Classification and Assessment of Power System Security Using Multiclass SVM
01 Sep 2011-Vol. 41, Iss: 5, pp 753-758
TL;DR: The proposed SVM-based pattern classifier system is implemented and tested on standard benchmark systems and the results are compared with least-squares, probabilistic neural network, extreme learning machine, and extreme SVM classifiers.
Abstract: Security assessment and classification are the major concerns in real-time operation of electric power systems. This paper proposes a multiclass support vector machine (SVM) classifier for static and transient security assessment and classification. A straightforward and quick procedure called the sequential forward selection method is used for a feature selection process. The security status of any given operating condition is classified into four modes, viz., secure, critically secure, insecure, and highly insecure, based on the computation of a security index. The proposed SVM-based pattern classifier system is implemented and tested on standard benchmark systems. The simulation results of the multiclass SVM classifier are compared with least-squares, probabilistic neural network, extreme learning machine, and extreme SVM classifiers. The feasibility of implementation of the proposed classifier system for online security evaluation is also discussed.
Citations
More filters
27 Jul 2014
TL;DR: This paper proposes a decision tree (DT)-based systematic approach for cooperative online power system dynamic security assessment (DSA) and preventive control that trains two contingency-oriented DTs on a daily basis by the databases generated from power system simulations.
Abstract: 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 daily basis by the databases generated from power system simulations. Fed with real-time wide area measurements, one DT about measurable variables is employed for online DSA to identify potential security issues and the other DT about 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.
175 citations
Cites background from "Classification and Assessment of Po..."
...Pattern recognition techniques, such as artificial neural networks (ANNs) [2], support vector machines (SVMs) [3], and decision trees (DTs) [4]–[16] can be applied in DSA of power systems....
[...]
TL;DR: In this paper, a binary support vector machine (SVM) classifier with combinatorial trajectories inputs was trained to predict the transient stability status of a power system following a large disturbance such as a fault, based on dynamic response trajectories of rotor angle, speed, voltage, electromagnetic power and imbalance power.
77 citations
Cites background or methods from "Classification and Assessment of Po..."
...Recently, a lot of work has been done in power system transient stability assessment by means of SVM [19]....
[...]
...[19] proposed a multiclass support vector machine (SVM) classifier for static and transient stability assessment and classification....
[...]
...Moreover, ANN is good in interpolation but not so good in extrapolation which decreases its generalization ability [19]....
[...]
TL;DR: In this article, a real-time transient stability assessment (TSA) approach based on prediction of area-based center-of-inertia (COI) referred rotor angles from phasor measurement unit (PMU) measurements is presented.
Abstract: Several smart grid applications have recently been devised in order to timely perform supervisory functions along with self-healing and adaptive countermeasures based on system-wide analysis, with the ultimate goal of reducing the risks associated with potentially insecure operating conditions. Real-time transient stability assessment (TSA) belongs to this type of applications, which allows deciding and coordinating pertinent corrective control actions depending on the evolution of post-fault rotor-angle deviations. This study presents a novel approach for carrying out real-time TSA based on prediction of area-based centre-of-inertia (COI) referred rotor angles from phasor measurement unit (PMU) measurements. Monte Carlo-based procedures are performed to iteratively evaluate the system transient stability response, considering the operational statistics related to loading condition changes and fault occurrence rates, in order to build a knowledge database for PMU and COI-referred rotor-angles as well as to screen those relevant PMU signals that allows ensuring high observability of slow and fast dynamic phenomena. The database is employed for structuring and training an intelligent COI-referred rotor-angle regressor based on support vector machines [support vector regressor (SVR)] to be used for real-time TSA from selected PMUs. Besides, the SVR is optimally tuned by using the swarm variant of the mean-variance mapping optimisation. The proposal is tested on the IEEE New England 39-bus system. Results demonstrate the feasibility of the methodology in estimating the COI-referred rotor angles, which enables alerting about real-time transient stability threats per system areas, for which a transient stability index is also computed.
74 citations
TL;DR: In this article, the authors proposed a methodology for identifying and classifying transmission line faults occurring at any location in a power grid from phasor measurement unit measurements at only one of the generator buses.
Abstract: Smart power grids (SPGs) entail comprehensive real-time smart monitoring and controlling strategies against contingencies such as transmission line faults. This study proposes a novel methodology for identifying and classifying transmission line faults occurring at any location in a power grid from phasor measurement unit measurements at only one of the generator buses. The proposed methodology is based on frequency domain analysis of equivalent voltage phase angle and equivalent current phase angle at the generator bus. Equivalent voltage and current phase angles are the angles made by three-phase equivalent voltage and current phasors with respect to reference axis. These angles are estimated through Park's transformation and frequency domain analysis is performed over a fixed time span equal to inverse of system nominal frequency using fast Fourier transformation. The proposed methodology can be utilised for relaying purposes in case of single transmission lines as well as for system protection centre (SPC) applications in power grid. The significance of the fault information from the methodology is for assisting SPC in SPGs for transmission line fault detection and classification to restore the transmission lines at the earliest and initiate wide-area control actions to maintain system stability against disturbances generated by occurrence and clearance of fault.
68 citations
TL;DR: An efficient approach based on the combination of dragonfly optimization (DFO) algorithm and support vector regression (SVR) has been proposed for online voltage stability assessment, which can successfully predict the VSI.
Abstract: In this paper, an efficient approach based on the combination of dragonfly optimization (DFO) algorithm and support vector regression (SVR) has been proposed for online voltage stability assessment. As the performance of the SVR model extremely depends on careful selection of its parameters, the DFO algorithm involves SVR parameters setting, which significantly ameliorates their performance. In the proposed approach, the voltage magnitudes of the phasor measurement unit (PMU) buses are adopted as the input data for the hybrid DFO–SVR model, while the minimum values of voltage stability index (VSI) are taken as the output vector. Using the data provided by PMUs as the input variables makes the proposed model capable of assessing the voltage stability in a real-time manner, which helps the operators to adopt the required measures to avert large blackouts. The predictive ability of the proposed hybrid model was investigated and compared with the adaptive neuro-fuzzy inference system (ANFIS) through the IEEE 30-bus and the Algerian 59-bus systems. According to the obtained results, the proposed DFO–SVR model can successfully predict the VSI. Moreover, it provides a better performance than the ANFIS model.
49 citations
References
More filters
TL;DR: Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.
Abstract: LIBSVM is a library for Support Vector Machines (SVMs). We have been actively developing this package since the year 2000. The goal is to help users to easily apply SVM to their applications. LIBSVM has gained wide popularity in machine learning and many other areas. In this article, we present all implementation details of LIBSVM. Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.
40,826 citations
TL;DR: A new learning algorithm called ELM is proposed for feedforward neural networks (SLFNs) which randomly chooses hidden nodes and analytically determines the output weights of SLFNs which tends to provide good generalization performance at extremely fast learning speed.
10,217 citations
"Classification and Assessment of Po..." refers methods in this paper
...It can be observed from the results of classification shown in Table V that SVM classifier gives a better performance in terms of highclassification accuracy and less misclassification rate compared with the MLS, PNN, ELM, and ESVM classifiers....
[...]
...The performance of SVM is also compared with extreme learning machine (ELM) [20] and online sequential extreme learning machine (OSELM) [21], newly developed algorithms to replace feedforward networks....
[...]
...In addition, in ELM and OS-ELM algorithms, the RBF activation function is used to compute the hidden layer output matrix....
[...]
...Computations for ELM and OSELM classifiers are performed in the environment of MATLAB 7.6....
[...]
TL;DR: Decomposition implementations for two "all-together" multiclass SVM methods are given and it is shown that for large problems methods by considering all data at once in general need fewer support vectors.
Abstract: Support vector machines (SVMs) were originally designed for binary classification. How to effectively extend it for multiclass classification is still an ongoing research issue. Several methods have been proposed where typically we construct a multiclass classifier by combining several binary classifiers. Some authors also proposed methods that consider all classes at once. As it is computationally more expensive to solve multiclass problems, comparisons of these methods using large-scale problems have not been seriously conducted. Especially for methods solving multiclass SVM in one step, a much larger optimization problem is required so up to now experiments are limited to small data sets. In this paper we give decomposition implementations for two such "all-together" methods. We then compare their performance with three methods based on binary classifications: "one-against-all," "one-against-one," and directed acyclic graph SVM (DAGSVM). Our experiments indicate that the "one-against-one" and DAG methods are more suitable for practical use than the other methods. Results also show that for large problems methods by considering all data at once in general need fewer support vectors.
6,562 citations
"Classification and Assessment of Po..." refers background in this paper
...After each of the binary classifiers makes its vote, the decision function assigns an instance x to a class having a largest number of votes [15]....
[...]
TL;DR: The results show that the OS-ELM is faster than the other sequential algorithms and produces better generalization performance on benchmark problems drawn from the regression, classification and time series prediction areas.
Abstract: In this paper, we develop an online sequential learning algorithm for single hidden layer feedforward networks (SLFNs) with additive or radial basis function (RBF) hidden nodes in a unified framework. The algorithm is referred to as online sequential extreme learning machine (OS-ELM) and can learn data one-by-one or chunk-by-chunk (a block of data) with fixed or varying chunk size. The activation functions for additive nodes in OS-ELM can be any bounded nonconstant piecewise continuous functions and the activation functions for RBF nodes can be any integrable piecewise continuous functions. In OS-ELM, the parameters of hidden nodes (the input weights and biases of additive nodes or the centers and impact factors of RBF nodes) are randomly selected and the output weights are analytically determined based on the sequentially arriving data. The algorithm uses the ideas of ELM of Huang developed for batch learning which has been shown to be extremely fast with generalization performance better than other batch training methods. Apart from selecting the number of hidden nodes, no other control parameters have to be manually chosen. Detailed performance comparison of OS-ELM is done with other popular sequential learning algorithms on benchmark problems drawn from the regression, classification and time series prediction areas. The results show that the OS-ELM is faster than the other sequential algorithms and produces better generalization performance
1,800 citations
"Classification and Assessment of Po..." refers methods in this paper
...The performance of SVM is also compared with extreme learning machine (ELM) [20] and online sequential extreme learning machine (OSELM) [21], newly developed algorithms to replace feedforward networks....
[...]
...Computations for ELM and OSELM classifiers are performed in the environment of MATLAB 7.6....
[...]
Book•
26 Oct 1999TL;DR: This book presents architectures for multiclass classification and function approximation problems, as well as evaluation criteria for classifiers and regressors, and discusses kernel methods for improving the generalization ability of neural networks and fuzzy systems.
Abstract: A guide on the use of SVMs in pattern classification, including a rigorous performance comparison of classifiers and regressors. The book presents architectures for multiclass classification and function approximation problems, as well as evaluation criteria for classifiers and regressors. Features: Clarifies the characteristics of two-class SVMs; Discusses kernel methods for improving the generalization ability of neural networks and fuzzy systems; Contains ample illustrations and examples; Includes performance evaluation using publicly available data sets; Examines Mahalanobis kernels, empirical feature space, and the effect of model selection by cross-validation; Covers sparse SVMs, learning using privileged information, semi-supervised learning, multiple classifier systems, and multiple kernel learning; Explores incremental training based batch training and active-set training methods, and decomposition techniques for linear programming SVMs; Discusses variable selection for support vector regressors.
1,002 citations
"Classification and Assessment of Po..." refers methods in this paper
...Thus, in each training session, the number of training data is considerably reduced compared with OVA–SVM, which uses all the training data [14]....
[...]