Requester: @ETSells, @SunDevilRocketry/dashboard
(Non-developer requests only): Is your feature request related to a problem? Please describe.
Our current state estimations provide a reliable pitch measurement. The yaw direction experiences drift due to a lack of compensation (it's purely integrated gyro values), and the roll direction is incredibly unreliable. In addition, the pitch angle glitches off by a significant angle every once in a while (root cause unknown).
Describe the solution you'd like
We should re-write the sensor_body_state function to provide a better estimate of these angles and their rates of change. Consider using the magnetometer for compensation in the yaw direction to try and reduce drift over time, and try to re-implement the roll estimation so that its a bit smoother.
This issue may be made obsolete by @dmortloff's work once it is successfully integrated. As a result, we should leave this unscheduled until we have a flight computer with a working USB port and revisit then, since this issue will be blocked until then.
ALSO even if we wait on the kalman filter for orientation, let's do some simple filtering on baro data (thanks Aaryan!)
Describe alternatives you've considered
We've considered holding off for @dmortloff to be ready. This will be revisited when this ticket is ready to be worked.
Additional context
N/A
Points of Contact
Kalman Filter & FC-R2 builder: @dmortloff
Sensor SME: @ETSells
We could also consider looping Aaryan in on this if he starts working with firmware.
Requester: @ETSells, @SunDevilRocketry/dashboard
(Non-developer requests only): Is your feature request related to a problem? Please describe.
Our current state estimations provide a reliable pitch measurement. The yaw direction experiences drift due to a lack of compensation (it's purely integrated gyro values), and the roll direction is incredibly unreliable. In addition, the pitch angle glitches off by a significant angle every once in a while (root cause unknown).
Describe the solution you'd like
We should re-write the
sensor_body_statefunction to provide a better estimate of these angles and their rates of change. Consider using the magnetometer for compensation in the yaw direction to try and reduce drift over time, and try to re-implement the roll estimation so that its a bit smoother.This issue may be made obsolete by @dmortloff's work once it is successfully integrated. As a result, we should leave this unscheduled until we have a flight computer with a working USB port and revisit then, since this issue will be blocked until then.
ALSO even if we wait on the kalman filter for orientation, let's do some simple filtering on baro data (thanks Aaryan!)
Describe alternatives you've considered
We've considered holding off for @dmortloff to be ready. This will be revisited when this ticket is ready to be worked.
Additional context
N/A
Points of Contact
Kalman Filter & FC-R2 builder: @dmortloff
Sensor SME: @ETSells
We could also consider looping Aaryan in on this if he starts working with firmware.