029"},"PeriodicalIF":5.7,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10495845","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142844426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The integration of photovoltaic power generation is a new development into the traction power supply system (TPSS). However, traditional research on the TPSS operation strategy has not fully considered the risk of uncertainty in photovoltaic power output. To this end, we propose an operation strategy for the rail transit green energy system that considers the uncertainty risk of photovoltaic power output. First, we establish a regenerative braking energy utilization model that considers the impact of time-of-use (TOU) electricity price on the utilization efficiency and economic profit of regenerative braking energy and compensates for non-traction load. Then, we propose an operation strategy based on the balance of power supply and demand that uses an improved light robust (ILR) model to minimize the total cost of the rail transit green energy system, considering the risk of uncertainty in photovoltaic power output. The model incorporates the two-step load check on the second-level time scale to correct the operational results, solve the issue of different time resolutions between photovoltaic power and traction load, and achieve the coordinated optimization of risk cost and operation cost after photovoltaic integration. Case studies demonstrate that the proposed model can effectively consider the impact of the uncertainty in photovoltaic power output on the operation strategy, significantly improving the efficiency and economy of the system operation.
{"title":"Operation Strategy of Rail Transit Green Energy System Considering Uncertainty Risk of Photovoltaic Power Output","authors":"Yanbo Chen;Haoxin Tian;Guodong Zheng;Yuxiang Liu;Maja Grbić","doi":"10.35833/MPCE.2023.000788","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000788","url":null,"abstract":"The integration of photovoltaic power generation is a new development into the traction power supply system (TPSS). However, traditional research on the TPSS operation strategy has not fully considered the risk of uncertainty in photovoltaic power output. To this end, we propose an operation strategy for the rail transit green energy system that considers the uncertainty risk of photovoltaic power output. First, we establish a regenerative braking energy utilization model that considers the impact of time-of-use (TOU) electricity price on the utilization efficiency and economic profit of regenerative braking energy and compensates for non-traction load. Then, we propose an operation strategy based on the balance of power supply and demand that uses an improved light robust (ILR) model to minimize the total cost of the rail transit green energy system, considering the risk of uncertainty in photovoltaic power output. The model incorporates the two-step load check on the second-level time scale to correct the operational results, solve the issue of different time resolutions between photovoltaic power and traction load, and achieve the coordinated optimization of risk cost and operation cost after photovoltaic integration. Case studies demonstrate that the proposed model can effectively consider the impact of the uncertainty in photovoltaic power output on the operation strategy, significantly improving the efficiency and economy of the system operation.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 6","pages":"1859-1868"},"PeriodicalIF":5.7,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10495886","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142841998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As the phasor measurement unit (PMU) placement problem involves a cost-benefit trade-off, more PMUs get placed on higher-voltage buses. However, this leads to the fact that many lower-voltage levels of the bulk power system cannot be observed by PMUs. This lack of visibility then makes time-synchronized state estimation of the full system a challenging problem. In this paper, a deep neural network-based state estimator (DeNSE) is proposed to solve this problem. The DeNSE employs a Bayesian framework to indirectly combine the inferences drawn from slow-timescale but widespread supervisory control and data acquisition (SCADA) data with fast-timescale but selected PMU data, to attain sub-second situational awareness of the full system. The practical utility of the DeNSE is demonstrated by considering topology change, non-Gaussian measurement noise, and detection and correction of bad data. The results obtained using the IEEE 118-bus system demonstrate the superiority of the DeNSE over a purely SCADA state estimator and a PMU-only linear state estimator from a techno-economic viability perspective. Lastly, the scalability of the DeNSE is proven by estimating the states of a large and realistic 2000-bus synthetic Texas system.
{"title":"Deep Neural Network-Based State Estimator for Transmission System Considering Practical Implementation Challenges","authors":"Antos Cheeramban Varghese;Hritik Shah;Behrouz Azimian;Anamitra Pal;Evangelos Farantatos","doi":"10.35833/MPCE.2023.000997","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000997","url":null,"abstract":"As the phasor measurement unit (PMU) placement problem involves a cost-benefit trade-off, more PMUs get placed on higher-voltage buses. However, this leads to the fact that many lower-voltage levels of the bulk power system cannot be observed by PMUs. This lack of visibility then makes time-synchronized state estimation of the full system a challenging problem. In this paper, a deep neural network-based state estimator (DeNSE) is proposed to solve this problem. The DeNSE employs a Bayesian framework to indirectly combine the inferences drawn from slow-timescale but widespread supervisory control and data acquisition (SCADA) data with fast-timescale but selected PMU data, to attain sub-second situational awareness of the full system. The practical utility of the DeNSE is demonstrated by considering topology change, non-Gaussian measurement noise, and detection and correction of bad data. The results obtained using the IEEE 118-bus system demonstrate the superiority of the DeNSE over a purely SCADA state estimator and a PMU-only linear state estimator from a techno-economic viability perspective. Lastly, the scalability of the DeNSE is proven by estimating the states of a large and realistic 2000-bus synthetic Texas system.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 6","pages":"1810-1822"},"PeriodicalIF":5.7,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10495872","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142844441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-09DOI: 10.35833/MPCE.2023.000882
Hang Shuai;Buxin She;Jinning Wang;Fangxing Li
This study investigates a safe reinforcement learning algorithm for grid-forming (GFM) inverter based frequency regulation. To guarantee the stability of the inverter-based resource (IBR) system under the learned control policy, a model-based reinforcement learning (MBRL) algorithm is combined with Lyapunov approach, which determines the safe region of states and actions. To obtain near optimal control policy, the control performance is safely improved by approximate dynamic programming (ADP) using data sampled from the region of attraction (ROA). Moreover, to enhance the control robustness against parameter uncertainty in the inverter, a Gaussian process (GP) model is adopted by the proposed algorithm to effectively learn system dynamics from measurements. Numerical simulations validate the effectiveness of the proposed algorithm.
{"title":"Safe Reinforcement Learning for Grid-forming Inverter Based Frequency Regulation with Stability Guarantee","authors":"Hang Shuai;Buxin She;Jinning Wang;Fangxing Li","doi":"10.35833/MPCE.2023.000882","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000882","url":null,"abstract":"This study investigates a safe reinforcement learning algorithm for grid-forming (GFM) inverter based frequency regulation. To guarantee the stability of the inverter-based resource (IBR) system under the learned control policy, a model-based reinforcement learning (MBRL) algorithm is combined with Lyapunov approach, which determines the safe region of states and actions. To obtain near optimal control policy, the control performance is safely improved by approximate dynamic programming (ADP) using data sampled from the region of attraction (ROA). Moreover, to enhance the control robustness against parameter uncertainty in the inverter, a Gaussian process (GP) model is adopted by the proposed algorithm to effectively learn system dynamics from measurements. Numerical simulations validate the effectiveness of the proposed algorithm.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"13 1","pages":"79-86"},"PeriodicalIF":5.7,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10495852","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143183965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-05DOI: 10.35833/MPCE.2023.000850
Bin Feng;Huating Xu;Gang Huang;Zhuping Liu;Chuangxin Guo;Zhe Chen
With the goal of low-carbon energy utilization, electric vehicles (EVs) and EV charging stations (EVCSs) are becoming increasingly popular. The economical operation strategy is always a primary concern for EVCSs, while users' behavior and operating data leakage problems in EVCSs have not been taken seriously. Herein, federated deep reinforcement learning, a privacy-preserving method, is applied to learn the optimal strategy for multiple EVCSs. However, it is prone to Byzantine attacks. It is urgent to achieve an economical operation strategy while preserving data privacy and defending against Byzantine attacks. Therefore, this paper proposes a Byzantine-resilient federated deep reinforcement learning (BR-FDRL) method to address these problems. First, the distributed EVCS data are utilized by the federated deep reinforcement learning to train an economical operation strategy while preserving privacy by only transmitting gradients. The sampling efficiency is enhanced by both federated learning and stochastically controlled stochastic gradient. Then, the Byzantine-resilient gradient filter (BRGF) designs two distance rules to keep malicious gradients out. The case study verifies the effectiveness of the proposed BRGF in resisting Byzantine attacks and the effectiveness of federated deep reinforcement learning in improving convergence speed and reward and preserving privacy. The resluts show that the BR-FDRL method minimizes the operation cost by an average of 35% compared with the rule-based method while meeting the state of charge demand as much as possible.
{"title":"Byzantine-Resilient Economical Operation Strategy Based on Federated Deep Reinforcement Learning for Multiple Electric Vehicle Charging Stations Considering Data Privacy","authors":"Bin Feng;Huating Xu;Gang Huang;Zhuping Liu;Chuangxin Guo;Zhe Chen","doi":"10.35833/MPCE.2023.000850","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000850","url":null,"abstract":"With the goal of low-carbon energy utilization, electric vehicles (EVs) and EV charging stations (EVCSs) are becoming increasingly popular. The economical operation strategy is always a primary concern for EVCSs, while users' behavior and operating data leakage problems in EVCSs have not been taken seriously. Herein, federated deep reinforcement learning, a privacy-preserving method, is applied to learn the optimal strategy for multiple EVCSs. However, it is prone to Byzantine attacks. It is urgent to achieve an economical operation strategy while preserving data privacy and defending against Byzantine attacks. Therefore, this paper proposes a Byzantine-resilient federated deep reinforcement learning (BR-FDRL) method to address these problems. First, the distributed EVCS data are utilized by the federated deep reinforcement learning to train an economical operation strategy while preserving privacy by only transmitting gradients. The sampling efficiency is enhanced by both federated learning and stochastically controlled stochastic gradient. Then, the Byzantine-resilient gradient filter (BRGF) designs two distance rules to keep malicious gradients out. The case study verifies the effectiveness of the proposed BRGF in resisting Byzantine attacks and the effectiveness of federated deep reinforcement learning in improving convergence speed and reward and preserving privacy. The resluts show that the BR-FDRL method minimizes the operation cost by an average of 35% compared with the rule-based method while meeting the state of charge demand as much as possible.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 6","pages":"1957-1967"},"PeriodicalIF":5.7,"publicationDate":"2024-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10494232","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142844556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fault diagnosis (FD) for offshore wind turbines (WTs) are instrumental to their operation and maintenance (O&M). To improve the FD effect in the very early stage, a condition monitoring based sample set mining method from super-visory control and data acquisition (SCADA) time-series data is proposed. Then, based on the convolutional neural network (CNN) and attention mechanism, an interpretable convolutional temporal-spatial attention network (CTSAN) model is proposed. The proposed CTSAN model can extract deep temporal-spatial features from SCADA time-series data sequentially by: ① a convolution feature extraction module to extract features based on time intervals; ② a spatial attention module to extract spatial features considering the weights of different features; and ③ a temporal attention module to extract temporal features considering the weights of intervals. The proposed CT-SAN model has the superiority of interpretability by exposing the deep temporal-spatial features extracted in a human-understandable form of the temporal-spatial attention weights. The effectiveness and superiority of the proposed CTSAN model are verified by real offshore wind farms in China.
{"title":"Fault Diagnosis Based on Interpretable Convolutional Temporal-Spatial Attention Network for Offshore Wind Turbines","authors":"Xiangjing Su;Chao Deng;Yanhao Shan;Farhad Shahnia;Yang Fu;Zhaoyang Dong","doi":"10.35833/MPCE.2023.000606","DOIUrl":"https://doi.org/10.35833/MPCE.2023.000606","url":null,"abstract":"Fault diagnosis (FD) for offshore wind turbines (WTs) are instrumental to their operation and maintenance (O&M). To improve the FD effect in the very early stage, a condition monitoring based sample set mining method from super-visory control and data acquisition (SCADA) time-series data is proposed. Then, based on the convolutional neural network (CNN) and attention mechanism, an interpretable convolutional temporal-spatial attention network (CTSAN) model is proposed. The proposed CTSAN model can extract deep temporal-spatial features from SCADA time-series data sequentially by: ① a convolution feature extraction module to extract features based on time intervals; ② a spatial attention module to extract spatial features considering the weights of different features; and ③ a temporal attention module to extract temporal features considering the weights of intervals. The proposed CT-SAN model has the superiority of interpretability by exposing the deep temporal-spatial features extracted in a human-understandable form of the temporal-spatial attention weights. The effectiveness and superiority of the proposed CTSAN model are verified by real offshore wind farms in China.","PeriodicalId":51326,"journal":{"name":"Journal of Modern Power Systems and Clean Energy","volume":"12 5","pages":"1459-1471"},"PeriodicalIF":5.7,"publicationDate":"2024-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10494233","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142324322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：