I was reading Dan Simon article about kalman filters http://www.embedded.com/story/OEG20010529S0118?pgno=1 and I decided to give it a go in scilab.I made the following changes in the code:reduced mesnoise from 10 to 0.1 (that makes my measurement sensor more accurate)Set u=0, that is no external accelerationsAnd the results I got are quite dissapointing :-)
According to the chart above in 60 seconds my sensor + kalman filter would report that I am 18 feet away from my original position.how can it be so bad? Is this was I was supposed to get?Thanks!
You need to be a member of diydrones to add comments!
Are you adding any noise to the process? If so, then you are looking at a random walk.
Also, is there a "true" value in the plot above? It looks like there is a "true" value in green and the red estimate follows it quite nicely. Perhaps it is working perfectly?
Replies
Also, is there a "true" value in the plot above? It looks like there is a "true" value in green and the red estimate follows it quite nicely. Perhaps it is working perfectly?
Tom