A number of contingencies simulated during dynamic security assessment do not generate unacceptable values of power system state variables, due to their small influence on system operation. Their exclusion from the se...A number of contingencies simulated during dynamic security assessment do not generate unacceptable values of power system state variables, due to their small influence on system operation. Their exclusion from the set of contingencies to be simulated in the security assessment would achieve a significant reduction in computation time. This paper defines a critical contingencies selection method for on-line dynamic security assessment. The selection method results from an off-line dynamical analysis, which covers typical scenarios and also covers various related aspects like frequency, voltage, and angle analyses among others. Indexes measured over these typical scenarios are used to train neural networks, capable of performing on-line estimation of a critical contingencies list according to the system state.展开更多
A static security assessment approach considering electro-thermal coupling of transmission lines is proposed in this paper. Combined with the dynamic thermal rating technology and energy forecasting, the approach can ...A static security assessment approach considering electro-thermal coupling of transmission lines is proposed in this paper. Combined with the dynamic thermal rating technology and energy forecasting, the approach can track both the electrical variables and transmission lines’ temperature varying trajectory under anticipated contingencies. Accordingly, it identifies the serious contingencies by transmission lines’ temperature violation rather than its power flow, in this case the time margin of temperature rising under each serious contingency can be provided to operators as warning information and some unnecessary security control can also be avoided. Finally, numerical simulations are carried out to testify the validity of the proposed approach.展开更多
Deep learning(DL)-based methods in pre-fault dynamic security assessment(DSA)have provided significant results,contributing to the safe operation of power systems.However,power systems often suffer from insuf-ficient,...Deep learning(DL)-based methods in pre-fault dynamic security assessment(DSA)have provided significant results,contributing to the safe operation of power systems.However,power systems often suffer from insuf-ficient,small,and imbalanced datasets,which significantly impact the performance of DL-based DSA models.Existing DSA frameworks typically operate as two-class black-box models,assessing only overall system security without providing insights into the causes of insecurity or identifying critical generators(CGs),and they fail to quantify prediction uncertainty.These challenges hinder the implementation of current methods in real-world power systems and reduce operators’confidence in them.To address these issues,this paper proposes an uncertainty-aware bi-level multitask learning framework based on transfer learning and SqueezeNet architec-ture.The framework assesses system security,identifies CGs during instability,and leverages fine-tuning of a pretrained SqueezeNet model to facilitate training with limited data.Additionally,evidential deep learning is incorporated to quantify classification uncertainty.Without relying on the complex and challenging data augmentation method,this framework uses a simple technique called optimal classification threshold determi-nation to mitigate the negative impact of imbalanced data on model performance.The optimal threshold is determined by maximizing the area under the receiver operating characteristic(ROC)curve.The application of the proposed method to the IEEE 118-bus system shows its strong performance.These results offer crucial technical insights for the implementation of DL-based DSA in real-world power systems.展开更多
In a high-risk sector,such as power system,trans parency and interpretability are key principles for effectively deploying artificial intelligence(AI)in control rooms.There fore,this paper proposes a novel methodology...In a high-risk sector,such as power system,trans parency and interpretability are key principles for effectively deploying artificial intelligence(AI)in control rooms.There fore,this paper proposes a novel methodology,the evolving sym bolic model(ESM),which is dedicated to generating highly in terpretable data-driven models for dynamic security assessment(DSA),namely in system security classification(SC)and the def inition of preventive control actions.The ESM uses simulated annealing for a data-driven evolution of a symbolic model tem plate,enabling different cooperative learning schemes between humans and AI.The Madeira Island power system is used to validate the application of the ESM for DSA.The results show that the ESM has a classification accuracy comparable to pruned decision trees(DTs)while boasting higher global inter pretability.Moreover,the ESM outperforms an operator-de fined expert system and an artificial neural network in defining preventive control actions.展开更多
The paper presents a practical dynamic security region (PDSR) based dynamic security risk assessment and optimization model for power transmission system. The cost of comprehensive security control and the influence o...The paper presents a practical dynamic security region (PDSR) based dynamic security risk assessment and optimization model for power transmission system. The cost of comprehensive security control and the influence of uncertainties of power injections are considered in the model of dynamic security risk assessment. The transient stability constraints and uncertainties of power injections can be considered easily by PDSR in form of hyper-box. A method to define and classify contingency set is presented, and a risk control optimization model is given which takes total dynamic insecurity risk as the objective function for a dominant con-tingency set. An optimal solution of dynamic insecurity risk is obtained by opti-mizing preventive and emergency control cost and contingency set decomposition. The effectiveness of this model has been proved by test results on the New Eng-land 10-genarator 39-bus system.展开更多
This letter proposes a reliable transfer learning(RTL)method for pre-fault dynamic security assessment(DSA)in power systems to improve DSA performance in the presence of potentially related unknown faults.It takes ind...This letter proposes a reliable transfer learning(RTL)method for pre-fault dynamic security assessment(DSA)in power systems to improve DSA performance in the presence of potentially related unknown faults.It takes individual discrepancies into consideration and can handle unknown faults with incomplete data.Extensive experiment results demonstrate high DSA accuracy and computational efficiency of the proposed RTL method.Theoretical analysis shows RTL can guarantee system performance.展开更多
New strategies and methods for assessing the security of protection systems to reduce the risk of unnecessary disturbances and blackouts are the main topic of the present paper. The system behavior of a protection sys...New strategies and methods for assessing the security of protection systems to reduce the risk of unnecessary disturbances and blackouts are the main topic of the present paper. The system behavior of a protection system and network is analyzed and assessed as a whole. Hence, the established algorithms are capable to handle complex network structures with regard to an intelligent data management as well as data validation. Protection security assessment comprised two different aspects: on the one hand the behavior regarding dependability and security in terms of speed and sensitivity, on the other hand the behavior regarding the response on dynamic network phenomena as voltage stability and transient stability. A new automated method for assessing the dependability and security of protection systems is shown. The short-circuit simulation tool is used to provide a simulation system including network and protection devices as a whole. The handling of the large amount of resulting data is done by an intelligent visualization method like a “fingerprint” analysis. Further on the paper is focused on the protection response on dynamic network phenomena and presents innovative strategies for this investigation aspect. The structure of simulation environment will be described. Results of a case study show the application of this method on a real network. The system tool which is concluding these two aspects of protection assessment is called SIGUARD? PSA.展开更多
Modern electric power grids face a variety of new challenges and there is an urgent need to improve grid resilience more than ever before. The best approach would be to focus primarily on the grid intelligence rather ...Modern electric power grids face a variety of new challenges and there is an urgent need to improve grid resilience more than ever before. The best approach would be to focus primarily on the grid intelligence rather than implementing redundant preventive measures. This paper presents the foundation for an intelligent operational strategy so as to enable the grid to assess its current dynamic state instantaneously. Traditional forms of real-time power system security assessment consist mainly of methods based on power flow analyses and hence, are static in nature. For dynamic security assessment, it is necessary to carry out time-domain simulations (TDS) that are computationally too involved to be performed in real-time. The paper employs machine learning (ML) techniques for real-time assessment of grid resiliency. ML techniques have the capability to organize large amounts of data gathered from such time-domain simulations and thereby extract useful information in order to better assess the system security instantaneously. Further, this paper develops an approach to show that a few operating points of the system called as landmark points contain enough information to capture the nonlinear dynamics present in the system. The proposed approach shows improvement in comparison to the case without landmark points.展开更多
Power systems transport an increasing amount of electricity,and in the future,involve more distributed renewables and dynamic interactions of the equipment.The system response to disturbances must be secure and predic...Power systems transport an increasing amount of electricity,and in the future,involve more distributed renewables and dynamic interactions of the equipment.The system response to disturbances must be secure and predictable to avoid power blackouts.The system response can be simulated in the time domain.However,this dynamic security assessment(DSA)is not computationally tractable in real-time.Particularly promising is to train decision trees(DTs)from machine learning as interpretable classifiers to predict whether the systemwide responses to disturbances are secure.In most research,selecting the best DT model focuses on predictive accuracy.However,it is insufficient to focus solely on predictive accuracy.Missed alarms and false alarms have drastically different costs,and as security assessment is a critical task,interpretability is crucial for operators.In this work,the multiple objectives of interpretability,varying costs,and accuracies are considered for DT model selection.We propose a rigorous workflow to select the best classifier.In addition,we present two graphical approaches for visual inspection to illustrate the selection sensitivity to probability and impacts of disturbances.We propose cost curves to inspect selection combining all three objectives for the first time.Case studies on the IEEE 68 bus system and the French system show that the proposed approach allows for better DT-selections,with an 80%increase in interpretability,5%reduction in expected operating cost,while making almost zero accuracy compromises.The proposed approach scales well with larger systems and can be used for models beyond DTs.Hence,this work provides insights into criteria for model selection in a promising application for methods from artificial intelligence(AI).展开更多
随着云计算、BYOD(Bring your own device)的流行,企业信息系统呈现出开放与动态互联的特征,这种趋势使得基于动态信任评估的零信任安全架构开始取代基于边界信任的一次性身份认证模式,成为工业界与学术界关注的研究热点。动态信任评估...随着云计算、BYOD(Bring your own device)的流行,企业信息系统呈现出开放与动态互联的特征,这种趋势使得基于动态信任评估的零信任安全架构开始取代基于边界信任的一次性身份认证模式,成为工业界与学术界关注的研究热点。动态信任评估模型为零信任架构提供持续信任评估的能力,可以对企业信息系统的安全性和隐私性进行有效的保护。然而,训练动态信任评估模型面临两个现实挑战:1)很多企业的用户异常登录行为数据很少,影响模型的训练效果,导致信任评估模型准确性不高,不利于身份认证系统的可靠性;2)用户行为数据中包含着用户的隐私信息,泄漏用户隐私的法律风险使得企业不愿意共享用户异常登录行为数据。针对这些问题,本文提出了一种基于联邦学习的动态信任评估身份认证方法,使得各个平台在不泄漏原始用户数据的情况下达到联合训练模型的目的,进而提高各平台身份认证系统的安全性。在假设各个平台提供了用户的行为原始数据的前提下,本方案会根据不同特征的实际含义提取离散型用户行为数据的统计学特征,并选取与风险用户相关性高的特征。为了保证数据安全性和训练数据的规模,本方法采用联邦学习技术联合多个企业进行训练,从而得到动态信任评估层的核心模型,其误识率和拒识率相较于单一平台有了一定的提升。通过该方案,身份认证系统可以在不泄露用户敏感信息的情况下,对用户身份进行有效评估,进而提升身份认证系统安全性和用户体验。本文还对不同的支持横向联邦学习的机器学习算法应用于动态信任评估模型的效果进行了比较,实验结果表明了在基于联邦学习的动态身份认证模型中使用SVM作为机器学习训练方法的效果优于其他机器学习训练方法。最后,本文从安全性和隐私性的角度出发还对动态信任评估系统自身以及联邦学习带来的安全性和隐私性的影响做了讨论。展开更多
文摘A number of contingencies simulated during dynamic security assessment do not generate unacceptable values of power system state variables, due to their small influence on system operation. Their exclusion from the set of contingencies to be simulated in the security assessment would achieve a significant reduction in computation time. This paper defines a critical contingencies selection method for on-line dynamic security assessment. The selection method results from an off-line dynamical analysis, which covers typical scenarios and also covers various related aspects like frequency, voltage, and angle analyses among others. Indexes measured over these typical scenarios are used to train neural networks, capable of performing on-line estimation of a critical contingencies list according to the system state.
文摘A static security assessment approach considering electro-thermal coupling of transmission lines is proposed in this paper. Combined with the dynamic thermal rating technology and energy forecasting, the approach can track both the electrical variables and transmission lines’ temperature varying trajectory under anticipated contingencies. Accordingly, it identifies the serious contingencies by transmission lines’ temperature violation rather than its power flow, in this case the time margin of temperature rising under each serious contingency can be provided to operators as warning information and some unnecessary security control can also be avoided. Finally, numerical simulations are carried out to testify the validity of the proposed approach.
基金The Imperial College London team was supported through the Reli-ability,Resilience and Defense technology for the grid(R2D2)Project co-funded by the Europe(EU)through the LCE Policy Support Program under Grant HORIZON-CL5-2021through the Competitiveness and Innovation Framework Program under Grant 101075714the EPSRC-funded programs“NetworkPlus-A green,connected and pros-perous Britain”under grant number EP/W034204/1.
文摘Deep learning(DL)-based methods in pre-fault dynamic security assessment(DSA)have provided significant results,contributing to the safe operation of power systems.However,power systems often suffer from insuf-ficient,small,and imbalanced datasets,which significantly impact the performance of DL-based DSA models.Existing DSA frameworks typically operate as two-class black-box models,assessing only overall system security without providing insights into the causes of insecurity or identifying critical generators(CGs),and they fail to quantify prediction uncertainty.These challenges hinder the implementation of current methods in real-world power systems and reduce operators’confidence in them.To address these issues,this paper proposes an uncertainty-aware bi-level multitask learning framework based on transfer learning and SqueezeNet architec-ture.The framework assesses system security,identifies CGs during instability,and leverages fine-tuning of a pretrained SqueezeNet model to facilitate training with limited data.Additionally,evidential deep learning is incorporated to quantify classification uncertainty.Without relying on the complex and challenging data augmentation method,this framework uses a simple technique called optimal classification threshold determi-nation to mitigate the negative impact of imbalanced data on model performance.The optimal threshold is determined by maximizing the area under the receiver operating characteristic(ROC)curve.The application of the proposed method to the IEEE 118-bus system shows its strong performance.These results offer crucial technical insights for the implementation of DL-based DSA in real-world power systems.
基金supported by the ENFIELD(European Lighthouse to Manifest Trustworthy and Green AI)project,European Union’s Horizon Research and Innovation Programme(No.101120657).
文摘In a high-risk sector,such as power system,trans parency and interpretability are key principles for effectively deploying artificial intelligence(AI)in control rooms.There fore,this paper proposes a novel methodology,the evolving sym bolic model(ESM),which is dedicated to generating highly in terpretable data-driven models for dynamic security assessment(DSA),namely in system security classification(SC)and the def inition of preventive control actions.The ESM uses simulated annealing for a data-driven evolution of a symbolic model tem plate,enabling different cooperative learning schemes between humans and AI.The Madeira Island power system is used to validate the application of the ESM for DSA.The results show that the ESM has a classification accuracy comparable to pruned decision trees(DTs)while boasting higher global inter pretability.Moreover,the ESM outperforms an operator-de fined expert system and an artificial neural network in defining preventive control actions.
基金Supported by the key research of the National Natural Science Foundation of China (Grant No. 50595413) The National Basic Research Program of China (973 Program) (Grant No. 2004CB217904)
文摘The paper presents a practical dynamic security region (PDSR) based dynamic security risk assessment and optimization model for power transmission system. The cost of comprehensive security control and the influence of uncertainties of power injections are considered in the model of dynamic security risk assessment. The transient stability constraints and uncertainties of power injections can be considered easily by PDSR in form of hyper-box. A method to define and classify contingency set is presented, and a risk control optimization model is given which takes total dynamic insecurity risk as the objective function for a dominant con-tingency set. An optimal solution of dynamic insecurity risk is obtained by opti-mizing preventive and emergency control cost and contingency set decomposition. The effectiveness of this model has been proved by test results on the New Eng-land 10-genarator 39-bus system.
基金supported by the Internal Talent Award(TRACS)with Wallenberg-NTU Presidential Postdoctoral Fellowship 2022the National Research Foundation,Singapore and DSO National Laboratories under the AI Singapore Program(AISG Award No:AISG2-RP-2020-019)+1 种基金the RIE 2020 Advanced Manufacturing and Engineering(AME)Programmatic Fund(No.A20G8b0102),SingaporeFuture Communications Research&Development Program(FCP-NTU-RG-2021-014).
文摘This letter proposes a reliable transfer learning(RTL)method for pre-fault dynamic security assessment(DSA)in power systems to improve DSA performance in the presence of potentially related unknown faults.It takes individual discrepancies into consideration and can handle unknown faults with incomplete data.Extensive experiment results demonstrate high DSA accuracy and computational efficiency of the proposed RTL method.Theoretical analysis shows RTL can guarantee system performance.
文摘New strategies and methods for assessing the security of protection systems to reduce the risk of unnecessary disturbances and blackouts are the main topic of the present paper. The system behavior of a protection system and network is analyzed and assessed as a whole. Hence, the established algorithms are capable to handle complex network structures with regard to an intelligent data management as well as data validation. Protection security assessment comprised two different aspects: on the one hand the behavior regarding dependability and security in terms of speed and sensitivity, on the other hand the behavior regarding the response on dynamic network phenomena as voltage stability and transient stability. A new automated method for assessing the dependability and security of protection systems is shown. The short-circuit simulation tool is used to provide a simulation system including network and protection devices as a whole. The handling of the large amount of resulting data is done by an intelligent visualization method like a “fingerprint” analysis. Further on the paper is focused on the protection response on dynamic network phenomena and presents innovative strategies for this investigation aspect. The structure of simulation environment will be described. Results of a case study show the application of this method on a real network. The system tool which is concluding these two aspects of protection assessment is called SIGUARD? PSA.
文摘Modern electric power grids face a variety of new challenges and there is an urgent need to improve grid resilience more than ever before. The best approach would be to focus primarily on the grid intelligence rather than implementing redundant preventive measures. This paper presents the foundation for an intelligent operational strategy so as to enable the grid to assess its current dynamic state instantaneously. Traditional forms of real-time power system security assessment consist mainly of methods based on power flow analyses and hence, are static in nature. For dynamic security assessment, it is necessary to carry out time-domain simulations (TDS) that are computationally too involved to be performed in real-time. The paper employs machine learning (ML) techniques for real-time assessment of grid resiliency. ML techniques have the capability to organize large amounts of data gathered from such time-domain simulations and thereby extract useful information in order to better assess the system security instantaneously. Further, this paper develops an approach to show that a few operating points of the system called as landmark points contain enough information to capture the nonlinear dynamics present in the system. The proposed approach shows improvement in comparison to the case without landmark points.
基金The authors were supported by a scholarship funded by the Nige-rian National Petroleum Corporation,NNPC,the TU Delft AI Labs Programme,NL,and the research project IDLES,UK(EP/R045518/1).
文摘Power systems transport an increasing amount of electricity,and in the future,involve more distributed renewables and dynamic interactions of the equipment.The system response to disturbances must be secure and predictable to avoid power blackouts.The system response can be simulated in the time domain.However,this dynamic security assessment(DSA)is not computationally tractable in real-time.Particularly promising is to train decision trees(DTs)from machine learning as interpretable classifiers to predict whether the systemwide responses to disturbances are secure.In most research,selecting the best DT model focuses on predictive accuracy.However,it is insufficient to focus solely on predictive accuracy.Missed alarms and false alarms have drastically different costs,and as security assessment is a critical task,interpretability is crucial for operators.In this work,the multiple objectives of interpretability,varying costs,and accuracies are considered for DT model selection.We propose a rigorous workflow to select the best classifier.In addition,we present two graphical approaches for visual inspection to illustrate the selection sensitivity to probability and impacts of disturbances.We propose cost curves to inspect selection combining all three objectives for the first time.Case studies on the IEEE 68 bus system and the French system show that the proposed approach allows for better DT-selections,with an 80%increase in interpretability,5%reduction in expected operating cost,while making almost zero accuracy compromises.The proposed approach scales well with larger systems and can be used for models beyond DTs.Hence,this work provides insights into criteria for model selection in a promising application for methods from artificial intelligence(AI).
文摘随着云计算、BYOD(Bring your own device)的流行,企业信息系统呈现出开放与动态互联的特征,这种趋势使得基于动态信任评估的零信任安全架构开始取代基于边界信任的一次性身份认证模式,成为工业界与学术界关注的研究热点。动态信任评估模型为零信任架构提供持续信任评估的能力,可以对企业信息系统的安全性和隐私性进行有效的保护。然而,训练动态信任评估模型面临两个现实挑战:1)很多企业的用户异常登录行为数据很少,影响模型的训练效果,导致信任评估模型准确性不高,不利于身份认证系统的可靠性;2)用户行为数据中包含着用户的隐私信息,泄漏用户隐私的法律风险使得企业不愿意共享用户异常登录行为数据。针对这些问题,本文提出了一种基于联邦学习的动态信任评估身份认证方法,使得各个平台在不泄漏原始用户数据的情况下达到联合训练模型的目的,进而提高各平台身份认证系统的安全性。在假设各个平台提供了用户的行为原始数据的前提下,本方案会根据不同特征的实际含义提取离散型用户行为数据的统计学特征,并选取与风险用户相关性高的特征。为了保证数据安全性和训练数据的规模,本方法采用联邦学习技术联合多个企业进行训练,从而得到动态信任评估层的核心模型,其误识率和拒识率相较于单一平台有了一定的提升。通过该方案,身份认证系统可以在不泄露用户敏感信息的情况下,对用户身份进行有效评估,进而提升身份认证系统安全性和用户体验。本文还对不同的支持横向联邦学习的机器学习算法应用于动态信任评估模型的效果进行了比较,实验结果表明了在基于联邦学习的动态身份认证模型中使用SVM作为机器学习训练方法的效果优于其他机器学习训练方法。最后,本文从安全性和隐私性的角度出发还对动态信任评估系统自身以及联邦学习带来的安全性和隐私性的影响做了讨论。
