[AR] Re: FOGgy idea \ ... vanguarduino?
- From: Henry Spencer <hspencer@xxxxxxxxxxxxx>
- To: Arocket List <arocket@xxxxxxxxxxxxx>
- Date: Tue, 19 Jan 2016 01:39:36 -0500 (EST)
On Tue, 19 Jan 2016, Ben Brockert wrote:
I talked with a guy who had done this with MEMS IMUs at the suborbital
conference and the general take is that you get a signal that is
sqrt(number of sensors) times as good as a single sensor.
Yes, that's the standard result for averaging multiple noisy data sources,
subject to certain assumptions about the details. Which is to say, it's
an expensive way to get big improvements in accuracy, especially since
those crucial assumptions might not be *exactly* true.
Henry
Other related posts: