Automatic speed control

[cesarBLG]

Several improvements to the cruise control subsystem, including:

• Interface with the TCS to implement Automatic Train Operation. Exposed parameters are SetSpeedMpS and SetSpeedAccelerationMpSS which determine the permitted speed and deceleration of the braking curve.
• Use a linear conversion from 'delta' speed (difference between set speed and current train speed) to deceleration. The legacy algorithm (which is kept as default to maintain compatibility) is non-linear and causes oscillations for small deltas (small changes in speed result in big changes in throttle demand).
• Simplified the code

https://blueprints.launchpad.net/or/+spec/cruise-control-extensions

https://www.elvastower.com/forums/index.php?/topic/33841-cruise-control/page__view__findpost__p__318664

[Csantucci]

Hi César, about this https://www.elvastower.com/forums/index.php?/topic/39072-u20250824-0738-restore-crash-using-a-locomotive-with-tcs-and-cc/#entry321033 I made some investigation and saw that the crash started to occur when I merged the present PR into ORNYMG. I checked the code and believe I detected a bug in the Save/Restore code of the MaxForceSelectorController within CruiseControl.cs.

The MaxForceSelectorController Save/Restore code calls two methods within MSTSNotchController.cs. However those two methods aren't symmetrical. The Save code within MSTSNotchController.cs has a line outf.Write((int)ControllerTypes.MSTSNotchController); which has no corresponding line in the Restore code. And in fact, if you check another place where the Restore code is called, that is MSTSWagon.cs where WeightLoadController data are restored, the restore code is as follows: var controllerType = inf.ReadInt32(); WeightLoadController.Restore(inf);

Moreover I don't know if a null check for the MaxForceSelectorController has to be done before saving and restoring it.

[cesarBLG]

Thanks for reporting Carlo. I'm travelling and I don't have access to my PC, I'll fix it as soon as I come back.

I think it's not wise to have non-symmetrical save/restore for the NotchController, which differs from the rest of the save/restore methods, so I'll take a look.

[cesarBLG]

I uploaded a fix. However, It looks like all the controller save/restore process is unnecessarily complicated. I don't know why it was done in such a way, but I'll try to simplify it.