[Solved] Flank protection

Moderator: dadolphs

Post Reply
rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 19.09.2015, 10:45

Hi Erik,

I cannot reproduce your issue; Both trains can lock and run parallel.
I tested with the block sensors only.

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 19.09.2015, 11:14

Hi Rob

First schedule
20150919.112131.202 r9999I cmdr029B OLoc 3319 processing cmd="useschedule" for [V100]

finds route
20150919.112131.389 r4201I 000013A8 OLcDrive 0115 Found destination for [V100]: [B12] by route [21-12]

checks flank protection (why ends S03a up with Rejecting all commands)
20150919.112131.405 r9999c V100 OSwitch 1017 check flank protection for switch [S03a]: S03b,straight,locked
20150919.112131.405 r9999c V100 OSwitch 1131 Switch [S03a] will be set to [straight,0,0]
20150919.112131.405 r9999c 000013A8 OXpressN 0360 switch 1-3 turnout
20150919.112131.405 r9999c V100 OSwitch 1131 Switch [S03b] will be set to [straight,0,0]
20150919.112131.405 r9999c V100 OSwitch 1017 check flank protection for switch [S03a]: S03b,straight,locked
20150919.112131.420 r9999I transact OXpressN 1318 processing post 0x029BD4CC...
20150919.112131.420 r9999c timedque OXpressN 1255 new timed command time=1841 delay=10 tick=1841
20150919.112131.420 r9999c V100 OSwitch 1458 switch[S03a] is flank protecting; Rejecting all commands

Second schedule
20150919.112134.478 r9999I cmdr029B OLoc 3319 processing cmd="useschedule" for [NFJ SM3]

finds route
20150919.112134.649 r4201I 00001658 OLcDrive 0115 Found destination for [NFJ SM3]: [B13] by route [22-13]

checks flank protection (ends with an error on switching S03a. There should be no switching at all)
20150919.112134.665 r9999c NFJ SM3 OSwitch 1017 check flank protection for switch [S03b]: S03a,straight,locked
20150919.112134.665 r9999c NFJ SM3 OSwitch 0951 switch [S03b] is locked by [V100]: reject any commands from others
20150919.112134.665 r9999c NFJ SM3 OSwitch 1458 switch[S03a] is flank protecting; Rejecting all commands
20150919.112134.665 r9999E NFJ SM3 ORoute 0478 Error on switching S03a:straight

and after the first scheduled is finished, the second starts.

This should document the problem.

And a small question. Message 1017 is unlogical to me. I would think that flank protection is checked for the switch in the route, not the flank switch.

Regards
Erik

rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 19.09.2015, 11:22

Hi Erik,

you switched of the option to skip switch commands if they are already in position...

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 19.09.2015, 12:13

Hi Rob

On thuesday you wondered why I have set the option to skip commands on. Now it is mandatory to have the option on.

I have understood that flank protection should work regardless of this option.

Regards
Erik

rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 19.09.2015, 13:44

Hi Erik,

I improved the flank protection so it will work with or without the skip option set. (9856+)

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 19.09.2015, 14:51

Hi Rob

Fine

I will test asap

Regards
Erik

cds
Moderator
Posts: 5231
Joined: 03.02.2012, 19:24
Location: Tullnerbach, Austria

Re: Flank protection

Post by cds » 21.09.2015, 11:15

Hello all!

Back from a longer absence I've read this thread at first by random. I appreciate this feature a lot! Another step forward to how real electronic interlockings work.
Bravo to Erik and Rob!

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 21.09.2015, 19:43

Hi Claus

Thank you for your comments. I am very pleased using Rocrail and the way new ideas can be developed. Rob and the rest of the Rocrail team do a great job.

Hi Rob

I have now had oppertunity to test.

And after I have read the wiki many times, I have understood, that the definition of flank protection is not on the protected switch, but on the protecting switch.
That is if a switch in a route needs protection, the protection must be defined on the protecting switch. In my first tests with only 2 switches, the definitions are the same but I have included new switches, and then the definition was clear to me.

But I think it is a bit difficult to understand the flank protection definition.
It says:
If the locked switch is locked in the locked state (the rightmost fields), this switch (the one in the heading) must be set to state (the leftmost field) and locked from being switched but not locked from being used in a route.

I realize that the definition has been like that all the time.

But back to the test.

The locking seems OK, but
  • if the protecting switch is occupied (that is locked by a occupancy sensor) then the protected switch can be locked by a route setting no matter what position the protecting switch is in. Then there is no protection at all. If the protecting switch is in the wrong position the route setting must wait for the switch to be switched.
  • similarly if the protecting switch is protecting a switch and is requested to protect another switch but in different position the route setting must wait. Again no protection
  • The locking does not seem to work using menu/routes control
If you need documentation in form of issues please tell me

Until now I have tested on switches without position sensors but the next tests will be with sensors which makes the switch movement slow, which will show if the flank protection waits for the protecting switch to be in correct position.

Regards
Erik

smitt48
Moderator
Posts: 6111
Joined: 04.04.2014, 01:07
Location: Kralendijk, Bonaire - Dutch Caribbean

Re: Flank protection

Post by smitt48 » 21.09.2015, 20:27

Hi Erik,
But I think it is a bit difficult to understand the flank protection definition.
It says:
If the locked switch is locked in the locked state (the rightmost fields), this switch (the one in the heading) must be set to state (the leftmost field) and locked from being switched but not locked from being used in a route.
Where is that definition in the wiki?

Thanks Tom

rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 22.09.2015, 06:14

ejp wrote:
  • if the protecting switch is occupied (that is locked by a occupancy sensor) then the protected switch can be locked by a route setting no matter what position the protecting switch is in. Then there is no protection at all. If the protecting switch is in the wrong position the route setting must wait for the switch to be switched.
Can't reproduce it with your example. (Route is electrically occupied...)
ejp wrote:
  • similarly if the protecting switch is protecting a switch and is requested to protect another switch but in different position the route setting must wait. Again no protection
Can't reproduce it with your example.
ejp wrote:
  • The locking does not seem to work using menu/routes control
This is only for testing.

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 22.09.2015, 10:05

Hi Tom and Rob

@Tom. The description is not in the wiki but is my interpretation to make it clear what happens.Perhaps the information in it should be written in the wiki.

@Rob. I have made issues describing both tests. Description of each test is in issue.txt.

And generally. I still think that the flank definition should be made on the route and not the switch. Protection is for routes not for the physical switches. This would be more understandable to users.
This way it would be easier to enhance flank protection to signals, narrow spaces, putting off uncouplers etc.

But the problems in my test are the same.

Regards
Erik
You do not have the required permissions to view the files attached to this post.

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 22.09.2015, 10:08

Hi Rob

I forgot.

Shouldn't using the menu Control/Routes control do the same as setting routes from schedule? Otherwise it is not a test.

Regards
erik

rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 22.09.2015, 10:16

Hi Erik,
ejp wrote: Shouldn't using the menu Control/Routes control do the same as setting routes from schedule? Otherwise it is not a test.
No it is not because this control uses some options which disable many other things in the route and switches.

rjversluis
Site Admin
Posts: 42326
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Re: Flank protection

Post by rjversluis » 22.09.2015, 10:18

ejp wrote:And generally. I still think that the flank definition should be made on the route and not the switch. Protection is for routes not for the physical switches. This would be more understandable to users.
Great! I implemented it for you as a switch option... :shock: I will remove the complete Flank Protection.

ejp
Posts: 117
Joined: 01.10.2012, 08:20
Location: Denmark

Re: Flank protection

Post by ejp » 22.09.2015, 10:35

Hi Rob

When we started talking about flank protection, I made a description on defining flank protection on the route. While I was making that description and before submitting it you made it on the switch and I accepted that. As a comment to Tom about my understanding/misunderstanding of the wiki I described that I still think it was better defined on the route.

But now it is not on the route!

Defining on the switch is definitely better than nothing. So please reconsider the removal one more time.

Regards
Erik

Post Reply

Return to “Basic topics”