Quickest change detection with unknown post-change distribution

Lau, Tze Siong; Tay, Wee Peng; Veeravalli, Venugopal V.

doi:10.1109/icassp.2017.7952892

Cited by 20 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For baseline methods specifically designed for outage detection with unknown post-outage distribution, we consider an approximated maximum likelihood estimation (MLE) proposed to learn the unknown parameters [3], [34], a generalized likelihood ratio test (GLRT) that only considers finite possibilities [35] of post-outage distributions f , a Shewhart test [36] that utilizes meanshift and covariance changes in the data to detect outages. For methods that are developed for unknown post-change distribution in the change point detection, we consider a nonparametric binned generalized statistic (BGS) proposed to approximate the original ratio test in classic CPD [37], a non-parametric uncertain likelihood ratio (ULR) proposed to replace the original ratio [38], a distributed approach (DIS) [1], and a deep Q-network approach (DCQ) [39].…”

Section: Validate On Extensive Outage Scenarios With Real-world Datamentioning

confidence: 99%

Distribution Grid Line Outage Identification With Unknown Pattern and Performance Guarantee

Xiao,

Liao,

Weng

2024

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

Line outage identification in distribution grids is essential for sustainable grid operation. In this work, we propose a practical yet robust detection approach that utilizes only readily available voltage magnitudes, eliminating the need for costly phase angles or power flow data. Given the sensor data, many existing detection methods based on change-point detection require prior knowledge of outage patterns, which are unknown for realworld outage scenarios. To remove this impractical requirement, we propose a data-driven method to learn the parameters of the post-outage distribution through gradient descent. However, directly using gradient descent presents feasibility issues. To address this, we modify our approach by adding a Bregman divergence constraint to control the trajectory of the parameter updates, which eliminates the feasibility problems. As timely operation is the key nowadays, we prove that the optimal parameters can be learned with convergence guarantees via leveraging the statistical and physical properties of voltage data. We evaluate our approach using many representative distribution grids and real load profiles with 17 outage configurations. The results show that we can detect and localize the outage in timely manner with only voltage magnitudes and without assuming the prior knowledge of outage patterns.

show abstract

Section: Validate On Extensive Outage Scenarios With Real-world Datamentioning

confidence: 99%

Distribution Grid Line Outage Identification With Unknown Pattern and Performance Guarantee

Xiao,

Liao,

Weng

2024

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

show abstract

“…A preliminary version of this work was presented in [46]. To the best of our knowledge, the only other recursive test known in the literature for the case where the post-change distribution is not completely specified and does not belong to a finite set of distributions is discussed in [47].…”

Section: B Our Contributionsmentioning

confidence: 99%

A Binning Approach to Quickest Change Detection With Unknown Post-Change Distribution

Lau

Tay

Veeravalli

2019

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

show abstract

“…For other problems where the pre-or postchange distribution is unknown or not parametrized most of the CUSUM-like algorithms that form a likelihood ratio test are not applicable. Since, in this case, the change in the distribution is arbitrary it is hard to perform optimally against all alternatives and the most widely used method in the literature is to form a generalized likelihood ratio test (GLRT) over the set of alternative distributions and then use it as a drift term [11], which can be computationally demanding or without any optimality guarantees. For this problem, we propose a computationally efficient method, called information projection test (IPT), that works in conjunction with an existing change detection algorithm as an additional filter to separate outliers from a change.…”

Section: Introductionmentioning

confidence: 99%

Separating an Outlier from a Change

Sargun

Köksal

2019

2019 53rd Asilomar Conference on Signals, Systems, and Computers

View full text Add to dashboard Cite

We study the quickest change detection problem with an unknown post-change distribution. In this scenario, the unknown change in the distribution of observations may occur in many ways without much structure, while, before change, an outlier (a false alarm event) is highly structured, following a particular sample path. We first characterize these likely events for the deviation of finite strings and propose a method to test the deviation, relative to the most likely way for it to occur as an outlier. Our method works along with other change detection schemes to substantially reduce the false positive rates associated with the plain scheme used without the heavy computation associated with the generalized likelihood ratio test. Finally, we apply our method on economic market indicators and climate data. Our method successfully captures the regime shifts during times of historical significance and identifies the current climate change phenomenon to be a highly likely regime shift.

show abstract

Quickest change detection with unknown post-change distribution

Cited by 20 publications

References 21 publications

Distribution Grid Line Outage Identification With Unknown Pattern and Performance Guarantee

Distribution Grid Line Outage Identification With Unknown Pattern and Performance Guarantee

A Binning Approach to Quickest Change Detection With Unknown Post-Change Distribution

Separating an Outlier from a Change

Contact Info

Product

Resources

About