Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Inertial Navigation Integration FIlter

Inertial Navigation Integration FIlter

Inertial Navigation Integration FIlter

I am trying to create a filter that can integrate inertial measurements with GPS measurements. My inertial sensors are a yaw axis gyro, an odometer, and possibly a 3axis accelerometer.
Could you please help me with some guidance? I have read papers, articles and books on the subject but I am still a little confused.
I consider the acceleration constant, so I have a state vector x=[E N v_E v_N a_E a_N], E and N are the positions in NED coordinate system, v_E and v_N are the speed on each axis, and a is the acceleration on each axis
The measurement vector z is [E_GPS N_GPS E_DR N_DR a_E a_N] where E_GPS and N_GPS are the coordinates obtained from the GPS receiver and E_DR and N_DR are the coordinates obtained from the inertial sensors through dead reckoning, and a is the acceleration obtained from the accelerometer.
F, the transformation matrix is: [1 0 dt 0 (dt^2)/2 0; 0 1 0 dt 0 (dt^2)/2; 0 0 1 0 dt 0; 0 0 0 1 0 dt; 0 0 0 0 0 1];
H, the measurement matrix is: [1 0 0 0 0 0; 0 1 0 0 0 0 ; 1 0 0 0 0 0; 0 1 0 0 0 0; 0 0 0 0 1 0; 0 0 0 0 0 1],
Does it make any sense so far? Should I chose a different state vector?
My problem is that I don't know, for this system how can I determine the process noise covariance and measurement noise covariance.
Can you give me some indications in that direction?
Thank you very much.

RE: Inertial Navigation Integration FIlter

So I should introduce an estimate in the equations the measurement and process noise? Like z=Hx+w (where w is the measurement noise) and x(+)=Fx(-)+v (where v is the process noise)?

RE: Inertial Navigation Integration FIlter

It's not for school. It's a project i'm trying to do im my spare time. So I'm far from the topic? Can you give me some guidance please, so I can clear some of the confusion?
Im trying to integrate the output from an IMU(3 axis gyro, yaw axis accelerometer and odometer if needed) with GPS.
Where do I go wrong in the system described before?

RE: Inertial Navigation Integration FIlter

The process noise is added to the state update equation:
x(+)=x(-)F+w (w process noise)
and the measurement noise is added to the measurement equation:
z=Hx+v (v measurement noise)
My question is how to determine the values of v and w?
You determine the values experimentally, analytically or you just assume some value by trial and error?

RE: Inertial Navigation Integration FIlter

Thx for the help. :)
For a Inertial system with a gyro and an odometer, should I introduce more states then [position velocity acceleration]?
Should I introduce in the state vector also the heading, gyro and odometer bias? and i get a state vector like [position velocity acc heading gyro_bias odo_bias] ?
Or in case I don't introduce the bias from the odometer and gyro, they will be part of the measurement noise, and they only downside is that I cannot correct the sensors?

RE: Inertial Navigation Integration FIlter

Should also the heading be part of the state vector? How can I determine what is vital to be in the state vector and what can be skipped?
Thank you

RE: Inertial Navigation Integration FIlter

No, it's not for school. I graduated 2 years ago from university.
Why is everybody asking me if it is for school? It shouldn't be?

RE: Inertial Navigation Integration FIlter

Sorry, I forgot I already asked.  The fact that you don't know what to do with heading raised a red flag.

Student postings are not allowed on this site.  This site is supposed to be for engineering professionals asking work related questions.  Some leeway is allowed for cross-discipline work.


FAQ731-376: Eng-Tips.com Forum Policies
Chinese prisoner wins Nobel Peace Prize

RE: Inertial Navigation Integration FIlter

So you don't have to worry about this..I haven't been a student for some time...
I'm asking if the heading needs to be part of the state vector as I saw different approaches that did not use it.
And my experience with Kalman filters doesn't go too far. I have read about it but it's the first time I'm trying to implement sth like this.
So then a better state vector would be: [position speed acceleration heading gyro_bias odometer_byas]
or should i leave out the acceleration?

RE: Inertial Navigation Integration FIlter

If you have no other attitude information, then why do you have heading and gyro_bias?  Seems like you picked a bigger problem that necessary for a first project.

A relatively complete set might be:
3 positions
3 velocities
3 accelerations
3 angle
3 angle rates
3 angle accelerations
same with biases

36 states is not unheard-of


FAQ731-376: Eng-Tips.com Forum Policies
Chinese prisoner wins Nobel Peace Prize

RE: Inertial Navigation Integration FIlter

Quote (IRstuff):

36 states is not unheard-of
And how...!

If you're adding in GPS, you'll want:
clock bias
clock drift
wheel speed scale factor (I'm assuming this is vehicle-based)
compass offset

Add all of that into what IRstuff listed and you have a pretty significant-sized matrix.  Luckily, a large percentage of the coefficients are zero (with the majority of non-zero coefficients along the diagonal), so the calculations are faster than one would expect from a typical 30-coefficient+ matrix.

Dan - Owner

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members! Already a Member? Login


Design for Additive Manufacturing (DfAM)
Examine how the principles of DfAM upend many of the long-standing rules around manufacturability - allowing engineers and designers to place a part’s function at the center of their design considerations. Download Now
Taking Control of Engineering Documents
This ebook covers tips for creating and managing workflows, security best practices and protection of intellectual property, Cloud vs. on-premise software solutions, CAD file management, compliance, and more. Download Now

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close