Dears, I was reading some posts about using Kalman filter with accelerometer sensor, and most of the questions were about the velocity and position. but my question is can i use Kalamn filter to filter the acceleration raw data itself ? for example my sensor has 3 axes and I want to remove the noise form them so can do that ? in this case the state vector should be (x,y,z) for three axes and the measurements z also would be x,y,z raw acceleration data that were read form the accelerometer. it is right what I am saying ? and what about the rest of the matrices how should they would be ? their values ... and the process noise, measurement noise ? for the measurement noise I think it should be taken form the sensor data sheet for the three axes ..... waiting for your suggestions and comments ... thank you very much _____________________________ Posted through www.DSPRelated.com
accelerometer and Kalman Filter.
Started by ●September 13, 2013
Reply by ●September 13, 20132013-09-13
On Fri, 13 Sep 2013 07:23:02 -0500, samertelawi wrote:> Dears, > I was reading some posts about using Kalman filter with accelerometer > sensor, and most of the questions were about the velocity and position. > but my question is can i use Kalamn filter to filter the acceleration > raw data itself ? > for example my sensor has 3 axes and I want to remove the noise form > them so can do that ? > in this case the state vector should be (x,y,z) for three axes and the > measurements z also would be x,y,z raw acceleration data that were read > form the accelerometer. > it is right what I am saying ? > and what about the rest of the matrices how should they would be ? > their values ... and the process noise, measurement noise ? > for the measurement noise I think it should be taken form the sensor > data sheet for the three axes ..... > waiting for your suggestions and comments ... > thank you very muchIf you just have acceleration in x, y and z and you want to filter out noise, then you don't want a Kalman filter -- you want low-pass filters on the three accelerometer channels. The Kalman filter gets useful in navigation problems when you have redundant data and no good intuitive way of figuring out how to get good information from the data. Then the Kalman filter (which usually ends up being an extended Kalman or worse) provides a framework for developing something that'll give you the answers that you need. -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
Reply by ●September 13, 20132013-09-13
On 9/13/2013 7:23 AM, samertelawi wrote:> Dears, > I was reading some posts about using Kalman filter with accelerometer > sensor, and most of the questions were about the velocity and position. > but my question is can i use Kalamn filter to filter the acceleration raw > data itself ?If you have dynamic model of a vehicle that accelerometer is attached to, and if accelerations along different axis are dependent in some way, then Kalman filter could be useful.> for example my sensor has 3 axes and I want to remove the noise form them > so can do that ?Idiot.> in this case the state vector should be (x,y,z) for three axes and the > measurements z also would be x,y,z raw acceleration data that were read > form the accelerometer. > it is right what I am saying ? > and what about the rest of the matrices how should they would be ? > their values ... and the process noise, measurement noise ? > for the measurement noise I think it should be taken form the sensor data > sheet for the three axes ..... > waiting for your suggestions and comments ... > thank you very much > > > > _____________________________ > Posted through www.DSPRelated.com >
Reply by ●September 14, 20132013-09-14
On Saturday, September 14, 2013 9:34:41 AM UTC+12, Vladimir Vassilevsky wrote:> On 9/13/2013 7:23 AM, samertelawi wrote: > > > Dears, > > > I was reading some posts about using Kalman filter with accelerometer > > > sensor, and most of the questions were about the velocity and position. > > > but my question is can i use Kalamn filter to filter the acceleration raw > > > data itself ? > > > > If you have dynamic model of a vehicle that accelerometer is attached > > to, and if accelerations along different axis are dependent in some way, > > then Kalman filter could be useful. > > > > > for example my sensor has 3 axes and I want to remove the noise form them > > > so can do that ? > > > > Idiot. > > > > > in this case the state vector should be (x,y,z) for three axes and the > > > measurements z also would be x,y,z raw acceleration data that were read > > > form the accelerometer. > > > it is right what I am saying ? > > > and what about the rest of the matrices how should they would be ? > > > their values ... and the process noise, measurement noise ? > > > for the measurement noise I think it should be taken form the sensor data > > > sheet for the three axes ..... > > > waiting for your suggestions and comments ... > > > thank you very much > > > > > > > > > > > > _____________________________ > > > Posted through www.DSPRelated.com > > >but he'll need to do a MATLAB simulation first to see.
