SYS_STATUS: fill correct attitude, horizontal position flags

[MaEtUgR]

Solved Problem

In #23001 the correct health components are updated to reflect the system state. However, during testing, I found that this causes the wrong or unexpected MAVLink system status flag to be set.

The AHRS flag should in my understanding indicate IMU and magnetometer health and be unrelated to position estimates. Since there are separate flags for gyroscope, accelerometer, and magnetometer, it seems the AHRS flag was meant to represent an estimator in ArduPilot (see mavlink/mavlink@66c4190). To me, its definition is too unclear to map to any specific EKF2 estimator state.

Solution

1. Map health_component_t::local_position_estimate to MAV_SYS_STATUS_SENSOR_XY_POSITION_CONTROL because with a local position estimate position control is possible
2. Map health_component_t::attitude_estimate to MAV_SYS_STATUS_SENSOR_ATTITUDE_STABILIZATION because with a valid attitude estimate attitude stabilization is possible
3. Don't set the MAV_SYS_STATUS_AHRS flags due to unclear definition.

Changelog Entry

Improvement: fill MAVLink's SYS_STATUS flags for horizontal position and attitude estimate

Test coverage

This was not tested yet. @cmic0 could you provide feedback after you tried?

[bkueng]

To me, its definition is too unclear to map to any specific EKF2 estimator state.

That was also my thinking, so I mapped it to local pos. Certainly not ideal, so removing it works too.

[bresch]

AHRS is by definition an attitude estimate, why not mapping it to health_component_t::attitude_estimate as well?

(source: https://www.vectornav.com/resources/inertial-navigation-primer/theory-of-operation/theory-inertial )

[MaEtUgR]

why not mapping it to health_component_t::attitude_estimate as well?

Not sure if this helps any use case, but it's the definition I'd guess from the name as well.

[MaEtUgR]

ok, I rebased and readded the AHRS to also report about the health_component_t::attitude_estimate