eduzhai > Applied Sciences > Engineering >

Unsupervised Event Detection Clustering and Use Case Exposition in Micro-PMU Measurements

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 12 pages

Abstract: Distribution-level phasor measurement units, a.k.a, micro-PMUs, report alarge volume of high resolution phasor measurements which constitute a varietyof event signatures of different phenomena that occur all across powerdistribution feeders. In order to implement an event-based analysis that hasuseful applications for the utility operator, one needs to extract these eventsfrom a large volume of micro-PMU data. However, due to the infrequent,unscheduled, and unknown nature of the events, it is often a challenge to evenfigure out what kind of events are out there to capture and scrutinize. In thispaper, we seek to address this open problem by developing an unsupervisedapproach, which requires minimal prior human knowledge. First, we develop anunsupervised event detection method based on the concept of GenerativeAdversarial Networks (GAN). It works by training deep neural networks thatlearn the characteristics of the normal trends in micro-PMU measurements; andaccordingly detect an event when there is any abnormality. We also propose atwo-step unsupervised clustering method, based on a novel linear mixed integerprogramming formulation. It helps us categorize events based on their origin inthe first step and their similarity in the second step. The active nature ofthe proposed clustering method makes it capable of identifying new clusters ofevents on an ongoing basis. The proposed unsupervised event detection andclustering methods are applied to real-world micro-PMU data. Results show thatthey can outperform the prevalent methods in the literature. These methods alsofacilitate our further analysis to identify important clusters of events thatlead to unmasking several use cases that could be of value to the utilityoperator.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×