Multi-Receiver GPS-based Direct Time Estimation for PMUs Sriramya - - PowerPoint PPT Presentation

β–Ά
multi receiver gps based
SMART_READER_LITE
LIVE PREVIEW

Multi-Receiver GPS-based Direct Time Estimation for PMUs Sriramya - - PowerPoint PPT Presentation

Aug 20, 2022 575 likes β€’1.05k views

Multi-Receiver GPS-based Direct Time Estimation for PMUs Sriramya Bhamidipati, Yuting Ng and Grace Xingxin Gao University of Illinois at Urbana-Champaign University of Illinois at Urbana-Champaign CREDC All Hands Meeting | Oct 14 2016

slide-1
SLIDE 1

University of Illinois at Urbana-Champaign University of Illinois at Urbana-Champaign

Multi-Receiver GPS-based Direct Time Estimation for PMUs

Sriramya Bhamidipati, Yuting Ng and Grace Xingxin Gao

CREDC All Hands Meeting | Oct 14 2016

slide-2
SLIDE 2

University of Illinois at Urbana-Champaign

Motivation

  • Supply and demand of electricity should be balanced to

maintain power grid stability

  • Power grid vulnerable to

External attacks Natural disasters Man-made errors

1

slide-3
SLIDE 3

University of Illinois at Urbana-Champaign

Massive power blackouts

2

Northeast USA 2003 Java-Bali 2005 Brazil 2009 India 2012

50 million people affected 670 million people affected 100 million people affected 87 million people affected

slide-4
SLIDE 4

University of Illinois at Urbana-Champaign

Goals of US power community

  • Synchronized phasor measurements
  • Reliable communication network
  • Real-time information monitoring
  • Automation of the power grid
  • Improving the security margins

3

Development of reliable and robust Smart Power Grid

slide-5
SLIDE 5

University of Illinois at Urbana-Champaign

Goals of US power community

  • Synchronized phasor measurements
  • Reliable communication network
  • Real-time information monitoring
  • Automation of the power grid
  • Improving the security margins

4

In use currently Supervisory Control and Data Acquisition (SCADA)

slide-6
SLIDE 6

University of Illinois at Urbana-Champaign

Goals of US power community

  • Synchronized phasor measurements
  • Reliable communication network
  • Real-time information monitoring
  • Automation of the power grid
  • Improving the security margins

4

In use currently Supervisory Control and Data Acquisition (SCADA) Switching to Phasor Measurement Units (PMUs)

slide-7
SLIDE 7

University of Illinois at Urbana-Champaign

Phasor Measurement Unit (PMU)

  • Highly synchronized

measurements

  • PMU measures current

and voltage in power grid

5

[NASPI]

slide-8
SLIDE 8

University of Illinois at Urbana-Champaign

GPS Timing for PMUs

6

GPS used for time synchronization

Power grid PMU GPS clock GPS Antenna

Advantages Global coverage Freely available πœˆπ‘‘-level accurate time

slide-9
SLIDE 9

University of Illinois at Urbana-Champaign

GPS Conventional Approach

  • Inputs
  • Center: 3D satellite position
  • Radius: Pseudoranges
  • Unknowns to be estimated:
  • 3D position 𝐲, 𝐳, π’œ
  • Methodology
  • Trilateration technique

7

Trilateration technique

slide-10
SLIDE 10

University of Illinois at Urbana-Champaign

GPS Conventional Approach

  • Inputs
  • Center: 3D satellite position
  • Radius: Pseudoranges
  • Unknowns to be estimated:
  • 3D position 𝐲, 𝐳, π’œ
  • Clock bias π’…πœΊπ’–
  • Methodology
  • Trilateration technique
  • Minimum 4 satellites required

7

Trilateration technique

slide-11
SLIDE 11

University of Illinois at Urbana-Champaign

GPS Timing for PMUs

8

GPS used for time synchronization

Power grid PMU GPS clock GPS Antenna

Advantages Disadvantages Global coverage Unencrypted structure Freely available Low signal power πœˆπ‘‘-level accurate time Vulnerable to attacks

slide-12
SLIDE 12

University of Illinois at Urbana-Champaign

GPS Timing Attacks

High-power noise signal Power sub-station Authentic GPS signals

Jamming: Makes timing unavailable for PMUs

Power sub-station Replay signal with high power Authentic GPS signals 9

Meaconing: Mislead PMU with wrong time

slide-13
SLIDE 13

University of Illinois at Urbana-Champaign

Objectives

Propose a robust GPS time transfer technique to:

  • Mitigate the effect of external timing attacks
  • Improve tolerance against noise and interference

10

slide-14
SLIDE 14

University of Illinois at Urbana-Champaign

Outline

Motivation and Objectives GPS Conventional approach Multi-Receiver Direct Time Estimation (MRDTE) Experimental setup Results and Analysis Ongoing Work Summary

11

slide-15
SLIDE 15

University of Illinois at Urbana-Champaign

MRDTE: Approach

12

Power substation, Sidney, IL

slide-16
SLIDE 16

University of Illinois at Urbana-Champaign

MRDTE: Approach

  • Multiple receivers
  • Geographical diversity

12 Receiver Receiver Receiver Receiver

Power substation, Sidney, IL

slide-17
SLIDE 17

University of Illinois at Urbana-Champaign

MRDTE: Approach

  • Multiple receivers
  • Geographical diversity
  • Position Aiding
  • Static receiver location

12 Receiver Receiver Receiver Receiver

Power substation, Sidney, IL

slide-18
SLIDE 18

University of Illinois at Urbana-Champaign

MRDTE: Approach

  • Multiple receivers
  • Geographical diversity
  • Position Aiding
  • Static receiver location
  • Direct Time Estimation (DTE)
  • Works with timing parameters
  • No intermediate pseudoranges

12 Receiver Receiver Receiver Receiver

Power substation, Sidney, IL

slide-19
SLIDE 19

University of Illinois at Urbana-Champaign

MRDTE: Approach

  • Multiple receivers
  • Geographical diversity
  • Position Aiding
  • Static receiver location
  • Direct Time Estimation (DTE)
  • Works with timing parameters
  • No intermediate pseudoranges
  • Triggered by common external

clock Reduction in no. of unknowns from 8 x, y, z, cΞ΄t, x, y, z, cΞ΄ t Γ— # of receivers to 2 (cΞ΄t, cΞ΄ t)

12 Receiver Receiver Receiver Receiver

Power substation, Sidney, IL

slide-20
SLIDE 20

University of Illinois at Urbana-Champaign

MRDTE: Architecture

Direct Time Estimation MRDTE Filter 1 3 4 2 Raw GPS signals from multiple receivers PMU Output from PMU: Synchronized phasor measurements Time MRDTE All receivers triggered by a common clock

13

slide-21
SLIDE 21

University of Illinois at Urbana-Champaign

MRDTE: Architecture

All receivers triggered by a common clock 1 3 4 2 Raw GPS signals from multiple receivers Output from PMU: Synchronized phasor measurements MRDTE Filter MRDTE PMU Time Direct Time Estimation

14

slide-22
SLIDE 22

University of Illinois at Urbana-Champaign

Direct Time Estimation

All satellites 3D position and velocity Receiver 3D position and velocity

Combined satellite signal replica Across the candidates in search space Clock Drift Clock Bias

15

slide-23
SLIDE 23

University of Illinois at Urbana-Champaign

Direct Time Estimation

All satellites 3D position and velocity Receiver 3D position and velocity

Combined satellite signal replica Across the candidates in search space Incoming raw GPS signal

Vector Correlation

Clock Drift Clock Bias

15

slide-24
SLIDE 24

University of Illinois at Urbana-Champaign

Direct Time Estimation

All satellites 3D position and velocity Receiver 3D position and velocity

Combined satellite signal replica Across the candidates in search space Incoming raw GPS signal Maximum likelihood clock state

Vector Correlation

Clock Drift Clock Bias

15

slide-25
SLIDE 25

University of Illinois at Urbana-Champaign

DTE: Vector Correlation

Code phase depends

  • n clock bias

Carrier frequency depends

  • n clock drift

Code residual (Ξ”πœšπ‘‘π‘π‘’π‘“) , Carrier residual (Δ𝑔

𝑑𝑏𝑠𝑠)

independently estimated in two parallel threads

16

slide-26
SLIDE 26

University of Illinois at Urbana-Champaign

DTE: Vector Correlation Continued

Direct correlation involves non-coherent summation

  • Non-coherent summation across satellites to track code

phase and carrier frequency.

17

slide-27
SLIDE 27

University of Illinois at Urbana-Champaign

DTE: Max Likelihood Estimation

Maximum likelihood clock state 𝑼𝑡𝑴𝑭

𝑑𝑝𝑠𝑠

π‘˜ = 𝑑𝑝𝑠𝑠

𝑆,

𝑗=1 𝑂

𝑍𝑗 π‘‘πœ€π‘’π‘˜, π‘‘πœ€ π‘’π‘˜ π‘ˆπ‘π‘€πΉ = 𝑛𝑏𝑦

π‘˜=1,..,𝑄 𝑑𝑝𝑠𝑠 π‘˜

= [π‘‘πœ€π‘’π‘π‘€πΉ, π‘‘πœ€ 𝑒𝑁𝑀𝐹] Where, 𝑄= number of grid points 𝑆= incoming raw GPS signal 𝑍𝑗= π‘—π‘’β„Žsatellite signal replica

18

slide-28
SLIDE 28

University of Illinois at Urbana-Champaign

DTE: Robustness

Strong signal environment

... ...

Weak signal environment Direct Time Estimation Direct Time Estimation more robust than Scalar Tracking

Across the satellites

19

slide-29
SLIDE 29

University of Illinois at Urbana-Champaign

MRDTE: Architecture

All receivers triggered by a common clock 1 3 4 2 Raw GPS signals from multiple receivers Output from PMU: Synchronized phasor measurements

MRDTE Filter

MRDTE

Direct Time Estimation

PMU Time

20

slide-30
SLIDE 30

University of Illinois at Urbana-Champaign

MRDTE Filter: Kalman Filter

Measurement error Measurement update

𝑓𝑒.𝑙 π‘ˆ

𝑒,𝑙

Overall Time update Overall measurement update

π‘ˆ

𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š

π‘ˆ

𝑒+1,𝑙

21

  • Prediction model:

π‘ˆπ‘’+1,𝑙 = 1 Ξ”π‘ˆ 1 π‘ˆπ‘’,π‘π‘€π‘“π‘ π‘π‘šπ‘š

  • State vector π‘ˆπ‘’,𝑙 = π‘‘πœ€π‘’π‘™

π‘‘πœ€ 𝑒𝑙

  • Error covariance matrix is

calculated by processing the last 19 measurement errors

slide-31
SLIDE 31

University of Illinois at Urbana-Champaign

MRDTE Filter: Overall Filter

Measurement error Measurement update

𝑓𝑒.𝑙 π‘ˆ

𝑒,𝑙

Overall Time update Overall measurement update

π‘ˆ

𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š

π‘ˆ

𝑒+1,𝑙

22

  • Overall filter to obtain the

final corrected clock state π‘ˆπ‘’,π‘π‘€π‘“π‘ π‘π‘šπ‘š

  • Measurement error matrix

𝑓𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š = π‘ˆ

𝑒,1 βˆ’

π‘ˆ

𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š

: π‘ˆ

𝑒,𝑙 βˆ’

π‘ˆ

𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š

π‘ˆ

𝑒,𝑀 βˆ’

π‘ˆ

𝑒,π‘π‘€π‘“π‘ π‘π‘šπ‘š

Where π‘ˆ

𝑒,𝑙 = π‘‘πœ€π‘’π‘™

π‘‘πœ€ 𝑒𝑙 𝑙 = 1. . 𝑀

slide-32
SLIDE 32

University of Illinois at Urbana-Champaign

Outline

Motivation and Objectives GPS Conventional approach Multi-Receiver Direct Time Estimation (MRDTE) Experimental setup Results and Analysis Ongoing Work Summary

24

slide-33
SLIDE 33

University of Illinois at Urbana-Champaign

Experimental Setup

1 2 3 4

25

  • 4 receivers on the rooftop of Talbot Lab, Urbana, Illinois
  • Placed along the corners of square with diagonal length 10m
  • Mimic the setup of a original power substation
slide-34
SLIDE 34

University of Illinois at Urbana-Champaign

Experimental Setup: Continued

26

  • 4 USRP’s used for collecting

GPS signals

  • All the receivers triggered by

a common external clock - Chip Scale Atomic Clock (CSAC)

  • For processing the data:

pyGNSS - object oriented python platform developed by

  • ur lab
slide-35
SLIDE 35

University of Illinois at Urbana-Champaign

Outline

Motivation and Objectives GPS Conventional approach Multi-Receiver Direct Time Estimation (MRDTE) Experimental Setup Results and Analysis Ongoing Work Summary

27

slide-36
SLIDE 36

University of Illinois at Urbana-Champaign

Jamming: Carrier Frequency

28

MRDTE Scalar MRDTE (loses track at 17dB added jamming)

  • ffers 5dB more noise tolerance than

Scalar Tracking (loses track at 12dB added jamming)

slide-37
SLIDE 37

University of Illinois at Urbana-Champaign

Jamming: Code Frequency

29

MRDTE Scalar MRDTE offers better convergence and smaller variance to external noise interference

slide-38
SLIDE 38

University of Illinois at Urbana-Champaign

Jamming: Different Levels

30

At 12𝑒𝐢 jamming, MRDTE maintains a residual in clock bias of < 100π‘œπ‘‘ and clock drift of < 1.5π‘œπ‘‘/𝑑 Clock bias Clock drift

slide-39
SLIDE 39

University of Illinois at Urbana-Champaign

Jamming: Single vs Multiple

Multiple receivers show smaller variance in the clock bias as compared to single receiver Clock bias Clock drift

31

slide-40
SLIDE 40

University of Illinois at Urbana-Champaign

Meaconing: Carrier Frequency

32

Scalar tracking is operational until 2dB of added meaconed signal while MRDTE is operational till 5dB MRDTE Scalar

Maintains track Locks onto meaconed signal

slide-41
SLIDE 41

University of Illinois at Urbana-Champaign

Outline

Motivation and Objectives GPS Conventional approach Multi-Receiver Direct Time Estimation (MRDTE) Experimental Setup Results and Analysis Ongoing Work Summary

34

slide-42
SLIDE 42

University of Illinois at Urbana-Champaign

Ongoing Work

  • Objective:
  • Comparison of the performance robustness of the

MRDTE and Scalar tracking using RTDS setup

Timing: Scalar tracking

Used to trigger virtual PMUs in RTDS RTDS stability analyzed

SEL-2488

Raw GPS signals

35

slide-43
SLIDE 43

University of Illinois at Urbana-Champaign

Ongoing Work

  • Raw GPS signals are supplied to SEL-2488 (external

clock) to trigger virtual PMU and the hardware PMU is triggered using our MRDTE algorithm.

Timing: Scalar tracking

Used to trigger virtual PMUs in RTDS Used to trigger hardware PMU connected to RTDS RTDS stability analyzed

Timing: MRDTE

Raw GPS signals

SEL-2488 PMU USRP-LFTX

36

slide-44
SLIDE 44

University of Illinois at Urbana-Champaign

Work done till now

  • Generated the IRIG-B000 timing pulse: Input to PMU
  • Created a voltage shifter to convert the transmitted USRP-

LFTX 0-1v IRIG-B signal to 0-5v IRIG-B000 signal

0-1v IRIG-B000 0-5v IRIG-B000

38

slide-45
SLIDE 45

University of Illinois at Urbana-Champaign

Upcoming Work

  • Timing attacks are simulated and added to the raw GPS

signals being supplied to the SEL-2488 and USRP-LFTX.

Timing: Scalar tracking

Used to trigger virtual PMUs in RTDS Used to trigger hardware PMU connected to RTDS Raw GPS signals RTDS stability analyzed

Timing: MRDTE Timing attacks introduced

37

slide-46
SLIDE 46

University of Illinois at Urbana-Champaign

Summary

  • Proposed a novel Multi-Receiver Direct Time Estimation

(MRDTE) algorithm

  • Verified the increased noise tolerance and successful

mitigation of meaconing attack

  • Work being done in evaluating the impact of the MRDTE
  • n power grid

39

Timing Attack MRDTE Scalar Jamming 17dB 12dB Meaconing 5dB 2dB

slide-47
SLIDE 47

University of Illinois at Urbana-Champaign

Acknowledgements: This material is based upon work supported by the Department of Energy under Award Number DE-OE000078

Thank You

Special Thanks to: Prosper and Jeremy for helping with the experimental setup

  • f power grid and in carrying out the evaluations

40