Project

General

Profile

Issues #5347

How should time correlated GOOSE change status in GOOSE Performance test

Added by Hua Qin almost 2 years ago. Updated over 1 year ago.

Status:
Resolved
Priority:
Normal
Due date:
10/04/2022 (about 18 months late)
Discuss in Upcoming Meeting:
No
Clause Reference:
61850 Standard:
8-1
Triggering Tissue:
Final Decision:
Initial Test Document:
Updated Test Document:
Test Case ID:
Closed Reason:
Test Procedure Update
Triggering Tissue 2:
Triggering Tissue 3:

Description

Please clarify how should the time correlated GOOSE change status in GOOSE performance test.

From what I read below and understand, there are 5 GOOSE which has the following scheme 1:
-4ms: GPFsubscribed1 only change status of 5th data
-2ms: GPFsubscribed2 only change status of 5th data
0ms: GPFsubcribed3 change status of 5th data, GPFPPing change the status
2ms: GPFsubscribed4 only change status of 5th data
4ms: GPFsubscribed5 only change status of 5th data

But there is another understanding - scheme 2:
-4ms: GPFsubscribed1/2/3/4/5 all change status of 5th data
-2ms: GPFsubscribed1/2/3/4/5 all change status of 5th data
0ms: GPFsubscribed1/2/3/4/5 all change status of 5th data, GPFPPing change the status
2ms: GPFsubscribed1/2/3/4/5 all change status of 5th data
4ms: GPFsubscribed1/2/3/4/5 all change status of 5th data

In device with contact inputs rather than GOOSE, there is a general setting like debounce time to filter the situation in scheme 2 and thus will not have 5 changes within 8ms.


Files

#1

Updated by Bruce Muschlitz almost 2 years ago

It is "obvious" (to me) that because clause 4.1.3 of UCATestProceduresGoosePerformanceEd2_Rev2p5.pdf says for the time correlated subscribed GOOSE:
Retransmission at 4, 32 and 256 ms (or more) that it is means that GPFsubscribed1 has an incremented state number ONLY at time= -4 ms; and not at -4 and -2 and 0 and 2 and 4 ms.
Similarly GPFsubscribed2 does not have incremented state numbers at -4 or 0 or +2 or +4.

This results in exactly 6 time-correlated subscribed GOOSE messages centered upon relative time 0 ms (both GPFPPpingNormal and GPFsubscribed3 are sent at 0 ms).
To produce the most accurate measurements, it is assumed that GPFPPpingNormal is sent slightly before GPFsubscribed3 to avoid queuing issue on the Ethernet media. In other words, the specification implies "these are simultaneous" but sending first the GPFsubscribed3 GOOSE would add a delay to the PING-PONG GOOSE name GPFPPpingNormal.

For the above, please substitute GPFPPpingLarge for tests Gpf4 and Gpf8.

#2

Updated by Hua Qin almost 2 years ago

Thank you Bruce.

I propose to change the "5 subscribed GOOSE each having one .... used for ping-pong" to the following words:

5 subscribed GOOSE each having one state change of the 5th data value element approximately as following before and after the subscribed GOOSE state change used for ping-pong:
At -4ms: subscribed GPFsubscribed1 GOOSE change status of 5th data
At -2ms: subscribed GPFsubscribed2 GOOSE change status of 5th data
At 0ms: subscribed GPFPPing GOOSE change the status and followed by subscribed GPFsubcribed3 GOOSE change status of 5th data
At 2ms: subscribed GPFsubscribed4 GOOSE change status of 5th data
At 4ms: subscribed GPFsubscribed5 GOOSE change status of 5th data

#3

Updated by Bruce Muschlitz almost 2 years ago

  • Discuss in Upcoming Meeting changed from No to Yes
#4

Updated by Joel Greene over 1 year ago

  • Status changed from New to In Progress
  • Assignee set to Richard Schimmel
  • Discuss in Upcoming Meeting changed from Yes to No

Richard to clarify test case.

#5

Updated by IEC 61850 TPWG over 1 year ago

  • Due date set to 10/04/2022
#6

Updated by Richard Schimmel over 1 year ago

Attached solution

#7

Updated by IEC 61850 TPWG over 1 year ago

  • Status changed from In Progress to Resolved

Also available in: Atom PDF