Line Outage Identification Based on AC Power Flow and Synchronized - PowerPoint PPT Presentation

1 Paper No: 20PESGM0226 Line Outage Identification Based on AC Power Flow and Synchronized Measurements Zhen Dai, Joseph Euzebe Tate Department of Electrical and Computer Engineering University of Toronto zhen.dai@mail.utoronto.ca Discovery Grant NSERC RGPIN-2016-06674

2 Background • Problems: inaccurate dc model, unrealistic assumptions, and misidentification • Proposed method: Identification + Rejection Filter filter threshold Stage 1: Identification -0.05 correct Compare expected voltage changes -0.051 filtered misidentified imag delta phasor (via ac power flow) to observations -0.052 filtered Stage 2: Rejection filter -0.053 Examine the distance between the -0.054 correct best candidate to the second best -0.055 misidentified -0.056 Tested on various systems using random -0.057 -0.03 -0.025 -0.02 -0.015 -0.01 -0.005 0 PMU placements and noisy measurements real delta phasor Fig. 1

3 Results Comparison of the dc and ac approaches Stage 1: Identification Stage 2: Filtering 100 100 • AC > DC 80 • Rejection filter 80 percentage percentage Misidentified Filtered 60 60 Correct Misidentified Misidentified 40 40 Correct Filtered MisidentifiedFiltered ac correct CorrectFiltered Correct dc correct 20 20 • More PMUs 0 0 5 15 20 25 30 10 5 15 20 25 30 10 nPMUs nPMUs Fig. 3 The 30-bus system

4 Results Identification results versus the filter threshold • Threshold Misidentified Correct • More PMUs Correct Correct Filtered Fig. 5 (a) 26 PMUs (c) 868 PMUs (220 kV+) The Ontario system (3488 buses)

5 Conclusions/Recommendations • Ac model necessary for higher accuracy • Rejection filter to overcome misidentification due to measurement uncertainties • Significant benefits of having a higher PMU coverage