Shop OBEX P1 Docs P2 Docs Learn Events
numerical integrator to get position information from accelerometer — Parallax Forums

numerical integrator to get position information from accelerometer

qiuqiuaaaqiuqiuaaa Posts: 37
edited 2012-02-16 22:00 in General Discussion
I am looking for a numerical integrator to integrate the acceleration getting from accelerometer, but I can not find a hardware like that. can someone give me a guide?

Comments

  • LeonLeon Posts: 7,620
    edited 2012-02-10 02:44
    That sort of thing is usually done in software. What MCU is the accelerometer connected to?
  • qiuqiuaaaqiuqiuaaa Posts: 37
    edited 2012-02-12 18:47
    Thank you. I will focus on software
  • Mark_TMark_T Posts: 1,981
    edited 2012-02-13 09:54
    And be aware that integrating twice leads to lots of drift.... If the accelerometer has analog output an op-amp integrator(s) (with software reset) could be employed.
  • qiuqiuaaaqiuqiuaaa Posts: 37
    edited 2012-02-13 17:43
    Thanks. Unfortunately, I only have digital output. I am also worried about the error.
    Mark_T wrote: »
    And be aware that integrating twice leads to lots of drift.... If the accelerometer has analog output an op-amp integrator(s) (with software reset) could be employed.
  • qiuqiuaaaqiuqiuaaa Posts: 37
    edited 2012-02-13 17:52
    I have a question. What do you mean by analog output, because I think we can only get a lot of numbers. It must be digital, how can it be analog?
    Mark_T wrote: »
    And be aware that integrating twice leads to lots of drift.... If the accelerometer has analog output an op-amp integrator(s) (with software reset) could be employed.
  • ercoerco Posts: 20,259
    edited 2012-02-13 20:26
    Without a gyro, you can't distinguish a tilt from an acceleration; integration of data from a standalone accelerometer is "problematic "at best. More like futile.
  • qiuqiuaaaqiuqiuaaa Posts: 37
    edited 2012-02-14 23:05
    Actually, I have gyro. What I want to do is to integrate the acceleration along each axis after calculating based on accelerometer and gyro.
    erco wrote: »
    Without a gyro, you can't distinguish a tilt from an acceleration; integration of data from a standalone accelerometer is "problematic "at best. More like futile.
  • ercoerco Posts: 20,259
    edited 2012-02-15 15:25
    Sensor fusion & Kalman filter. You have some Googling to do, my friend!
  • JimInCAJimInCA Posts: 80
    edited 2012-02-16 07:10
    qiuqiuaaa wrote: »
    Actually, I have gyro. What I want to do is to integrate the acceleration along each axis after calculating based on accelerometer and gyro.

    Here are a couple of good tutorials to start with.

    http://www.starlino.com/imu_guide.html

    http://www.starlino.com/dcm_tutorial.html

    Jim...
  • qiuqiuaaaqiuqiuaaa Posts: 37
    edited 2012-02-16 22:00
    Thanks a lot!
    JimInCA wrote: »
    Here are a couple of good tutorials to start with.

    http://www.starlino.com/imu_guide.html

    http://www.starlino.com/dcm_tutorial.html

    Jim...
Sign In or Register to comment.