Mode Shape Estimation using Complex Principal Component Analysis and k-Means Clustering

Haugdal, Hallvar; Uhlen, Kjetil

doi:10.1109/sgsma.2019.8784556

Cited by 5 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The continually decreasing angle before t = 310 s indicates that the frequency of oscillations is slightly lower than the mean frequency used as input to the method, while the opposite is true for the period after t = 310 s. Observing that the trajectories of both the amplitude and angle make distinct turns at the same time could indicate that some remedial action was applied that changed the operating point. The same observation is also made in [8].…”

Section: Discussionsupporting

confidence: 78%

“…The work presented is a continuation of previous research on power oscillation monitoring. In [8], [9], it is shown that the frequency and observability mode shape of oscillatory modes can be estimated within seconds after standing oscillations appear in measurements. This motivates the development of monitoring and control applications that make use of this information.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Estimation of Oscillatory Mode Activity from PMU Measurements

Haugdal

Uhlen

Müller

et al. 2020

2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe)

Self Cite

View full text Add to dashboard Cite

We propose a method for estimating the activity of oscillatory modes in power systems. The frequencies and mode shapes of the modes of interest are assumed to be known beforehand, either from linear modal analysis or from empirical mode estimation methods, and are used in combination with measurements from Phasor Measurement Units to estimate the instantaneous mode excitation in terms of amplitude and phase. The estimation is carried out using non-linear least squares to fit a set of curves to the measured data. Combining mode shapes with measured data allows the activity to be estimated from only a low number of consecutive measurement snapshots, resulting in a problem of low computational complexity that can be solved fast enough for the method to run online.The purpose of estimating the mode activity is, firstly, to contribute to increased situational awareness and facilitate methods that build further upon this information, and secondly, to be able to synthesize signals that can serve as input to controllers for power oscillation damping. It is expected that using this excitation measure will result in a more robust controller that is less prone to disturbances and noise.Index Terms-Empirical modal analysis, non-linear least squares, power oscillations, wide area monitoring and control

show abstract

Section: Discussionsupporting

confidence: 78%

Section: Introductionmentioning

confidence: 99%

Estimation of Oscillatory Mode Activity from PMU Measurements

Haugdal

Uhlen

Müller

et al. 2020

2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe)

Self Cite

View full text Add to dashboard Cite

show abstract

“…The description of the technique is presented below based on what is presented in the researches [1,19,20]. Each row of the data matrix represents the temporal evolution of the monitored measurements or data processed by WAMS system at time t. That is, the variables obtain operation level, and it is appropriately structured so operation matrix name X (instantaneous) of m x n dimensions, as shown in section 2.3.3.…”

Section: Power System Stabilizersmentioning

confidence: 99%

Small Signal Stability Enhancement of a Multimachine Power System Using Probabilistic Tuning PSS Based in Wide Area Monitoring Data

Oscullo¹,

Gallardo²

2020

EJEE

View full text Add to dashboard Cite

Nowadays, the operation of electrical systems presents technical challenges associated with the operational dynamics present in increasingly charged systems. Due to the lowfrequency oscillations generated given the continuous change in demand and the exchange of power between the different areas of the system, under this context, it is interesting to analyze the incorporation into the power system stabilizer (PSS) of information received and processed in the wide-area measurement system (WAMS). To address the problems of small-signal stability; present in the electric power system of manner off-line to have a PSS tuning bank for a group of dispatch. Since the state variables in the power system are stochastic, this paper proposes a probabilistic method based on Monte Carlo that together with the heuristic algorithm of mean-variance mapping optimization (MVMO) allows determining the number of PSSs, their location and the tuning parameters of PSSs this guarantee stability of the system at the new operating point. The proposed methodology is applied in the New England 39 bus system. The results show that PSSs adjusted using the proposed method to improve the response of the small-signal stability.

show abstract

“…The basic structure of the method (CPCA dynamics decomposition combined with clustering for averaging) has been described previously in [11], where it is applied to analysis of large disturbances/ringdowns. Continuing this work, we propose further development of the method by introducing the DBSCAN clustering algorithm, in addition to further testing on a medium size grid and on ambient conditions.…”

Section: Introductionmentioning

confidence: 99%

“…Section II gives a description of the method, based on the more thorough description in [11], however with some modifications. Section III describes results from applying the method to simulated and recorded PMU data, followed by discussions and conclusions in Sections IV and V.…”

Section: Introductionmentioning

confidence: 99%

Power Oscillation Monitoring using Statistical Learning Methods

Haugdal

Uhlen

2019

2019 IEEE Milan PowerTech

Self Cite

View full text Add to dashboard Cite

This paper describes development and testing of a method for estimation of electromechanic modes and corresponding mode shapes from frequency or voltage angle measurements in Wide Area Monitioring Systems. The method uses Complex Principal Component Analysis to perform a decomposition of the dynamics captured in the measurements, reducing any detected oscillations into sets of parameters. The numerous sets of parameters generated after a time period can be interpreted as points or observations in high-dimensional space, on which a clustering algorithm can be applied to pinpoint areas where high densities of points are accumulated. Results show that highly accurate estimates of oscillatory modes in the power system and their mode shapes can be derived by averaging the point observations belonging to each cluster. Important development described in this paper includes the use of voltage angle as input and testing on a medium size power system simulation model with synchronized measurements from 44 nodes. Furthermore, introduction of the DBSCAN clustering algorithm shows very promising results when applying the method to recorded phasor measurements from the Nordic power system.

show abstract

Mode Shape Estimation using Complex Principal Component Analysis and k-Means Clustering

Cited by 5 publications

References 12 publications

Estimation of Oscillatory Mode Activity from PMU Measurements

Estimation of Oscillatory Mode Activity from PMU Measurements

Small Signal Stability Enhancement of a Multimachine Power System Using Probabilistic Tuning PSS Based in Wide Area Monitoring Data

Power Oscillation Monitoring using Statistical Learning Methods

Contact Info

Product

Resources

About