A new approach for online coherency identification in power systems based on correlation characteristics of generators rotor oscillations

In this paper, a new approach for identifying coherent groups of generators in power systems based on the correlation coefficients between bus phase angles is presented. The method uses a new clustering index based on the correlation coefficients of bus phase angles. The proposed approach uses real time data of bus phase angles via PMUs, so it is able to easily take into account the effect of system detailed modeling, generators and system controllers and type of events. The proposed correlation index evaluated from the real time behavior of power system in time-domain following disturbances are used to evaluate the degree of coherency between any pair of buses. The bus phase angle can be obtained from synchronized measurements of system quantities using PMUs. Hence, the proposed method could be integrated into a wide-area measurement system enabling fast identification of coherent groups of generators. The proposed method is tested on the IEEE 39-bus with 10 generators.

VANTAGE: A Lyapunov exponents based technique for identification of coherent groups of generators in power systems

Optimized virtual inertia of wind turbine for rotor angle stability in interconnected power systems

The development of the modern Global Positioning System (GPS) technique makes the real time monitoring of the generator status workable. However it is impractical and unnecessary to install GPS devices at all generators due to the high cost. Observations and calculations indicate that most of the generators in a large power system have some similar transient power angle characteristics. Therefore the generators can be grouped according to the similarity of their power angle characteristics, and within each group only one generator is required to install GPS device to monitor the real time status of all the generators in the group. Based on the coherence group theory, the coherence groups of the generators are irrelevant to the magnitude of disturbance and the details of the generator unit model. Therefore, by building a classical synchronous generator linear system model, an accessible Gram matrix model can be derived. Then according to the generator identification rule of ε-coherence (or coherent group), the generator coherence groups can be identified in a large power system. The methodology can facilitate the real time monitoring of generators' status and the real-time control based on the GPS technique.

The coherent generator groups identified method was proposed via Empirical Mode Decomposition (EMD) and Stochastic Subspace Identification (SSI) method in this paper. Only the generator rotor speed gathered from the Wide Area Measurement System (WAMS) is used in the proposed method, and the detailed model and parameters of power system components are not needed. And the phase diagram obtained by using the SSI was employed in the proposed method to identify the coherent generator groups. At the end, the simulation was tested on the CEPRI system with 8-generator. The results of test system testify the efficiency of the proposed method.

This paper presents the application of fuzzy c-means (FCM) clustering to the recognition of coherent generators in power systems. A coherency measure, which is derived from the time-domain dynamic responses of generators, is first proposed for evaluating the property of generator coherency. From the coherency measure a fuzzy relation matrix describing the degree of coherency between any pair of generators is constructed. Fuzzy c-means clustering analysis is applied on coherency measure. The result of various coherent generator groups can thus be obtained, showing results of clustering for different prescribed number of coherent groups. Application results of a sample power system are presented to show the validity and effectiveness of the proposed method.

With the deregulation of the electrical power system and competitive business environment, power system monitoring and control have been emerged as a potential research area. Operation and control of power system are based on several strategies under normal, critical and emergency states of a power system. Control actions can be initiated either at load end (load shedding) or at generator end (generation rescheduling). Both of these actions required the knowledge of the swing of the generators at different operating/contingency conditions. This paper presents an application of supervised learning model to detect coherent machines in a large power network. A method to find the real-time transient stability state and identification of the coherent generator groups by predicting the rotor angle values following a large disturbance through Support Vector Machines (SVMs) is proposed in this work. The proposed method enables to determine synchronism state of the individual machine in real-time. The validity of proposed method is investigated on IEEE-39 bus test system.

In this paper, a new approach for detecting realtime matching for controlled islanding based on the correlation coefficient of generators is proposed. In this approach, by online measuring generators rotor angle oscillations, the correlation coefficients between all pairs of generators are evaluated. Based on the evaluated correlation coefficients, coherent groups of generators are identified. Then, using correlation coefficient of coherent groups, splitting indices between coherent groups are evaluated. Following a disturbance, in the case of the risk of splitting coherent groups towards islanding, the values of splitting indices start to increase detecting tendency of coherent groups towards splitting state. The proposed approach is demonstrated on the IEEE 39-bus system with promising results.

Power system monitoring and control in real time is a challenging task for modern power system due to large operational constraints. The deployment of phasor measurement units (PMUs) at key locations provides an opportunity for devising effective power system monitoring and control measures. In this study, a new method is proposed to determine the real-time transient stability status and identification of the coherent generator groups by predicting the rotor angle values following a large disturbance through radial basis function neural network. The first six cycles of synchronously sampled post-fault data measurements from PMUs consisting of rotor angles and voltages of generators are taken as the input to the neural network to predict the future state of the system. The proposed method can also determine the synchronism state of the individual machine in real time. The proposed scheme is demonstrated on the IEEE-39 bus test system at different operating conditions.

Inter-area oscillations and cascading failures are the most serious threats to the security of the electric power system. Uncontrolled islanding will occur in the event of an unstable inter-area oscillation or a progressive cascading failure. The establishment of uncontrolled islands with a deficiency in load-generation balance is the main reason for system blackout. Controlled islanding has been proposed as a preventive strategy for reducing the risk of blackout in this regard. A new algorithm for applying the controlled islanding strategy is proposed in this paper, based on load coherency and nearest electrical distances between coherent groups of generators. Coherent generators as the main core for controlled islands are identified in this method, which is based on the correlation coefficients between generators and the DBSCAN clustering algorithm. The sub-networks are then created by applying mixed-integer linear programming to each coherent group. Once this is accomplished, non-linear programming is used to construct the stable sub-networks associated with islands that meet the requirements of load-generation balance, voltage limitations, and transmission limits. The proposed scheme is implemented on the small-scale IEEE 39-bus system and a large-scale realistic power system which is the Iran power grid. The results demonstrate that the proposed method is capable of being implemented in a real power system.

This paper illustrates a new technique that identifies coherent clusters of synchronous generators in multi-machine power systems. The technique can be easily implemented in a wide-area measurement system. The importance of fast identification of coherent groups of generators based on wide-area signal measurements lies in its contribution to the design of wide-area based stability control schemes that aim to enhance the overall performance of the power system. The method uses a hierarchical clustering technique (Agglomerative Method) to classify any number of synchronous generators into a number of coherent groups. A new technique is proposed to take into account the influence of the type of events on the clustering. The clustering is based on coherency measures obtained from the time domain responses of the generators following system disturbances. These measures are used to evaluate the degree of coherency between any pair of generators. Thereafter, the clustering algorithm is used to cluster these coherent generators into coherent groups. It is suggested that generators' rotor measurements can be obtained and therefore synchronised measurements of these quantities using Phasor Measurement Units (PMUs) technology can be acquired. Hence, the proposed clustering method could be integrated into a wide-area measurement system that enables fast identification of coherent clusters of generators. The proposed method is tested on the standard 16 generator 68 bus system.

In this paper, a new approach for recognizing coherent groups of generators based on the correlation characteristics of generator rotor angle oscillation is proposed. In this approach, by online measuring generators rotor angle oscillations, the correlation coefficients between all pairs of generators are evaluated. Based on the evaluated correlation coefficients, coherent groups of generators are identified. Comparing the result with time domain transient stability simulation following a disturbance and also modal analysis, leads to similar coherent groups. The proposed approach is demonstrated on the IEEE 39-bus system with promising results.

Identification of coherent generators groups in multi-machine power systems is an important step for both operation and control needs. This paper presents a new identification method based on the generators' rotor angular positions. The methodology consists of extracting the time-domain dynamic responses of generators to build a relationship matrix indicating the degree of coherency between any pair of generators. Then, by applying the Fuzzy C Mean (FCM) clustering, coherent groups are determined. The Centre of Inertia (COI) is used to represent the coherent group in order to visualize the global oscillations of the rotor angle of that particular group following a disturbance. Applications on the 10-machine New England power system show the feasibility and the validity of the proposed methodology.

Coherency detection of generators using recurrence quantification analysis

An eigensystem realization algorithm based data-driven approach for extracting electromechanical oscillation dynamic patterns from synchrophasor measurements in bulk power grids