Imu acceleration to position exclude gravity effect

asked 2016-05-26 08:47:06 -0500

jafecasu gravatar image

updated 2016-05-27 07:52:36 -0500

Airuno2L gravatar image

Im working with a Pixhawk 3DR. And I need to get the position of the robot. I'm trying to calculate with the linear accelerations and a delta time, but I can't exclude the gravity effect.

Im trying with this:

x_posicion_var = (0.5*alx)*(delta_time*delta_time)
y_posicion_var = (0.5*aly)*(delta_time*delta_time)
z_posicion_var = (0.5*alz)*(delta_time*delta_time)

does some know how to resolve it?


edit retag flag offensive close merge delete



This is not an answer, but if you're trying to get position by integrating the acceleration from an IMU twice you're not going to have a good time. Errors/noise are exacerbated when a signal is integrated, and doing it twice leads to huge errors.

Airuno2L gravatar image Airuno2L  ( 2016-05-27 07:51:29 -0500 )edit

Many thanks for your suggestion.

Maybe do you have or know another method to get the position?

jafecasu gravatar image jafecasu  ( 2016-05-27 09:19:23 -0500 )edit

It depends on the robot and what sensors are available.

Airuno2L gravatar image Airuno2L  ( 2016-05-27 11:27:52 -0500 )edit

Im using this device:

And Im working with ROS, so I get the linear accelerations but with the gravity include and I need to exclude it.

jafecasu gravatar image jafecasu  ( 2016-05-27 11:46:36 -0500 )edit

Is this a ground robot or an aerial robot?

Airuno2L gravatar image Airuno2L  ( 2016-05-31 06:28:11 -0500 )edit

I'm using Pixhawk 3D robotics. I get from the device the angular velocity, quaternion and linear acceleration. I'm good with the rotation but not with translation. So.I'm trying to get it from linear acceleration but I can't exclude the gravity. Do you have some suggestions?

jafecasu gravatar image jafecasu  ( 2016-05-31 10:22:24 -0500 )edit