Ascending Technologies Pelican Accelerometer Modeling

[hanley6]

When collecting data with Ascending Technologies Pelican quadrotor for 4 hours, the following Allan variance and power spectral density is produced:

This result does not reflect the anticipated Allan variance curve presented in the IMU Sensor Model Instructions. The issue then is (1) why does this discrepancy exist? and (2) can the sensor model code be modified to also model this discrepancy?

[hanley6]

Some discussion related to this issue may be found in:

Introduction_to_Sensor_Errors.pdf

starting on page 26 until page 30.