Coherency Identification in Power Systems Through Principal Component Analysis

Abstract

Since all columns include only zeroes, all state variables are observable.Numerical tests carried out using the IEEE Reliability Test System [8], including voltage, active and reactive power injection, and active and reactive power flow measurements, show the effective behavior of the proposed algorithm. IV. CONCLUSIONThis letter proposes and illustrates a novel technique to state estimation observability analysis. This technique relies on computing the null space of the Jacobian measurement matrix. The computational burden involved is, therefore, that pertaining to the null space calculation, which is similar to the burden of solving a linear homogeneous system (O(N 3 )). A simple example is used to demonstrate the efficacious functioning of the technique. Extensive computational analysis involving voltage and active and reactive power measurements as well as current measurements is the subject of a future paper. App. Syst., vol. 100, no. 2, pp. 691-698, Feb. 1981. [6] , "An orthogonal row processing algorithm for power system sequential state estimation," IEEE Trans. Power App. Syst., vol. 100, no. 8, pp. 3791-3800, Aug. 1981. [7] REFERENCES I. INTRODUCTIONIncreasing interconnection of power plants in modern large electric power systems has made power system dynamic studies much more complex. Under the deregulated business environment, the interconnections are increasingly being used for trading between utilities. This stresses the interconnections through large power transfer. As a result, the system is prone to low-frequency interarea oscillations. In such an operating scenario, the system behaves coherently with groups of coherent generators separated from other groups of coherent generators linked through weak interconnections.The analysis of an interconnected power system is generally for a specified portion, called the study system, while the rest is the external portion of the system approximated to its equivalent of lower dimension. Since the impacts of major disturbances propagate through tie lines to neighboring systems, it is important to represent not only the power system in question but also the neighboring utilities or the external system in terms of its dynamic equivalent.Coherency-based approaches to dynamic equivalents have been adopted in reducing the size of the power system model [1]- [3]. However, coherency was identified with the help of the linearized model of the power systems. More often than not, the access to system dynamic data involving a number of utilities is incredibly difficult. The accuracy of the system model influences the accuracy of the results. Also, owing to the dimension of the interconnected systems, it is neither easy nor desirable to represent the entire system model in detail.In this letter, coherency is obtained from measured data, obviating the need for detailed modeling information. The method is based on principal component analysis (PCA), which is computationally simple and fast. An industrial application of this method will take the wide area measur...

Controlled islanding schemes for interconnected power systems based on coherent generator group identification and wide-area measurements

“…9 that the generators are clustered as two coherent groups which are denoted as CG1 00 and CG2 00 . CG1 00 = {G 1 , G 2 , …, G 13 }, and CG2 00 = {G 14 , G 15 , G 16 }. The generators in the same coherent group are connected each other, which means that these generators are strongly coherent.…”

“…An approach for identifying coherent generators of power systems is presented in [15] by using spectrum analysis of the generators velocity variations. In [16], an algorithm based on principal component analysis is presented for identifying coherent generators of an interconnected power system by using the measured data sets of generator speeds and bus angles. In [17], the independent component analysis method is applied for coherency identification of interconnected power systems with WAMS.…”

Comparative review of methods for stability monitoring in electrical power systems and vibrating structures

“…Spectral methods have also been used for mode shape and coherency estimation for example in [6] where the mode shapes are estimated using synchrophasor measurements from cross-spectral densities derived from the FFTs of the signals, and in [7] where PCA is used to obtain coherent groups of generators by clustering the weightings of the PCs obtained using simulated speed measurements at the rotors of generators.…”

Two-stage controlled islanding strategy based on Stoer-Wagner and improved Dinic algorithms

“…The first type is analyzing power-angle curves of the perturbed generators by mathematical methods, such as k-medoids cluster analysis [4] and the principal component analysis (PCA) [5]. However, the physical meanings of the above algorithms are not clear enough, which cannot explain the swinging process of the generators.…”