Navigation coordinations confused:

[bilucent]

I`m really got engrossed in working with an INS/GPS integration.pls help me in the basis,with which coordinate the INS and GPS work usually,and how should I convert them,as an usall wowrk,i`m not familiar with coordinates.
pls help if any information availble,thanx a lot

 

[Chris Hillman]

Suggest some reading

What is your mathematical background? It would be difficult indeed to discuss GPS without substantial mastery of coordinates.

I assume you are studying an engineering textbook, but you'll also need to have a solid grasp of the relevant relativistic physics if you really want to understand how GPS works. A good short introduction is
http://www.arxiv.org/abs/gr-qc/0601110
and a longer one is
http://relativity.livingreviews.org/Articles/lrr-2003-1/index.html

As the ancient Greeks well knew, calendars and coordinates on the sphere are beset by difficulties which modern mathematicians have recognized as topological. In this context, it is interesting to note that physicists have recently begun laying the theoretical foundation for 2nd generation GPS, which unlike the current systems (which are only useful very near the Earth's surface) would be useful for navigation in space, even deep space. In an unheralded milestone in intellectual history, the two ancient problems I just mentioned have finally been satisifactorily resolved by Bartolome Coll:
http://www.arxiv.org/abs/gr-qc/0606044
Here we see another example of the well-known phenomenon in which a seemingly harder problem is sometimes much easier to solve!

 

[bilucent]

Actually this topic(INS/GPS integration)is my BS thesis,and fortunatly I`ve learned about them sufficiently,but my problem is that I have never seen a real one!and I don't know,when for example using GPS,the row data is what?and in which coordination it will give my position,and as the same for INS,I need a to know how are the datas at first,and how should I compare them,I mean in which coordination.
thanx again,I`m looking forward

 

[Chris Hillman]

Bilucent, IMO you're not yet making much sense. Please type more slowly and explain your question as clearly and concisely as possible. I suggest you explain what you mean by the following:

1. "INS/GPS integration"

2. "is my BS thesis" (perhaps you meant to say that your BS honors thesis is entitled---well, you tell me!)

3. "I've learned about them sufficiently" (them? sufficiently? so why are you asking us?)

4. "I have never seen a real one!" (one what?)

5. "when using GPS" (using a handheld GPS device to determine your geolocation?)

6. "row data"

7. "in which coordination it will give my position"

8. "I need a to know how are the datas at first" (which data? whaddya mean, "at first"? initial data?)

For example, you might be confusing various mathematical models of the geoid with various coordinate systems.

 

[bilucent]

OK,sorry for being hasty!my main problem,is,when I`m working with GPS,in which coordinate it shows my position,for example (X,Y,Z) or (north,east,down) or attitudly.
and as the same for any inertial navigation system,
because I need to combine the two gained positions,for having a much more accurate positioning.

 

[D H]

The coordinates used depends on your GPS receiver and the INS. The GPS devices sold to the public report Earth-fixed longitude, geodetic latitude, and altitude. The relevant coordinate systems are WGS84 coordinate system and the WGS84 reference geoid. INS typically use an inertial reference frame internally, most likely J2000, and translate this to Earth-fixed. To combine the states you will need to know more than just a little bit about Kalman filters.

Stop being so hasty. Take a little time to type full sentences with correct spelling. Doing so is much appreciated here.

 

[Chris Hillman]

You don't seem to realize that you need to clarify your question. I suggested that you clarify eight points.

As DH implies, unfortunately the engineering literature (and manuals for common GPS devices) tend to confuse geoid models and coordinate systems. These are not the same thing!

The references I cited do discuss "earth-centered" vs. "inertial" coordinates, but I am still puzzled by what you said about not knowing anything about coordinates, yet you say you are writing a BS thesis about... what?... and have a satisfactory knowledge of... what? Do you see why I am puzzled?

 

[D H]

The WGS84 coordinate system refers to a rotating cartesian reference frame, not the geoid. The geoid is a simplified representation of the Earth's surface, something suitable for embedding so that altitude can be easily calculated.

 

[Chris Hillman]

Where "Earth centered" is most likely a type of "rotating coordinate system" (since the Earth is rotating). I am baffled how to explain "geoid" if the OP doesn't know about spherical harmonics. (And how could he know about them if he doesn't know about coordinates, e.g. polar spherical coordinates on the round sphere?)

 

[bilucent]

sorry I missed the 8 questions:

1&2. "INS/GPS integration means to combine the results of both GPS reciever and INS outage using an optimization tool usually kalman filter for positioning on a 3d moving vehicle,more likly a simple car.Its the name of my BS thesis.

3. "I have studied couple of ebooks and refrences about GPS,INS,KALMAN filtering,the integration of GPS and INS,I have passed units of algebra and improved algebra,meaning I have enough knowledge about the matrixes,coordinates and transformation matrixes.(I think they are enough,if I've missed a necessary learning pls tell me).

4. "I have never seen a real GPS reciever and IMU(main computing part of an INS)
so I have no preview of working with them,unfortunately I don't have the chance to see one of them.

5. "yes,a hand held GPS device,as you hinted up.

6. "row data are the first data gathered by GPS reciever,which probably after some calculation gives us the pleasure outage.

7. "in which coordination" I meant before computations of GPS reciever on data,what form do the datas have,and after the computations,for example GPS says that your position as (X,Y,Z) or ...or I can choose my ideal coordinate from possible ones on the reciever?

8. "Datas at first,I mean for INS,do I get the position in the form of (roll,pitch,yaw)?and in GPS for example I get in (X,Y,Z)?and how should I combine them,I mean in which coordination?what sort of transformation is needed?

thanks for your patience

 

[D H]

An IMU typically measures state derivatives: acceleration and angular rate. You have to integrate these to propagate the state (position, attitude). This creates a problem: You are integrating noise and bias. Integrating noise results in a random walk. Integrating bias results in a random ramp. The computed state will drift from the true state. To counter this, you need some other measurement such as that from a GPS receiver to update the state.

I recommend you read Brown and Hwang, "Introduction to Random Signals and Applied Kalman Filtering" and Gelb, "Applied Optimal Estimation". Your library almost certainly has these books.

 

[bilucent]

thanks a lot,I try for that:)

 

[Chris Hillman]

Hi, bilucent,

I guess your answers are moot since I think DH has told you what you needed to know, but thanks anyway! For future, this kind of context is invaluable when asking questions.

Regarding what you said about never having seen a GPS unit, these are ubiquitous in the U.S., many chains of stores carry several popular brands (typical units cost perhaps U.S. $130 on sale), a GPS unit is required in every taxicab, and so on. I confess that I had previously assumed that this is true pretty much anywhere in the developed world. Does anyone know otherwise?

 

[bilucent]

thanx mr Hillman for your notice.
but I thought I`ve improved my questioning.
and I want to let u know that here these kind of devices exist,but not as much as where u live,
and the last interface is not valuable for me,just a screen,I wanted too know more about its internal functioning.
anyway,goodluck