Support vector clustering‐based direct coherency identification of generators in a multi‐machine power system

The identification of critical machines (CMs) is the first and most important step of Equivalent‐Single Machine Infinite Bus based methods such as Extend Equal Area Criterion and SIngle Machine Equivalent (SIME) for transient stability assessment and control. This paper presents a novel real‐time CM identification method based on the Phase‐Plane Trajectory Vector (PTV) with Phase Measurement Unit (PMU) information support. The proposed method overcomes the drawbacks of conventional CM identification methods and enables the tracking of the change of CMs during the dynamic process. Only two sample points of all generators from PMUs are required to obtain the PTVs at each moment, and the computation of the feature matrix, as well as the k‐means clustering, is fast and accurate. The application of the PTV‐based method is verified through cases studied in the IEEE 39 bus New England power system. Results are compared with those obtained by conventional methods. © 2018 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

show abstract

“…The details of the K ‐means method are available in the literature . First, two clustering centers are given at the beginning:

c_{1} : ((),,, θ_{c 1}, Δ {\overset{true˜}{ω}}_{c 1}, {\overset{φ}{true}}_{c 1})

and

c_{2} : ((),,, θ_{c 2}, Δ {\overset{true˜}{ω}}_{c 2}, {\overset{φ}{true}}_{c 2})

.…”

Section: Phase‐plane Trajectory Vector‐based Methods For Real‐time CM mentioning

confidence: 99%

“…The subsequent simulations in this article also verify the applicability of K -means clustering. The details of the K -means method are available in the literature [16]. First, two clustering centers are given at the beginning: c 1 : (θ c1 , ω c1 , ϕ c1 ) and c 2 : (θ c2 , ω c2 , ϕ c2 ).…”

Section: Ptv-based Real-time Identification Of Cms Tomentioning

confidence: 99%

A phase‐plane trajectory vector‐based method for real‐time identification of critical machines

Yang

Hojo

Zhang

2018

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

show abstract

“…TSA includes evaluating the rotor swings future behaviour after failure or significant disturbance to accurately predict transient stability [6]. The data received from PMUs at the control centre provides sufficient information about the present and future state of the system after disturbance from a stability point of view, it is possible to deal with the stability in real time [7,8,9]. A Multilayer Perceptron neural network with three layers consisting of one input layer, one hidden layer and one output layer is proposed in [10] and compared with radial basis neural networks (RBFNN) in [11].…”

Section: Introductionmentioning

confidence: 99%

Development of Back-Propagation Neural Network for Transient Stability Assessment of Power Systems

Anirudh*,

Janibasha,

Chaitanya

et al. 2019

IJRTE

View full text Add to dashboard Cite

Transient Stability Assessment (TSA) is aspect of the power system dynamic stability assessment, which includes measuring the capacity of the system to stay synchronized under extreme disturbances. This research work shows the transient stability status of the power system following a major disturbance, such as a faults, line switchings, generator voltages. It can be predicted early based on response trajectories of rotor angle. This early prediction of transient stability is achieved by training a Back Propagation Neural Network (BPNN) taking trajectory of rotor angles as training features. Transient stability index (TSI) proposed in [4] is utilized as a target feature. The proposed methodology is tested with wide range of fault data collected from simulated IEEE 39-Benchmark system. The simulation results shows, utilization of BPNN for transient stability prediction resulted in better performance when compared to Radial Basis Neural Network (RBFNN) [4]..

show abstract

“…E-mail: masa_hojo@tokushima-u.ac.jp *State Key Laboratory of Electrical Insulation and Power Equipment, Xi'an Jiaotong University, Xi'an 710049, China **Department of Electrical and Electronic Engineering, Tokushima University, 2-1 Minami-josanjima Tokushima 770-8506, Japan of generator coherency under different disturbances, which is, however, inconsistent with practical cases. The measurement-based methods identify the generator coherency based on the hierarchical clustering method [11], independent component analysis [12,13], spectral clustering method [14,15], wavelet phase difference [16], robust principal component analysis [17], Koopman mode analysis [18], graph theory [19,20], and intelligent method [21,22]. These methods can adapt to various operation conditions, topology changes, and different disturbances by using the real-time measurement data because all the influences of these factors are reflected in the transient response of the power system.…”

Section: Introductionmentioning

confidence: 99%

A phase‐plane‐based dynamic coherency real‐time identification scheme for controlled islanding

Yang

Hojo

Zhang

2018

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

Controlled islanding often acts as the last resort against a severe blackout. Generator coherency is the primary constraint to determine an effective controlled islanding strategy. This article proposes a scheme to identify the dynamic coherency of generators for controlled islanding. The generator coherency is identified based on the phase-plane trajectory vectors (PTVs) on the phase plane for generators (PPG). Then, a phase plane for buses (PPB) is proposed to assign the nongenerator buses to coherent generator groups following the minimum distance principle. The separated islands are formed by disconnecting certain transmission lines according to the identified coherent generators and areas. The case studies on the IEEE 39-bus 10-machine power system show that the proposed scheme cannot only adapt to different disturbances and changes of system conditions and network topology but also succeed in identifying the generator coherency at different times and developing proper islanding strategies according to current system states.

show abstract

Support vector clustering‐based direct coherency identification of generators in a multi‐machine power system

Cited by 40 publications

References 44 publications

A phase‐plane trajectory vector‐based method for real‐time identification of critical machines

A phase‐plane trajectory vector‐based method for real‐time identification of critical machines

Development of Back-Propagation Neural Network for Transient Stability Assessment of Power Systems

A phase‐plane‐based dynamic coherency real‐time identification scheme for controlled islanding

Contact Info

Product

Resources

About