Author Topic: IMU alignment procedure  (Read 772 times)

Offline CorvusCorax

  • Full Member
  • ****
  • Posts: 425
  • Germany
  • Liked: 448
  • Likes Given: 340
IMU alignment procedure
« on: 01/30/2018 03:22 PM »
In some countdowns you hear "vehicle is in self align"

I tried to find information what that actually means.

Common knowledge is that the IMU responsible for vehicle guidance runs through some sort of initialization procedure, but what exactly is initialized?


Is it:

a) Calibrating the IMU itself using the known forces and rotational moments (earth rotation) at the launch location and mount?
b) Finetuning the vehicles orientation and position assuming it is stationary, but with unknowns in the orientation
c) Finetuning the offset of IMU sensors in respect to the vehicle, assuming the vehicle is in known loacation and stationary, as well as a)

Or a mixture of the above?

Or in different words, I see 3 sets of unknowns:

1. The sensor calibration itself
2. The pose (orientation and position) of the vehicle in the universe
3. The pose of the IMU sensors within the vehicle (there might always be manufacturing and mounting discrepancies on the order 10s or 100s  of a degree) (or 180 degree in case of Proton ;) )

Some of which might be done in the lab or during integration, while others apparently happen during this "self align" phase. But which?

Anyone with inside knowledge?

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32331
  • Cape Canaveral Spaceport
  • Liked: 10986
  • Likes Given: 327
Re: IMU alignment procedure
« Reply #1 on: 01/30/2018 04:58 PM »
In some countdowns you hear "vehicle is in self align"

I tried to find information what that actually means.

Common knowledge is that the IMU responsible for vehicle guidance runs through some sort of initialization procedure, but what exactly is initialized?


Is it:

a) Calibrating the IMU itself using the known forces and rotational moments (earth rotation) at the launch location and mount?
b) Finetuning the vehicles orientation and position assuming it is stationary, but with unknowns in the orientation
c) Finetuning the offset of IMU sensors in respect to the vehicle, assuming the vehicle is in known loacation and stationary, as well as a)

Or a mixture of the above?

Or in different words, I see 3 sets of unknowns:

1. The sensor calibration itself
2. The pose (orientation and position) of the vehicle in the universe
3. The pose of the IMU sensors within the vehicle (there might always be manufacturing and mounting discrepancies on the order 10s or 100s  of a degree) (or 180 degree in case of Proton ;) )

Some of which might be done in the lab or during integration, while others apparently happen during this "self align" phase. But which?

Anyone with inside knowledge?

In the past, a light from a theodolite on the ground was used to align the IMU.  But that was with mechanical gyros. 

Now with strap down laser ring gyros, they just have to know that the vehicle is not moving except for earth's rotation and that is what they are self aligning to
« Last Edit: 01/30/2018 05:00 PM by Jim »

Offline CorvusCorax

  • Full Member
  • ****
  • Posts: 425
  • Germany
  • Liked: 448
  • Likes Given: 340
Re: IMU alignment procedure
« Reply #2 on: 01/30/2018 08:56 PM »
In the past, a light from a theodolite on the ground was used to align the IMU.  But that was with mechanical gyros. 

Now with strap down laser ring gyros, they just have to know that the vehicle is not moving except for earth's rotation and that is what they are self aligning to

Thanks, Jim!

I found this papers with additional background information on strap down ring laser gyros:

http://research.ijcaonline.org/volume116/number2/pxc3902354.pdf

Edit:

This paper has additional information an required bias accuracy for different applications as well as technology providing them:

https://www.asee.org/documents/sections/middle-atlantic/fall-2009/01-Evaluation-Of-Ring-Laser-And-Fiber-Optic-Gyroscope-Technology.pdf
« Last Edit: 01/30/2018 09:35 PM by CorvusCorax »

Tags: