Power System Dynamic Security Assessment Research Articles

A domain adaptation‐based convolutional neural network incorporating data augmentation for power system dynamic security assessment

AbstractRecently, deep learning (DL) based dynamic security assessment (DSA) methods have been very successful. However, although a DSA model can be trained well for a specific topology, it often does not perform well for other topologies. Since the topology in real‐world power systems is frequently changing, the performance reduction of DL‐based DSA methods is very serious, which is a challenging and urgent problem. This paper proposes a novel DSA method based on a convolutional neural network (CNN) to solve this problem. In the proposed method, a strong yet simple domain adaptation approach named adaptive batch normalization (AdaBN) is used, which significantly enhances the extensibility and generalizability of the DSA model when the topology changes and eliminates the need to train a large number of models. This approach achieves a deep adaptation effect by modulating the statistics from the source domain to the target domain in all batch normalization layers across the model. Unlike other domain adaptation methods, this method is parameter‐free, requires no additional components, and has advanced performance despite its simplicity. In addition, this paper introduces TGAN‐based data augmentation to deal with the difficulty of costly data collection and labelling. This data augmentation makes the proposed model applicable to small databases. The test results of the proposed method on IEEE 39‐bus and IEEE 118‐bus systems show that this method can evaluate system dynamic security during topology changes and in the face of data noise with high accuracy.

A domain adaptation‐based convolutional neural network incorporating data augmentation for power system dynamic security assessment

AbstractRecently, deep learning (DL) based dynamic security assessment (DSA) methods have been very successful. However, although a DSA model can be trained well for a specific topology, it often does not perform well for other topologies. Since the topology in real‐world power systems is frequently changing, the performance reduction of DL‐based DSA methods is very serious, which is a challenging and urgent problem. This paper proposes a novel DSA method based on a convolutional neural network (CNN) to solve this problem. In the proposed method, a strong yet simple domain adaptation approach named adaptive batch normalization (AdaBN) is used, which significantly enhances the extensibility and generalizability of the DSA model when the topology changes and eliminates the need to train a large number of models. This approach achieves a deep adaptation effect by modulating the statistics from the source domain to the target domain in all batch normalization layers across the model. Unlike other domain adaptation methods, this method is parameter‐free, requires no additional components, and has advanced performance despite its simplicity. In addition, this paper introduces TGAN‐based data augmentation to deal with the difficulty of costly data collection and labelling. This data augmentation makes the proposed model applicable to small databases. The test results of the proposed method on IEEE 39‐bus and IEEE 118‐bus systems show that this method can evaluate system dynamic security during topology changes and in the face of data noise with high accuracy.

Understanding Discrepancy of Power System Dynamic Security Assessment with Unknown Faults: A Reliable Transfer Learning-based Method

Understanding Discrepancy of Power System Dynamic Security Assessment with Unknown Faults: A Reliable Transfer Learning-based Method

A two-stage parallel hybrid method for power system dynamic security assessment

Abstract The present paper introduces a hybrid method based on both the Lyapunov’s direct approach using the Potential Energy Boundary Surface method and parallel time-domain simulation for fast power system dynamic security assessment. The main contribution in the presented method is an approach that combines the directional derivative of the potential energy function and the maximum potential energy criterion in order to reduce the conservativeness in the estimation of the stability region at the direct method stage. For detailed investigation of the critical contingencies, a parallel approach is applied at the time-domain simulation stage. From the simulation results, the proposed method achieves a high level of accuracy in classifying network contingencies and shows great computational performance potential in light of the requirement for fast dynamic security assessment.

An improved constraint-inference approach for causality exploration of power system transient stability

Transient stability is the key aspect of power system dynamic security assessment, and data-driven methods are becoming alternative measures of assessment. The current data-driven methods only construct correlations between variables while neglecting causal relationships. Therefore, they face problems such as poor robustness, which restrict their practical application. This paper introduces an improved constraint-inference approach for causality exploration of power system transient stability. Firstly, a causal structure discovery method of power system transient stability is proposed based on a PC-IGCI algorithm, which addresses the shortage caused by Markov equivalence and massive variables. Then, a relative average causal effect index is proposed to reveal the relationship between relative intervention strength and causal effects. The results of a case study verify that the proposed method can identify the causal structure between the transient stability variables entirely based on data. In addition, the causal effect sorting between “cause” and “outcome” of transient stability variables is revealed. This paper provides a new approach for data mining to uncover the causal mechanisms between variables in power systems and expand the capabilities of data-driven methods in power system application.

Pre-Fault Dynamic Security Assessment of Power Systems for Multiple Different Faults via Multi-Label Learning

With the advancement of machine learning techniques, data-driven power system dynamic security assessment (DSA) has received great research interests. Traditional methods usually apply one DSA model for one specific fault and cannot simultaneously address different multiple faults by one model. To solve this issue and further improve the DSA accuracy performance, this paper proposes a novel DSA method based on multi-label learning (MLL) with a training database for sufficient and incomplete/uneven coverage labels scenarios. By considering fault correlation between different faults, the proposed MLL-based DSA method can handle multiple faults simultaneously with the high stability assessment accuracy performance. Moreover, this paper provides the detailed optimization process and tight mathematical proof for the proposed MLL-based DSA method. Comprehensive simulation tests and comparisons were conducted on the New England 10-machine 39-bus testing system and the synthetic Illinois 49-machine 200-bus testing system.

Data‐driven power system dynamic security assessment under adversarial attacks: Risk warning based interpretation analysis and mitigation

AbstractPower system dynamic security assessment (DSA) has long been essential for protecting the system from the risk of cascading failures and wide‐spread blackouts. The machine learning (ML) based data‐driven strategy is promising due to its real‐time computation speed and knowledge discovery capacity. However, ML algorithms are found to be vulnerable against well‐designed malicious input samples that can lead to wrong outputs. Adversarial attacks are implemented to measure the vulnerability of the trained ML models. Specifically, the targets of attacks are identified by interpretation analysis that the data features with large SHAP values will be assigned with perturbations. The proposed method has the superiority that an instance‐based DSA method is established with interpretation of the ML models, where effective adversarial attacks and its mitigation countermeasure are developed by assigning the perturbations on features with high importance. Later, these generated adversarial examples are employed for adversarial training and mitigation. The simulation results present that the model accuracy and robustness vary with the quantity of adversarial examples used, and there is not necessarily a trade‐off between these two indicators. Furthermore, the rate of successful attacks increases when a greater bound of perturbations is permitted. By this method, the correlation between model accuracy and robustness can be clearly stated, which will provide considerable assistance in decision making.

Robustness Verification for Machine-Learning-Based Power System Dynamic Security Assessment Models Under Adversarial Examples

Based on machine learning (ML) technique, the data-driven power system dynamic security assessment (DSA) has received significant research interest. Yet, the well-trained ML-based models with high training and testing accuracy may be vulnerable to the adversarial example, which is a modified version of the original sample that is intentionally perturbed but retains being very close to the original one. Such adversarial examples can mislead the DSA results and lead to catastrophic consequences. Thus, the accuracy index alone is not enough to represent the performance of the ML-based DSA models. To evaluate the ML-based DSA models and provide formal robustness guarantee for real-time DSA, this article proposes an adversarial robustness verification method to quantify the ability of ML-based DSA models against all kinds of adversarial examples. A model-free and attack-independent robust index is defined for both differentiable and nondifferentiable attack scenarios. Simulation results have verified the effectiveness of the proposed adversarial robustness verification method and the superiority of robust index compared with the upper bound of the adversarial perturbations computed by existing adversarial attack methods.

Machine Learning in Power System Dynamic Security Assessment

Recent growing energy crisis in Europe, coupled with the rising energy prices worldwide, is a clear indication of the many difficulties awaiting the transition of modern societies away from fossil fuels [...]

Online multi-fault power system dynamic security assessment driven by hybrid information of anticipated faults and pre-fault power flow

Online dynamic security assessment is one of the important applications of online situation awareness for power systems, providing essential information for secure operation and preventive control. Different from the existing methods which are fault-dependent or requiring for transient-state data, this paper proposes a novel deep learning model for online power system multi-fault dynamic security assessment driven by hybrid information of anticipated faults and static operating points. The proposed model is universal and suitable for different fault situations. In addition, this model avoids the acquisition of transient information, thus reducing the time-consuming process of online evaluation. For the discrete nominal features in the hybrid information, one-hot encoding is introduced to preprocess the discrete nominal features. Both the performance of the dynamic security assessment model and the principal component analysis are promoted. Furthermore, an improved adaptive synthetic sampling algorithm is proposed and applied to alleviate the problem of power system data imbalance. Finally, the feasibility and effectiveness of the proposed method is verified by testing on the anticipated faults in the New England 39-bus system and the IEEE 54-machine 118-bus system.

An Integrated Transfer Learning Method for Power System Dynamic Security Assessment of Unlearned Faults With Missing Data

This letter proposes an integrated transfer learning (TL) method for pre-fault dynamic security assessment (DSA) of power systems, which aims to simultaneously achieve fault transfer and address missing data issue for unlearned faults. Moreover, this letter provides the tight mathematic proof for the guaranteed DSA performance of the proposed integrated TL method. Comprehensive simulation results on the benchmark testing system have shown that the proposed integrated TL method can achieve a high DSA accuracy for unknown faults with complete data and can also maintain a satisfactory DSA accuracy for unknown faults even with incomplete data inputs.

Incremental broad learning for real‐time updating of data‐driven power system dynamic security assessment models

This work aims to real-time update the data-driven dynamic security assessment model for accuracy maintenance and improvement. Based on the random vector functional link neural network, an incremental broad learning model is developed as an alternative and much faster approach for deep learning. Besides, the proposed method can significantly reduce the computation burden caused by abundant model parameters of layers and filters. Three real-time increment scenarios are achieved: enhancement hidden nodes, features, and training instances. Distinguished from existing methods, the proposed method can fast remodel in broad expansion without retraining the whole model, since it can only calculate the pseudoinverse of new state matrix without the computations of the whole output weights matrix. Numerical simulation results show that the proposed method can improve the accuracy step by step and update the model within a very short time at the online stage, verifying its real-time computational efficiency and accuracy improvement.

Time Series Data-Driven Batch Assessment of Power System Short-Term Voltage Security

For power system dynamic security assessment (DSA), the conventional dynamic security region method is able to provide valuable information on security margins for preventive control. However, its event-based nature is likely to induce heavy computational burdens, especially in the presence of substantial presumed events. To tackle this challenging problem, this article develops an efficient time series data-driven scheme for batch DSA in a divide-and-conquer manner. First of all, with emphasis on short-term voltage stability, a novel u-shapelet (representative local trajectory)-based hierarchical clustering method is proposed to automatically divide various training cases into a handful of typical transient scenarios. Then, regressive shapelet learning is efficiently carried out to conquer individual scenarios, resulting in a group of high-precision security margin estimation models. With a desirable data-driven nature, the proposed scheme avoids time-consuming dynamic security region (DSR) characterization for each event, thereby achieving a significant speed-up for batch DSA. Test results on the realistic China Southern Power Grid illustrate its excellent performances on batch DSA.

A Fully Data-Driven Method Based on Generative Adversarial Networks for Power System Dynamic Security Assessment With Missing Data

This paper proposes a fully data-driven approach for PMU-based pre-fault dynamic security assessment (DSA) with incomplete data measurements. The generative adversarial network (GAN), which is an emerging unsupervised deep learning technique based on two contesting deep neural networks, is used to address the missing data. While the state-of-the-art methods for missing data are dependent on PMU observability, they are limited by the placement of PMU and network topologies. Distinguished from existing methods, the proposed approach is fully data-driven and can fill up incomplete PMU data independent on PMU observability and network topologies. Therefore, it is more generalized and extensible. Simulation results show that, under any PMU missing conditions, the proposed method can maintain a competitively high DSA accuracy with a much less computation complexity.

Online power system dynamic security assessment with incomplete PMU measurements: a robust white‐box model

With the development of synchronised measurement technique, online dynamic security assessment (DSA) is of great significance to prevent power system blackout. Recently, based on the phasor measurement unit (PMU) data, intelligent data-driven techniques have been rapidly developed for online DSA owing to their fast decision speed, less data requirement, and decision rule discovery ability. The interpretable decision rules provide useful information for preventive control and post-event auditing. However, in case of incomplete measurement events, such as PMU failure, communication loss, and phasor data concentrator failure, the accuracy and the transparency of the intelligent models can be significantly impaired by such incomplete data. To overcome this problem, this paper proposes a robust white-box model that can sustain DSA accuracy and survive the model interpretability and transparency against incomplete PMU data. The proposed model is tested on New England 39-bus system and demonstrates higher robustness over existing methods.

Real-time power system dynamic security assessment based on advanced feature selection for decision tree classifiers

This paper proposes a novel algorithm based on an advanced feature selection technique for the decision tree (DT) classifier to assess the dynamic security in a power system. The proposed methodology utilizes symmetrical uncertainty (SU) to reduce the data redundancy in a dataset for DT classifier-based dynamic security assessment (DSA) tools. The results show that SU reduces the dimension of the dataset used for DSA significantly. Subsequently, the approach improves the performance of the DT classifier. The effectiveness of the proposed technique is demonstrated on the modified IEEE 30-bus test system model. The results show that the DT classifier with SU outperforms the DT classifier without SU. The performance of the proposed algorithm indicates that the DT classifier with SU is able to assess the dynamic security of the system in near real time. Therefore, it is able to provide vital information for protection and control applications in power system operation.

Analyzing the Static Security Functions of a Power System Dynamic Security Assessment Toolbox

This paper presents the results from the iTesla offline workflow for a given set of contingencies applied to Nordic 44 power system model and are verified against the results from a similar workflow implemented separately using Python/Matlab that uses PSS/E analysis function. 21 (20 ‘N-1’ and 1 ‘N-2’) contingencies were created on the Nordic 44 power system model and was executed on 2928 snapshots (April to July 2015). This verification is only performed for the (static) steady state stability assessment results of the offline workflow. The generated decision trees (DTs) by the iTesla platform were verified for different network operating conditions. It was observed that the generated DT’s are consistent for the given set of operating conditions. It would be beneficial to check how the rules generated from DT’s with four months of data (April to July 2015) vary from the rules generated with one year of data (2015). The verification approach adopted provides a useful means to test and verify dynamic security assessment (DSA) tools’ with an independent implementation of some of the tools’ functions, this can be of value to other DSA tool developers.

Advanced Pattern Discovery-based Fuzzy Classification Method for Power System Dynamic Security Assessment

Dynamic security assessment (DSA) is an important issue in modern power system security analysis. This paper proposes a novel pattern discovery (PD)-based fuzzy classification scheme for the DSA. First, the PD algorithm is improved by integrating the proposed centroid deviation analysis technique and the prior knowledge of the training data set. This improvement can enhance the performance when it is applied to extract the patterns of data from a training data set. Secondly, based on the results of the improved PD algorithm, a fuzzy logic-based classification method is developed to predict the security index of a given power system operating point. In addition, the proposed scheme is tested on the IEEE 50-machine system and is compared with other state-of-the-art classification techniques. The comparison demonstrates that the proposed model is more effective in the DSA of a power system.

A vulnerability model for power system dynamic security assessment

Traditional security assessment based on risk takes into account the failure probability and the failure consequence. This paper provides a modification on the risk calculation by replacing the failure probability with the current status to overcome the dependence on mass statistical data. The vulnerability model is proposed that incorporates the two key factors of the risk along with the deterioration trend factor to define the system security level, which efficiently indicates the current and potential danger. Grey system theory is used to predict the equipment deterioration trend and the grey correlation moment is defined to evaluate equipment state with multi-parameters. The classification and quantification of equipment state, failure consequence, risk, deterioration trend and vulnerability are discussed. A feed water pump case demonstrates the effectiveness of vulnerability in system dynamic security assessment.

A Systematic Approach for Dynamic Security Assessment and the Corresponding Preventive Control Scheme Based on Decision Trees

This paper proposes a decision tree (DT)-based systematic approach for cooperative online power system dynamic security assessment (DSA) and preventive control. This approach adopts a new methodology that trains two contingency-oriented DTs on a daily basis by the databases generated from power system simulations. Fed with real-time wide-area measurements, one DT of measurable variables is employed for online DSA to identify potential security issues, and the other DT of controllable variables provides online decision support on preventive control strategies against those issues. A cost-effective algorithm is adopted in this proposed approach to optimize the trajectory of preventive control. The paper also proposes an importance sampling algorithm on database preparation for efficient DT training for power systems with high penetration of wind power and distributed generation. The performance of the approach is demonstrated on a 400-bus, 200-line operational model of western Danish power system.