Robot Reckoning Question
Dear Lazyweb,
I’m working on an project where we will have 5-10 robots moving around a large field… 100-1,000 yards. Right now we’re planning on using a mix of differential GPS, accelerometers and and a kalman filter so the robots can determine their position to within 1 meter or less. This may or may not be the best way to go about this.
Do you have some ideas as to some other solutions we might go with?
Take a peek at our main webpage…. http://www.orbswarm.com/
Umm…. Kalman Filtering is a little complicated and it has the nasty side effect of compounding the uncertainties in your state position vector (position, velocity, acceleration, etc.). If you are setting up the field yourself, how about a nice simple triangulation. You can have a minimum of two transmitters of different frequencies and as long as you can resolve the angle between your receiver and each of the transmitters, you’ll probably be a little better off (I’m guessing cheaper and more accurate).