Range prediction, fast fire solution.

[NavyUSA]

Hi commanders,
With the below image, i show you one of a few tables for prediction range during first phase of tma.
http://oi60.tinypic.com/kar71l.jpg
This posted image is usable only if relative bearing of target is 75<=B<=85, and ownship and target are along lag los.
For obtain prediction range must be calculated bearing rate after 2 min from first relative bearing B.
Example: if bearing rate is 1.7, you have the 68.23% probability that target is on interval [5.5 , 6.5] nm. But the probability that target is on interval [6 , 8] nm is 95.4% (Gaussian).
This is a great information (statistic) and no a magic solution.
For calculate bearing rate with good accuracy i have find simple method in a manual of tma from betasom forum, in italian language. I haven't had problem understand with google translate.
Last point important, this tables must be used with ownship speed between 6,7 knots, and when target speed is between 5,7 knots. In all other situations the error of prediction will increase.

[Pisces]

Can you explain the rational behind those standard deviation numbers? What sort of probability is that? Also how do you get from probability percentages to standard deviations and back in the first place. (I like math, but I suck at statistics, so it's still magic to me)

While my Italian is even worse than my 'statistic', it would have been nice if you could have provided a link to that forum. Maybe it is public enough to get a peek inside on that method. And let Google Translator have a go at it. Could be fun. :)

Still, only with relative target bearing between 75 and 85, and speeds between 6 and 7 , or target needing to be between 5 and 7, seems to be quite limiting in use to me. Such narrow bounds.

[NavyUSA]

Can you explain the rational behind those standard deviation numbers? What sort of probability is that? Also how do you get from probability percentages to standard deviations and back in the first place. (I like math, but I suck at statistics, so it's still magic to me)

Speak about the statistical theory in this forum is off-topic. In literature you can find the whole theory.:)

You know that the bearing rate has a very good correlation with the distance of the contact. It's inversely proportional to the range. These tables are born, from a statistical analysis of Bearing rate variation, by varying the range, route target and relative bearing.


While my Italian is even worse than my 'statistic', it would have been nice if you could have provided a link to that forum. Maybe it is public enough to get a peek inside on that method. And let Google Translator have a go at it. Could be fun. :)

If Cristian from Betasom, wants personally explain in English this method, would be better. But I think it's very clear. The link below:
http://www.betasom.it/forum/index.php?showtopic=33792
The direct link is this: http://goo.gl/S7iAUy
Page 24.


Still, only with relative target bearing between 75 and 85, and speeds between 6 and 7 , or target needing to be between 5 and 7, seems to be quite limiting in use to me. Such narrow bounds.

For all relative target bearings between 20 and 160 there is a table. I can send you all tables on your e-mail address.
About it, I think that speeds between 5 and 7 knots are not very restrictive, for several reasons.

[Pisces]

Speak about the statistical theory in this forum is off-topic. In literature you can find the whole theory.:)Well, I didn't mean to ask about what standard deviations are in general. Just in this context, in that table. Standard deviations are the summed squares of differences to a mean value, divided by the population size. But what sort of values and means is that table based on? 'Barnacle size' on the sperical array?

You know that the bearing rate has a very good correlation with the distance of the contact. It's inversely proportional to the range. These tables are born, from a statistical analysis of Bearing rate variation, by varying the range, route target and relative bearing.



If Cristian from Betasom, wants personally explain in English this method, would be better. But I think it's very clear. The link below:
http://www.betasom.it/forum/index.php?showtopic=33792
The direct link is this: http://goo.gl/S7iAUy
Page 24.Fair enough. I'll see how far I'll get this this document. Looks juicy. Too bad it isn't already in english.

For all relative target bearings between 20 and 160 there is a table. I can send you all tables on your e-mail address.
About it, I think that speeds between 5 and 7 knots are not very restrictive, for several reasons.Would you care to elaborate on that? Are you expecting all opponents to move at speeds between 5 and 7 knots? Sure, submerged opponents that know that you are there would want to stay silent, and thus slow. But that doesn't account for all,or most situations.

Anyway, I think I'll have pizza tonight and set my teeth in this document. Thanks :03:

[NavyUSA]

Well, I didn't mean to ask about what standard deviations are in general. Just in this context, in that table. Standard deviations are the summed squares of differences to a mean value, divided by the population size. But what sort of values and means is that table based on? 'Barnacle size' on the sperical array?


The population is made by misure of the bearing rate for different target course, range and relative bearing, a large number of measurements.
The standard deviations in that tables shows how much is variation or dispersion of the bearing rate from the average value, for the particular situation fixed by relative bearing, type of los and range.


Would you care to elaborate on that? Are you expecting all opponents to move at speeds between 5 and 7 knots? Sure, submerged opponents that know that you are there would want to stay silent, and thus slow. But that doesn't account for all,or most situations.


Good observation. During a research phase, the submerged opponents probably want to stay silent at speeds between 5 and 7 knots, best search speed is 7.
If a player is moving at speeds less than 5 knots, means that its towed array will be curved in depth, this will not give him a good solution, and the transducers can be below the layer. I think that an experienced player will never move at speeds less to 4 knots in research or attack phase.
Still, if the target has a velocity greater than 8 knots, is very likely that you will hear him with two sensors or you can get information from the demon. In these cases it's easy to proceed with the usual methods, Triangulation, Strip plotting ecc..


I have answered to your doubt?



Thank you for your interest :03:

[NavyUSA]

Pisces, what methods you use to calculate the Br?

I to calculate a history of Br use the method described by Cristian, but if I want to calculate a single value of Br I determine the angle between two lines of bearing directly from the screen of the Tma station (trigonometric).

[Pisces]

I don't in SC or DW really. Back in the beginning I made bearing history charts to plot it. But lost interest in this arduous task. Also, I didn't really know how to use this plot effectively.

Then some years ago I figured out I could used the TMA ruler to estimate the fraction of a bearing to a 10th degree by setting the course of the TMA ruler to the degree before and after the bearing line, and then measuring/eyeballing the ruler-tip distance to the bearing line. This provides much better accuracy than the 1 degree resolution of the bearing readouts. But is still quite arduous, and not very useful in tense combat situations. It takes too much time for me to be comfortable.

So, I still just fumble around trying to get the ruler to lineup with the bearings after enough zig-zags. Just like everyone else, with only a faint idea of what I'm supposed to do with the TMA dotstack. I still let the Auto-TMA do most of the work also. In other words, I suck at it. :)

But this method on that Italian forum (which I still have not found the time to decypher) does make me curious what you can do with it.

In SH3 and SH4 though, I rely much more on the graphical methods of determining course and speed based on bearing alone. (search subsim for 3 bearing and 4 bearing method) And my self-created 3-bearing AOB finder sliderule disk (http://www.subsim.com/radioroom/showthread.php?t=147719). Unfortunately the drawing tools in SC and DW leave much to be desired to be able to do this effectively in the game. (no ability to plot lines on the map, only during mission editing)

[NavyUSA]

I don't in SC or DW really. Back in the beginning I made bearing history charts to plot it. But lost interest in this arduous task. Also, I didn't really know how to use this plot effectively.

Then some years ago I figured out I could used the TMA ruler to estimate the fraction of a bearing to a 10th degree by setting the course of the TMA ruler to the degree before and after the bearing line, and then measuring/eyeballing the ruler-tip distance to the bearing line. This provides much better accuracy than the 1 degree resolution of the bearing readouts. But is still quite arduous, and not very useful in tense combat situations. It takes too much time for me to be comfortable.


It is easier to calculate the angle between two bearing (in LAG los)with the pocket calculator :D:
angle = arctan((So*2/60)/(length segment bearing from ownship and the intersection point))
Br= angle/2 [DEG/min] ''always approximated for defect''


So, I still just fumble around trying to get the ruler to lineup with the bearings after enough zig-zags. Just like everyone else, with only a faint idea of what I'm supposed to do with the TMA dotstack. I still let the Auto-TMA do most of the work also. In other words, I suck at it. :)


Zig-zags method is useful when target is very far. Allen Schrader in his video show a TMA solution of one target (civil) at range 35 nm. In this situation there is insufficient relative motion between the two platforms and is necessary to move along three/four legs (LAG and LEAD) with 4/5 bearings for leg, exactly Zig-zags.
So, this is a unusual scenery. Method inefficient for target range less then 11 nm (probable scenery).


In SH3 and SH4 though, I rely much more on the graphical methods of determining course and speed based on bearing alone. (search subsim for 3 bearing and 4 bearing method) And my self-created 3-bearing AOB finder sliderule disk (http://www.subsim.com/radioroom/showthread.php?t=147719). Unfortunately the drawing tools in SC and DW leave much to be desired to be able to do this effectively in the game. (no ability to plot lines on the map, only during mission editing)

Plotting Procedures in Dangerous waters aren't necessary, but certainly in the Second World War :)

[Pisces]

It is easier to calculate the angle between two bearing (in LAG los)with the pocket calculator :D:
angle = arctan((So*2/60)/(length segment bearing from ownship and the intersection point))
Br= angle/2 [DEG/min] ''always approximated for defect'' Sure, a pocket calculator does make it easier. I didn't find that much math though in that document that you suggested. Or it is hidden in the Italian text. Do you have any other suggestions where I can find how this formula is used and explained? Feel free to make us a tutorial. ;)


Zig-zags method is useful when target is very far. Allen Schrader in his video show a TMA solution of one target (civil) at range 35 nm. In this situation there is insufficient relative motion between the two platforms and is necessary to move along three/four legs (LAG and LEAD) with 4/5 bearings for leg, exactly Zig-zags.
So, this is a unusual scenery. Method inefficient for target range less then 11 nm (probable scenery).Well, speed changes only also helps to change the bearing rate. But not as much as a lag-los. I can't comprehend this method that you suggest in the first message, so I can't really say anyting about it's preference for short range TMA.


Plotting Procedures in Dangerous waters aren't necessary, but certainly in the Second World War :)Until you frequently patrol close to the bottom and near to the shores. Then you'll like to do a bit more depth contour-plotting by yourself. It's a shame it isn't possible during play really, as the code is obviously built into the engine. If I can plot circles and crosses, why not lines? Time comparison between modern and WW2 eras is just no excuse. I'm sure real life operators would have this function.

I know there is MoBo, the electronic manoeuvring-board application. But it is external to the game.

[NavyUSA]

Sure, a pocket calculator does make it easier. I didn't find that much math though in that document that you suggested. Or it is hidden in the Italian text. Do you have any other suggestions where I can find how this formula is used and explained? Feel free to make us a tutorial. ;)


The formula is not mentioned in the manual, also the table of the first post is not in the manual. I mentioned the manual (by cristian) just to recommend a precise method for Br calculation.

About formula: (simple trigonometry)
With the law of sines you can calculate the angle § = arcsin( B/L*sin(w)) , w lower angle between B and bearing at time 0.
B/L << 1 therefore § is approximately equal to §* = arctan(B/L) OR §* = arcsin(B/L) [error=<10^-2] ''more easier to calculate''. In the example B is SpeedOwnship/60*2. For obtain L is sufficient bring the cursor on intersection of the bearings when the bearing appear on TMA display.
http://oi61.tinypic.com/mj5ts7.jpg

[Pisces]

Oh ok, I remember having seen that drawing. Yep, in chapter 10 of some US Navy tracking training guide (Operations Specialist, Volume 1 NAVEDTRA 14308).

Not sure anymore where or how I found it. I think I can share it to all that are interested, as the frontpage says " Approved for public release; distribution is unlimited."

I guess I'll have to read that again.

[NavyUSA]

Oh ok, I remember having seen that drawing. Yep, in chapter 10 of some US Navy tracking training guide (Operations Specialist, Volume 1 NAVEDTRA 14308).

Not sure anymore where or how I found it. I think I can share it to all that are interested, as the frontpage says " Approved for public release; distribution is unlimited."

I guess I'll have to read that again.

Yes, the image is taken from that document.
Do you understand how to calculate simply the Br?
Just a clarification: the cited formula is fine for a single calculation of the Br, for a Br historian use the cristian's method plotting on navmap.

Now you can use the ekelund range or tables for range predictions, understand if the target range is decreasing or increasing, apply the Spiess graphical method, ecc.
The bearing rate is essential for an efficient TMA. :ping:

[Pisces]

Yes, using that picture your earlier posts start to make sense. Also the requirements for near 90 degree relative bearing. But if you use a calculator then you might as wel correct for different relative bearings.


...
Just a clarification: the cited formula is fine for a single calculation of the Br, for a Br historian use the cristian's method plotting on navmap. I don't think you meant this one:

http://en.wikipedia.org/wiki/Cristian's_algorithm

[EDIT]Nevermind, different Cristian. You meant from the Italian forum.

[NavyUSA]

Yes, using that picture your earlier posts start to make sense. Also the requirements for near 90 degree relative bearing. But if you use a calculator then you might as wel correct for different relative bearings.


I don't think you meant this one:

http://en.wikipedia.org/wiki/Cristian's_algorithm

[EDIT]Nevermind, different Cristian. You meant from the Italian forum.


Are you kidding me? :o :)

[Pisces]

I had a brainfart for a moment. I didn't understand what Cristian method you were talking about, until I checked your earlier messages.

[NavyUSA]

For completeness, I list the procedural sequence that should be followed with the tables: (repeat for those who are interested, I can send you by email all tables for all relative bearings ad all LOS type)

Scenario: bearing-only TMA, track target with TA, ownship and target(submerged) in search phase.

i) Measure the Br at time 2 and 4. (target range is decreasing or increasing?)
ii) Recognize LOS type. (LAG, LEAD)
iii) Whit the Br measured at time 4 and relative target bearing measured at time 2 --> choose the corresponding table. (RelativeB+LOS+Br --> Table)
iv) Obtained the probable interval target range, seek a possible alignment of the points in the dotstack in the interval range. (Strip plotting)
v) If the probability is friendly, the target will not be far from the solution found:up:. Otherwise you will fall into the 5% probability that the interval be wrong:down:.

[Pisces]

Sure, send them. Expect a pm.

[NavyUSA]

In SH3 and SH4 though, I rely much more on the graphical methods of determining course and speed based on bearing alone. (search subsim for 3 bearing and 4 bearing method) And my self-created 3-bearing AOB finder sliderule disk (http://www.subsim.com/radioroom/showthread.php?t=147719). Unfortunately the drawing tools in SC and DW leave much to be desired to be able to do this effectively in the game. (no ability to plot lines on the map, only during mission editing)

Please note that "3 bearing and 4 bearing methods" are not usable, not because you can't draw lines but for other obvious reasons.

[Seppel]

"About formula: (simple trigonometry)
With the law of sines you can calculate the angle § = arcsin( B/L*sin(w)) , w lower angle between B and bearing at time 0.
B/L << 1 therefore § is approximately equal to §* = arctan(B/L) OR §* = arcsin(B/L) [error=<10^-2] ''more easier to calculate''. In the example B is SpeedOwnship/60*2. For obtain L is sufficient bring the cursor on intersection of the bearings when the bearing appear on TMA display."

Can someone explain how i get the angle w?
Is it possibel to get the tabels for relative bearings and los typs?
Regards

[THEBERBSTER]

A Warm Welcome To The Subsim Community > Seppel
Subsim <> Make A Donation <> See The Benefits <> Support The Community

[FPSchazly]

"About formula: (simple trigonometry)
With the law of sines you can calculate the angle § = arcsin( B/L*sin(w)) , w lower angle between B and bearing at time 0.
B/L << 1 therefore § is approximately equal to §* = arctan(B/L) OR §* = arcsin(B/L) [error=<10^-2] ''more easier to calculate''. In the example B is SpeedOwnship/60*2. For obtain L is sufficient bring the cursor on intersection of the bearings when the bearing appear on TMA display."

Can someone explain how i get the angle w?
Is it possibel to get the tabels for relative bearings and los typs?
Regards

The part in bold is important (B/L is much less than 1). It's the small angle approximation, so you don't need "w".

What is this formula in reference to? I just skimmed this thread briefly and didn't see anything.

[Seppel]

If I understand correctly, the formula should determine a more accurate bearing rate. The formula is taken from an Italian tma manual (see link) I am looking for a decent method to get a better beraing rate, but so far without success. all self-determined beraing rates, for example, with the stopwatch are firstly inaccurate and secondly not suitable for a combat situation. other methods such as MoBo, in my opinion, ****. The general problem is the indication of opponent bearing. in all subsimgames I get, this is given as a whole number and not as a number with decimals.
We need a method with which we could determine the bearing rate relatively quickly after the first bearing. Maybe this formula is suitable? but I am still in the process of understanding them.
Best regards

[Aktungbby]

Seppel! :Kaleun_Salute:

[Seppel]

<s> mate