Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 34

Thread: One activity with Fall Harvest on CN Ruel Subdivision

  1. #21
    Join Date
    Jun 2004
    Location
    Winnipeg, MB, CA.
    Posts
    187

    Default

    Ok, fair point, but what I was thinking of is this:

    Say I'm designing an activity (and I am working on a few, one of them for the LIRR, excellent route, btw!) and I use the autopilot to time the traffic. Then the traffic won't sync up when a player controls the train, because the autopilot generally accelerates/decelerates faster and reaches higher speeds than under player control. Conversely, I can go the other way, and time the traffic by not using the autopilot and strictly driving under player control, then anyone using autopilot will be out of sync. Seems like either way, someone's going to be disappointed. It might not be a big deal when the traffic is just running past on a more-or-less regular schedule, but when I try to incorporate advanced shunting, it gets complicated.

    Say I've got an AI that's going to pick up some cars and deliver them to a yard for the player to pick up as he passes, but I'd like to have him see it working as he's taking care of some other switching earlier on. So the player is switching the south side (say) and the AI is switching the north side (and maybe sometimes they interact, even), and then drops his cut where the player can pick them up. Right now, I pretty much have to time it to accommodate the autopilot, because if I sync it to a human player, someone using the autopilot will arrive too early (and they'll miss seeing the AI dance I have choreographed, lol!) Meanwhile, there's all the other traffic going by as challenges to both the north and south side switching tasks!

    All I'm saying is it would be nice if we could "weigh down" the autopilot to make it act more like a real driver.

    Bob.

  2. #22

    Default

    Hi,

    I just want to point out that OR has a few logic errors, and for the purpose of this thread and the discussion of AI behaviour, the logic error is about trains chasing trains ( whether it's an AI chasing an AI... or a player train chasing an AI or an AI chasing a player train ) where the chasing train will always take the passing siding... which means the chasing train will be put off of it's schedule. I tested it about a year ago and all my paths have passing paths, and every chasing train went into the siding... oh yes!!... I should mention that there was no oncoming traffic. I did test without passing paths and for the life of me cannot remember the outcome. For the doubters... create a test activity and see for yourself.

    Bottom line... AI traffic will present some unpredictable results as it stands now.

    Peter

  3. #23

    Default

    Never saw anything as you describe in building, testing and running of the activities on the PRR Eastv2.
    I always use the latest testing verions of Open Rails

    regards,
    ............Vince ..............
    ...... Author NECv4 .......
    .... LIRR BUILD PHOTOS ....
    .............LIRR VIDEO.............
    ...... Eschew Obsfucation ......

    On the The Statue of Liberty in New York Harbor there is a Tablet. On it is written:
    "Give me your tired, your poor, your huddled masses yearning to breathe free,
    the wretched refuse of your teeming shore, send these, the homeless, tempest-tossed to me,
    I lift my lamp beside the golden door!"

  4. #24
    Join Date
    Feb 2013
    Location
    known universe
    Posts
    1,585

    Default

    Quote Originally Posted by peterman View Post
    Hi,

    I just want to point out that OR has a few logic errors, and for the purpose of this thread and the discussion of AI behaviour, the logic error is about trains chasing trains ( whether it's an AI chasing an AI... or a player train chasing an AI or an AI chasing a player train ) where the chasing train will always take the passing siding... which means the chasing train will be put off of it's schedule. I tested it about a year ago and all my paths have passing paths, and every chasing train went into the siding... oh yes!!... I should mention that there was no oncoming traffic. I did test without passing paths and for the life of me cannot remember the outcome. For the doubters... create a test activity and see for yourself.

    Bottom line... AI traffic will present some unpredictable results as it stands now.
    Peter
    Vince: Never saw anything as you describe in building, testing and running of the activities on the PRR Eastv2.
    Me neither (in testing activities), although I don't doubt what you observed...what my questions are...
    1. How did you set up the testing environment?
    2. Was there something in your method of testing that provoked the result you observed?
    3. Can you provide a testing environment that anyone could use to repeat your results?
    Cheers, R. Steele [Gerry] It's my railroad and I'll do what I want! Historically accurate attitude of US Railroad Barons.


  5. #25
    Join Date
    Nov 1999
    Location
    Netherlands.
    Posts
    155

    Default

    Quote Originally Posted by peterman View Post
    Hi,

    I just want to point out that OR has a few logic errors, and for the purpose of this thread and the discussion of AI behaviour, the logic error is about trains chasing trains ( whether it's an AI chasing an AI... or a player train chasing an AI or an AI chasing a player train ) where the chasing train will always take the passing siding... which means the chasing train will be put off of it's schedule. I tested it about a year ago and all my paths have passing paths, and every chasing train went into the siding... oh yes!!... I should mention that there was no oncoming traffic. I did test without passing paths and for the life of me cannot remember the outcome. For the doubters... create a test activity and see for yourself.

    Bottom line... AI traffic will present some unpredictable results as it stands now.

    Peter
    It happens if the chasing train tries to clear the route through the siding area while the preceding train has not yet cleared the main line in that area. As the main path is not available while trying to set the route, the path selection logic selects the alternative path.
    It only occurs when the option 'location linked path processing' is switched on, and alternative paths are defined for that location.
    The problem has been around ever since 'location linked path processing' was introduced and has been known for some time.
    It's somewhere on the list of things to do, but that's a very long, long list.

    Regards,
    Rob Roeterdink

  6. #26
    Join Date
    Nov 1999
    Location
    Netherlands.
    Posts
    155

    Default

    Quote Originally Posted by tooltroll View Post
    So all the focus on accurate physics only applies to the player train? That seems self-defeating: If traffic ignores physics, then all the player interactions with traffic won't be accurate.

    How does the AI/autopilot currently calculate acceleration and braking? It seems to have some sense of the weight it's hauling, so could we not just multiply that by an empirically determined fudge factor to slow acceleration and braking performance?

    Bob
    Basically, what happens when a train is starting is that the program increases the TrainThrottlePercentage (this, and TrainBrakePercentage, are the only controls available to regulate an AI train). Simplified physics are used to move the train.
    The program checks the train's speed and acceleration, and if these are not as what is expected, the program will keep increasing the TrainThrottlePercentage. When the percentage has reached 100% and the train is still not moving as expected, the program 'intervenes' and just forces the train to move.
    So if the train does move properly before the throttle percentate reaches 100%, true physics are used to move the train (be it a simplified version as compared to the player train). In that situation, you may indeed see a difference in acceleration between lightweight and heavy trains.
    Braking is more or less the same, but the control is more strict to prevent run-away trains.

    Regards,
    Rob Roeterdink

  7. #27
    Join Date
    Jun 2004
    Location
    Winnipeg, MB, CA.
    Posts
    187

    Default

    Quote Originally Posted by roeter View Post
    Basically, what happens when a train is starting is that the program increases the TrainThrottlePercentage (this, and TrainBrakePercentage, are the only controls available to regulate an AI train). Simplified physics are used to move the train.
    The program checks the train's speed and acceleration, and if these are not as what is expected, the program will keep increasing the TrainThrottlePercentage. When the percentage has reached 100% and the train is still not moving as expected, the program 'intervenes' and just forces the train to move.
    So if the train does move properly before the throttle percentate reaches 100%, true physics are used to move the train (be it a simplified version as compared to the player train). In that situation, you may indeed see a difference in acceleration between lightweight and heavy trains.
    Braking is more or less the same, but the control is more strict to prevent run-away trains.

    Regards,
    Rob Roeterdink
    Thank you for the info. It's about what I expected, given the behaviour in-sim. Next silly question: You say that "The program checks the train's speed and acceleration, and if these are not as what is expected, the program will keep increasing the TrainThrottlePercentage." and so I'm wondering how the program knows what to "expect"? Could we just lower its expectations?

    Also, how fast does it advance the throttle? A player should be advancing the throttle notch by notch, waiting for the RPMs to increase, or so my stack of EMD/GE operator's manuals suggest. Seems like the AP goes from idle to notch 8 almost instantaneously, and keeps it there until the speed limit is reached. Would it break anything to slow the throttle action to something more prototypical? Surely that'd just need a delay loop to prevent the throttle from advancing for a few seconds after the last advance? If there already is a delay, could it not be lengthened a few seconds more? (I get that the brakes need to be a bit 'prompt' to prevent runaways/collisions.)

    Bob

  8. #28

    Default

    You know how OR will load a consist even when a car is missing from the trainset folder? It will just load that train without that car. One car shorter. But what happens if all the locomotives from an AI consist get inadvertently deleted?

  9. #29

    Default

    Quote Originally Posted by roeter View Post
    It happens if the chasing train tries to clear the route through the siding area while the preceding train has not yet cleared the main line in that area. As the main path is not available while trying to set the route, the path selection logic selects the alternative path.It only occurs when the option 'location linked path processing' is switched on, and alternative paths are defined for that location.The problem has been around ever since 'location linked path processing' was introduced and has been known for some time.It's somewhere on the list of things to do, but that's a very long, long list.Regards,Rob Roeterdink
    Hi Rob,
    I'm assuming that... "the preceding train has not yet cleared the main line in that area" is in reference to the NumSigAhead (4). If so, in my case, the preceding train is outside that number.

    Your statement... " It only occurs when the option 'location linked path processing' is switched on "... I believe is incorrect. I don't have that option checked nor have I ever had it checked. It causes problems with switching activities. I first noticed what I'm talking about around the time that random activity and random weather were introduced. Don't quote me on that. 8-)

    Gerry,

    Pick a route, make sure it has passing sidings, add an AI that goes from one end to the other. Now add a player train that follows the same path and have it start 5 - 10 minutes after the AI. Start OR and drive your train or watch it on the track viewer.

    Peter

  10. #30
    Join Date
    Feb 2013
    Location
    known universe
    Posts
    1,585

    Default

    Quote Originally Posted by peterman View Post
    ...Gerry,
    Pick a route, make sure it has passing sidings, add an AI that goes from one end to the other. Now add a player train that follows the same path and have it start 5 - 10 minutes after the AI. Start OR and drive your train or watch it on the track viewer.
    Peter
    Thank you Peter, will do.
    Regards, Gerry
    Cheers, R. Steele [Gerry] It's my railroad and I'll do what I want! Historically accurate attitude of US Railroad Barons.


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •