Page 1 of 1

IMU Output Analysis

Posted: Mon May 18, 2015 6:52 pm
by t_walker_21
Hello community,

I am trying to look through the Crazyflie 2.0's firmware and figure out how I can obtain the quad's angular velocity in each axis. I have searched through imu.c,stablizer.c,sensfusion6.c, and particularly in the latter, I found some crazy mathematical gymnastics within the code. Can someone run me through or provide a guide on how the imu algorithm integrates/manipulates the sensor data to obtain roll/pitch/yaw. I also need to obtain the rate of change of roll/pitch/yaw.

Thanks

Re: IMU Output Analysis

Posted: Sat May 23, 2015 3:51 am
by whoenig
It basically uses freely available code to fuse the raw sensor data of accelerometer and gyroscope. The magnetometer is not used for the sensor fusion. The code is available here, alongside with a technical report and paper explaining the details: http://www.x-io.co.uk/open-source-imu-a ... lgorithms/.
The gyro gives you the rate of change directly, however you might need to filter the raw data to avoid outliers.