OHM RV Orientation

[streb2001]

It's very hard to make Auto-docking 'realistically careful' without taking forever (for a 'gamer') to complete.

Try 10x timewarp. Scary.

 

[ADSWNJ]

Thanks for the feedback guys. I'm posting a 1.01 release today, with the source for the quick fix I made earlier in this thread, and updates to documentation.

Szymon - sorry about the typo. I knew how to spell your name, honest! On the free software ... there's a philosophy of free software that I didn't quite capture in my simple terms bit, and I'll try again! It's still GPL though, per our earlier discussion on this vs LGPL, and we want to set an example for developers to open-source the code, so it lives on after they have left the community.

Boogabooga - you and Enjo were both right. The previous release was a work-in-progress beta, that confirmed for me that I needed a totally new approach to the autopilot. I deleted the beta from OrbitHangar yesterday to replace with this new one. <-- (I think this deleted the whole beta thread too ... interesting!)

Axel and C3PO - I've added the HUD Drawer link to the documentation. I've also added commentary about this missing VesselHooking.dll (indicates you have not installed the HUD Drawer), and missing MSVCP100 or MSVCR100 (indicates you have not installed the MS VC++ 2010 redistributables).

On making a realistically careful auto-dock (C3PO) - yeah ... I kept thinking "overshoot is not an option"! The interesting thing is the massive (e.g. 30x) difference in thruster performance between realistic and non-realistic vessels. Working with the CEV-Orion, I really had to work hard to get the main AP code to handle such low thrust levels. I'm going to continue working on further refining the AP thruster sequence, to damp some of the unnecessary (to a human) thrusting.

I also want to sort out my own thruster groups for each vessel to be invariant of the RCS Docking Mode, etc. I know which ship coordinate direction I want to go, and I can see where the thrust vectors for each thruster go, so I don't need to rely on the default thruster groups (e.g. THGROUP_ATT_FORWARD) which are intentionally remapped by the vessel coder when you change docking mode.

Oh and I want to make a 3D docking corridor now for the whole flight sequence, so you see a corridor from where you are to the target, curving in space. And if it works there, I'll add it to Glideslope too. W00t!

---------- Post added at 02:18 PM ---------- Previous post was at 01:55 PM ----------

Version 1.01 now released: [ame="http://www.orbithangar.com/searchid.php?ID=6221"]http://www.orbithangar.com/searchid.php?ID=6221[/ame]

 

[garyw]

Oh and I want to make a 3D docking corridor now for the whole flight sequence, so you see a corridor from where you are to the target, curving in space. And if it works there, I'll add it to Glideslope too. W00t!http://www.orbithangar.com/searchid.php?ID=6221

One word.

WANT.

Sounds truly awesome ADSWNJ!

 

[Star sailer]

missing dll

I get an error when trying to activate the module: vesselhooking.dll was not found. The module activates but I cant see the mfd in the mfd selection screen.

 

[Kendo]

Brilliant. :tiphat:
I don,t know how important this is, but would the word Docked be useful when actually docked.

 

[Zatnikitelman]

I also want to sort out my own thruster groups for each vessel to be invariant of the RCS Docking Mode, etc. I know which ship coordinate direction I want to go, and I can see where the thrust vectors for each thruster go, so I don't need to rely on the default thruster groups (e.g. THGROUP_ATT_FORWARD) which are intentionally remapped by the vessel coder when you change docking mode.

I'm really looking forward to this part, since it seems that despite reconfiguring the thrusters, Shuttle Fleet doesn't want to keep the thrusters configured that way for use in MFDs.

 

[ADSWNJ]

I'm really looking forward to this part, since it seems that despite reconfiguring the thrusters, Shuttle Fleet doesn't want to keep the thrusters configured that way for use in MFDs.

The code is a bit convoluted, but I'll get there. It'll also improve safety when a ship has an incomplete set of thrusters and I blindly try to use them!

For those interested in the code - here's what I'm thinking:

GetThrusterCount() gives you how many thrusters (or engines) are on the ship. Max dimension for a thruster array.

ThurusterGroupDefined(thgp) gives you a yes or no to whether the thruster group is defined. (If not - disable the AP.)

GetThrusterGroupHandle(thgp) gets you a handle to a thruster group (e.g. THGROUP_ATT_RIGHT). From here, find out how many thrusters are in the group, via GetGroupThrusterCount(thgph). Then iterate through the thrusters pulling one at a time from the group via GetGroupThruster(thgph,ix). For each thruster, you can then pull out GetThrusterRef (the point on the ship), and GetThrusterDir (the direction we want). For the hell of it, I'll also pull out GetThrusterMax0 to see if I can find a direct way to calculate translation coefficients (F = ma, so there should be a correlation to this Max0 value!).

Once I have all that, I'll work out what each thruster group is actually doing, and what the ship-invariant thruster group names would be. (Note at this point, the port may be anywhere, and the reference thrusters could be mapped to the ship or the port, so I have to work it out myself).

Finally - I'll create 12 user defined thruster groups using CreateThrusterGroup() to give the desired set of translations and rotations relative to the ship's axis.

One concern is always "what happens if a ship has a port not in a cardinal direction"? I have yet to see this, but it's possible. In this case, I'll just disable the AP I think, and may your chosen deity have mercy on anyone trying to dock such a beast by hand.

---------- Post added at 07:40 PM ---------- Previous post was at 07:32 PM ----------

I'm really looking forward to this part, since it seems that despite reconfiguring the thrusters, Shuttle Fleet doesn't want to keep the thrusters configured that way for use in MFDs.

Can you help explain a bit more about how you dock the Shuttle. Is this the stock Atlantis in the distribution or an add-on (if so, which one)? Do you have a mode that looks out of the docking port (how?). And do you have a mode to flip the RCS thrusters to the docking orientation (how?).

I know how to do this on the XR-5, but not on the Shuttle. (Or at least it's been that long since I flew it!).

I'm thinking I need a CAO button (coordinate axis orientation) that allows you to select SHIP or PORT orientation. This would let the commander dock from his seat, looking out the front, with the MFD and HUD either in ship mode (so translating UP to dock), or in PORT mode (so translating FWD to dock) as he prefers. With the invariant thruster setup, it would make no difference to the AP engine, just to the visuals.

---------- Post added at 07:44 PM ---------- Previous post was at 07:40 PM ----------

Try 10x timewarp. Scary.

Try 100x ... it's even more aggressive. It does a nice line-up, but it struggles to dock at that rate due to the safety back-off coding. So you may need to step it back to 10x for the last 2 meters!

Another nice thing to see is the "grand tour" of the ISS. Set it to port 1, distance 100m, and let the AP take it in to say 2 meters. Then hit PRT and reengage the APP and flip to outside view. The ship figures out the rotation, backs out, loops round and picks up the next port. Repeat through all ports. The Port 2 to 3 transition is particularly interesting.

---------- Post added at 07:45 PM ---------- Previous post was at 07:44 PM ----------

Brilliant. :tiphat:
I don,t know how important this is, but would the word Docked be useful when actually docked.

Will do. I switch off the AP when I'm docked, and the aural cue is of course the docking latches grabbing hold of us, but I'll flash something on the HUD and MFD as well.

 

[orb]

I get an error when trying to activate the module: vesselhooking.dll was not found.

Read a few posts back from yours.

 

[Zatnikitelman]

Can you help explain a bit more about how you dock the Shuttle. Is this the stock Atlantis in the distribution or an add-on (if so, which one)? Do you have a mode that looks out of the docking port (how?). And do you have a mode to flip the RCS thrusters to the docking orientation (how?).

I know how to do this on the XR-5, but not on the Shuttle. (Or at least it's been that long since I flew it!).

I'm thinking I need a CAO button (coordinate axis orientation) that allows you to select SHIP or PORT orientation. This would let the commander dock from his seat, looking out the front, with the MFD and HUD either in ship mode (so translating UP to dock), or in PORT mode (so translating FWD to dock) as he prefers. With the invariant thruster setup, it would make no difference to the AP engine, just to the visuals.

It's Dave413's Shuttle Fleet v4.8 on Simviation. When you switch into the docking view mode with SHIFT-Z, the thrusters are reconfigured to act like the Deltaglider, or any nose-dock ship. However, the MFD can't seem to trigger it.

As a test, I took the XR5, and put it into its docking mode and from the start it worked perfectly! I then re-calibrated the thrusters, like I tried with Shuttle Fleet, and the results are night and day. It's kind of hard to explain, but the XR5 performed exactly as advertised, like when the yaw right phase was ongoing, in the XR5, the stars all slid toward screen left, while in the Shuttle Fleet Orbiter, the stars rotated clockwise on the screen.

 

[ADSWNJ]

It's Dave413's Shuttle Fleet v4.8 on Simviation.

It that still a live addon? Per here: http://www.orbiter-forum.com/showpost.php?p=340040&postcount=1947 I think it's dead.

Anyway - if you can PM me a link, I'll download and install it and debug exactly what it's doing.

 

[boogabooga]

It that still a live addon? Per here: http://www.orbiter-forum.com/showpost.php?p=340040&postcount=1947 I think it's dead.

Anyway - if you can PM me a link, I'll download and install it and debug exactly what it's doing.

There was a whole "thing" with Shuttle Fleet last year. It was taken off of OH and moved to another site. Though new development stopped, it is excellent as it is and you should assume that it will remain very popular for quite some time.

Don't consider it "dead."

 

[ADSWNJ]

Don't consider it "dead."

Can't find it though. If you can give me a URL for it, then I'll debug and fix whatever it needs to work completely with RV Orientation (HUD not guaranteed - as that depends on how old a vessel class was used by the original coder.)

 

[Cras]

Shuttle Fleet is far from dead. It just isnt posted here, or on the hangar. You have to go elsewhere.

I always had some issue with the docking cam with SF myself, getting it to kick over correctly. the hud is nice, but really, I just end up staying in the VC and dock by looking out the aft workstation windows.

 

[Keatah]

Do you have a program to interpret the contents of the autopilot dump? Or a bit of documentation?

 

[ADSWNJ]

Do you have a program to interpret the contents of the autopilot dump? Or a bit of documentation?

You win the prize for the first to ask for this . I'll post back the column definitions when I get a chance.

 

[Keatah]

You could just include them as a chapter or appendix in the PDF documentation file that is distro'd with add-on itself. That way everybody gets a copy.

 

[ADSWNJ]

You could just include them as a chapter or appendix in the PDF documentation file that is distro'd with add-on itself. That way everybody gets a copy.

Will do.

There's just something cool about this set of people on this forum. Interested in every line of documentation, wanting to understand the debug files, and giving great feedback. I'm proud of you guys !!!

---------- Post added at 01:53 AM ---------- Previous post was at 01:26 AM ----------

Here's the decode for the AP dump file:

SimT - simulation time
StepAvg - average step duration of the past 10 intervals (for calculating thrust durations)
Dx, Dy, Dz ... directions in degrees (rotation about X = Pitch, about Y = Yaw, about Z = roll)
Px,Py,Pz ... positions in meters (X = right, Y = up, Z = fwd)


Dx = Pitch offset in degrees
Dx.V = Velocity (rate) of the pitch offset in degrees/sec
Dx.L = AP fuzzy logic determined limit for the rate of pitch in degrees/sec
Dx.Th = Thruster magnitude (-1.00 to 1.00) for pitch up (-ve for down)
Dx.C = Estimated sim cycles to attain desired offset at full thrust
Dx.A = Reference calibrated acceleration of pitch in degrees/sec^2 for our current weight at full thrust
Dx.B = AP fuzzy logic accumulator bias (used to nudge the rate when very low thrust needed)

Same applies to Dy, Dz, Px, Py, Pz.

Pz.Tgt overrides the Z direction target to hold the ship or back it off when out of cone.

AppInCone = 1 when approach is in the allowed cone
MaxDirErr = degrees of orientation error permitted at this range
MaxPosErr = meters of translation error permitted at this range

RotActive = AP working on ROT in this cycle
AttActive = AP working on ATT this cycle (interlocked to not do both on same cycle)
AppActive = AP working on FWD/BACK this cycle
AppAttAllow = Interlock to prevent ATT mode when ROT errors are too big
ApToggle = Counter for how many ATT cycles per ROT when we are settled down

Range = 3D spherical range from target
Our Lat, Long = Lat Long from target port's frame to our port
WP Lat, Long = current waypoint in SWP mode
Mode = -1 for AP Calibrate, 0 for dock, 1 for WP, 2 for SWP, 3 for RVEL.


And that's just the dump of how much it takes to do a nice approach AP

 

[Axel]

So I installed Hudddrawer and the both links of the needed runtime-librarys, because I use win7 64bit prof. I activated then in the orbiter start menu Huddrawer and RV orientation, but i got a error message again:
"Can't jump to the needed sequence XXX at msvcr100.dll" (its a translation from german, maybe not 100% the message in english then).
Later in the simulation I did not see your MFD.

After that I downloaded the msvcr100.dll seperate and put them direct into the orbiter-directory. And Vola, no error message at start-menu of orbiter and later in the sim your MFD!
The documentation of the DLL says that a mounting to the windows should be possible too, to avoid placing it in the orbiter directory itself alltimes. But I did not try it.
Now I'm going to begin the test phase of your MFD itself. Some trys with the DG was very succesfully. Next step I test the calibration mode of my favorite vessels.

Thanks for the addon, its looking good!

 

[ADSWNJ]

@Axel - I'm glad it's working, but you really should not need to put that runtime file into the Orbiter directory. I'm not sure what is going on without looking more closely, but you could search your whole hard disk for that file and see which directories it is in. I'm guessing you have installed the runtimes into other directories in the past maybe?

Anyway - great that it now works for you, and let me know how you get on with those new vessels.

 

[Iacomus Maximus]

I am having the same issue and was just looking to see if anyone else was.
Thank you, @Axel, for your research, I will try that also.
-----------------------------------------------------------------
UPDATE:
I copied a newer version of msvcr100.dll into my orbiter directory and now everything works fine.