[REL] Tutorial for tracking with sonar

[XLjedi]

Actually I dunno...

(funny how we're arguing on behalf of each other now )

After I typed that response (early this morning) it dawned on me that I was thinking in reference to a relative motion plot on a maneuvering board where all lines would radiate from the center.

In your case, you're plotting bearings on a DRT (dead reckoning tracer). The inherent motion I was thinking about I believe is accounted for in your update of OwnShip's position. Your plot looks more like the TMA lines in Dr.Sids Least Squares TMA solution example rather than one of my relative motion examples.

In MoBo there is a TMA tool that will do the measuring between the lines to help determine the target TC. Would you mind if I worked up an example or two on MoBo? ...for the benefit of those who like to use it. Of course, it would be even better if I could enlist you to do it for me.

Your example and Dr.Sid's original examples are both currently not represented with tutorials in MoBo. However, since MoBo is both a DRT and a Maneuvering Board, it is perfectly capable of handling the technique.

Incidentally, what graphics program are you currently using to create your tutorial images?

[jerm138]

I've never used MoBo, so I'm not sure I'd be much help there. I didn't even know it existed until this thread.

If you could come up with a way to prove or disprove this method, that'd be great.

I know the part about an intersecting line with equidistant segments is correct. The only part that would really need tested is the method for finding said line when OwnShip is not stationary.

For my tutorial images, I'm just using PowerPoint and exporting to .pdf. PowerPoint is an excellent vector drawing tool if you know how to use it. I'm a vector-art fanboy... raster doesn't cut it for me. I've tinkered with powerpoint for many years and I've made some pretty complex images with it. I even used it to re-make Cap'n Scurvy's AoB circular slide rule. I posted pictures of it here. It comes in pretty handy.

[XLjedi]

I don't think it needs further proving or disproving... However, I could easily prove that a two-legged series of equidistant observations while moving will yield a correct solution for true course.

I need to review Dr.Sid's logic now. There was a time when I was considering adding a least-squares TMA tool to MoBo, and I just haven't gotten around to it. I think Sid's least-squares method didn't even require that the observations be spaced at equivalent time intervals. On the one hand, that's pretty cool, on the other hand, purests might argue that the method would require more sophisticated computing power than may have been available at the time.

Your example strikes a nice balance in that it appears to be a valid TMA solution using methods that could have been employed with the technology available to tracking parties of the WWII era.

If you like, I can post a single MoBo image that will simply prove the concept works. If there is any interest from people wanting to know how I did it, I can follow up later or perhaps on another thread with more detailed MoBo instructions.

[XLjedi]

PowerPoint?

The only vector drawing tools I see in PowerPoint are the same lame autoshape and drawing tools that are available in Excel.

Are you saying you used those tools to create that AoB dial?

[jerm138]

All I'm looking to do now is to find that "magic line" using the conventional in-game tools.

I racked my brain with it all day at work. Knocking down some geometry cobwebs. I'll get it eventually.

Yes, my tutorial, and that AoB calculator were made in Powerpoint. They're the same drawing tools used in most MS Office programs. They're definitely not lame once you learn how to use them... I even used powerpoint when kicking around ideas for my next tattoo! (It's a cover-up for a Batman symbol)... but I think I'm going to scrap this one and go with a more photo-realistic approach.

[XLjedi]

I thought it would take only one pic, until I plotted this first one and it didn't happen to work so perfect.


In this case we see that a 30° course change away from the hydro bearings would produce a fairly decent representation of target heading but the distance in this case remains ambiguous.


Turning 30° into the hydro bearings produces a better result with a little more certainty on distance.


The best results are achieved with more aggressive turns into the hydro bearings that result in lattice-type contact bearing lines. Here we can see the TMA tracking tool lines up best when overlaying the true target track. Of course the only problem with this is that if you maneuver your sub until you get lattice bearings, you will most certainly be left in a poor position assuming the very slow submerged speed of WWII era subs. You may then be stuck doing a surface run end-around.

Do you want me to post one for constant OwnShip heading with increase-decrease speed?

[Pisces]

The technique Jerm described in his document IS actually a two legged manouvre, combining a 'imaginary' relative plot (without range) with a dead-reckoning tracer. First leg is with speed of 0 knots. You do not really need to be stationary, just creeping slow to keep depth is ok too if you are pointed along a bearing line. That messes it up in the least way. From this you get only target true course, and are able to predict the next series of equal-time bearings. (since you were stationary the target-course is also the 'direction of relative movement', DRM in nautical terms) Knowing where the 4th (imaginary) bearing line will point to from the stationary location, you can move to a different vantage-point and create a form of 'stereoscopic view'. You can also go on, skiping the moment of the 4th bearing, taking a bearing at the time of the 5th bearing line, just as long as the track intervals and bearinglines are progressed appropriately. It is perhaps even desireably. You'll want as much as possible a 90 degree angle (a good crossing) between the imaginary line from the stationary location and real bearing line from the displaced location. A small shift of position, or small angle between real and imaginary bearingline makes the range and speed measurement from the plot very unreliable.

If you were actually moving during the first leg you would need to make the same bearing fan drawing (originating from the same center), but translate (move parallel) the line resembling the target course to the head of your speedvector (that originates also from the same center). This translated line is actually the 'direction of relative motion' or DRM on a manouvering board. Put the oldest part of the line at the head of your speedvector. (use the circle tool for this as Jerm suggested in his document, you can drag the arrow anywhere when moving the center) Anything 'in the past' (whatever was before bearing 1) has no meaning in this drawing. The 'future side' (every moment after bearing 1) of the DRM line is actually a collection of all possible endpoints of the target's speedvector. Since you do not know what distance the target travelled during each interval, as range is still unknown, you can also not know yet which point on the DRM line the target speedvector points to. You can however make an estimate on the maximum speed and minimum speed of the target, and show those as speedcircles (like a ring with outer radius 12kts and inner radius 5kts). This should cut-off the DRM line at certain points, indicating the course it would have had at those speeds of the borders of the speedring.

To figure out a definate target course and speed if you moved during the first leg (I mean like cruise speed, 1kts to keep depth isn't that big of a deal), you need another leg, moving in such a way as to create a direction of relative movement that crosses nicely the old DRM. What course and speed is required on the 2nd leg is tricky to figure out beforehand. Use trial-and-error, or when in doubt simply turn parallel to the initial DRM, and move at the estimated/guessed target speed.

This is why I like this method of a stationary 1st leg. I allready get to know it's course. For the second leg I then cruise at the topspeed of the reported speed-class(slow, medium,fast, etc). No need for flank if you're going parallel. You'll probably stay at or more likely gain forward position even with a moderate speed. Once you get to figure out range and speed (from the 4th real bearing) you can start to draw a propper intercept drawing at leisure speed.

[-Danse-]

Played around with this the other night to a satisfying conclusion. For the first three bearings, I kept it simple and remained stationary, and then poured on flank speed to get a good cross-reference point for that predicted fourth bearing.

Because of the distances involved, which were huge, I knew I wasn't going to be exact, but it gave me direction, approximate distance (within half a K as it turned out), and what I thought would be approximate speed, but actually turned out to be bang on.

After setting a timed intercept to arrive early for caution's sake, I let him get closer, repeated the process, confirmed the heading, narrowed down the distance to an accurate level and confirmed his speed.

From there it was shockingly simple to work out AoB from his heading and set up the Hydrophone shot.

Fired two, hit with two, on a small 1600 tonne freighter at a distance of 1500m.

Must say, it's very satisfying to be able to hunt a target and tear its belly out without having to put eyes on target. (I did anyway for enemy/friendly confirmation. Didn't need any "Sorry sir!" moments.)

Found this tremendously useful, so thanks a bunch. While I'm on the subject, thanks Pisces for linking Nefelodamon's tutorial. Downloaded and watched that the other night also. Again, very informative, very useful.

Nearly caused me to sleep in for work because I simply "had" to finish watching it at 1am.

I guess my next step is to try it on the move against a faster target to see if I can track and achieve a good firing position.

Fun fun!

Thanks again guys.

[Hitman]

USS Buffalo (SSN-715) "Buffy Love"
E-Div 1999-2003

Says it all about why you are so good at doing these sound bearing only tracking, uh?

Nice tutorial, many thanks

[jerm138]

@ Aaronblood:
Thanks for taking the time to do all that. It looks like the 30-degree course change has less accuracy because you're introducing a relatively minor change. I like how the 90-degree one turned out

@ Pisces:
Thanks for the extra detail. It's definitely more than I knew about it.

@ Danse:
I'm glad it worked out so well for you. I've never tried a blindfolded shot, but maybe I'll try now.

@ Hitman:
Yes, we avoid fresh air as much as possible.

[Diopos]

Tools, tools, tools ...
Better to have them than not!
I think that all of you interested in this kind of stuff will find the original (and official) US Navy Department Maneuvering Board Manual of 1941 (!!) an interesting read (and adaptable to SH4's map/navigation tools)

download it from here

Solve your tracking problems the way our forefathers did!!!

(check cases VIII to XI)

[jerm138]

Wow... excellent find, Diopos!

I spent a long time searching for something like this online.
I found the US Navy Botswain's Mate manuals at NHSA, but it was all about OwnShip plotting, not how to intercept enemies.

Thanks for the link!

[I'm goin' down]

I highly recommend proceeding to the MoBo form and downloading the MoBo tutoroial and program. They are both masterpieces. Professionally presented, easy to use, well explained, and in short, remarkable.

If I could only master celestial navigaton my life would be complete.

Big wave to aarronblood. (see, somebody did read your tutorial!)

[I'm goin' down]

Cap'n Scurvy's art work (I offered to pay him to build me one)? power point? Take a look at gutted Solution Solver!! (I can go to sleep now!)

Waves to Hitman!

[Diopos]

A more modern version of the Maneuvering Board Manual (1984) can be downloaded from here