Re: Quaternions vs. Euler Angles
- From: Philip Hahn <hahnp@xxxxxxxxxxxxx>
- Date: Mon, 12 Sep 2005 18:53:50 -0400
Gordan Sikic wrote:
Hello,
How much of an error should I expect to see between the two models?
I have made couple of (real time) simulations and did not have difficulties using quats instead of Euler angles. One was simulation of FAGOT antitank missile, with "fly" time cca 10-12 s, integration period 2ms.
The other was simulation of MiG-21 aircraft, with "fly" time up to 40-50 minutes.
You should be aware that using quaternions, you are no more dealing with ODE's (ordinary differential equations) but DAE's (differential algebric equations). Basically, you use 4 parameters to represent 3 mutually independent variables, meaning that you need to involve algebric constraints( i.e. use the fact that sum of the squares should be equal to unity).
bast regards, Gordan
Right. I renormalize the quaternion every iteration.
I've done hand calculations so I know that I implemented the appropriate equations; I'm pretty sure the error is coming from the different calculations (IE: Quaternions involve only multiplication between the quaternion components and the rotation rates, whereas Euler angles involve trig functions).
So you are saying in these simulations you wrote, your answers were identical to 6 significant figures?
thanks, Phil
.
- Follow-Ups:
- Re: Quaternions vs. Euler Angles
- From: Gordan Sikic
- Re: Quaternions vs. Euler Angles
- References:
- Quaternions vs. Euler Angles
- From: Philip Hahn
- Re: Quaternions vs. Euler Angles
- From: Gordan Sikic
- Quaternions vs. Euler Angles
- Prev by Date: Re: Quaternions vs. Euler Angles
- Next by Date: Re: Quaternions vs. Euler Angles
- Previous by thread: Re: Quaternions vs. Euler Angles
- Next by thread: Re: Quaternions vs. Euler Angles
- Index(es):
Relevant Pages
|