Peter Barker
e1a033b296
AP_NavEKF3: make logging a core concern
...
Also dissolve some methods only used for logging
4 years ago
Andrew Tridgell
1e4b1d7563
AP_NavEKF3: convert to using common buffer classes
...
this saves a considerable amount of flash
4 years ago
Randy Mackay
5eb3875ebb
AP_NavEKF3: rename posxy_source_last and posxy_source_reset
4 years ago
Randy Mackay
9716b09cad
AP_NavEKF3: minor indenting fix
4 years ago
Randy Mackay
6c0bb6a198
AP_NavEKF3: add convert_params for source
4 years ago
Randy Mackay
ba9f60abc7
AP_NavEKF3: add getVelInnovationsAndVariancesForSource
4 years ago
Randy Mackay
58aa6086df
AP_NavEKF3: always calculate GPS innovations and variances
4 years ago
Randy Mackay
e7e91b1c3b
AP_NavEKF3: always calculate extNav innovations and variances
4 years ago
Randy Mackay
6b2b5c4ca0
AP_NavEKF3: add CalculateVelInnovationsAndVariances
4 years ago
Randy Mackay
4cadaa5194
AP_NavEKF3: readyToUseOptFlow, Beacon, ExtNav check source
4 years ago
Randy Mackay
faed58a027
AP_NavEKF3: integrate Source for yaw
4 years ago
Randy Mackay
8931e50166
AP_NavEKF3: integrate Source for alt
4 years ago
Randy Mackay
c21d58ebea
AP_NavEKF3: integrate Source for position
4 years ago
Andrew Tridgell
f5275fd0ad
AP_NavEKF3: fixed use of pointers in ringbuffers
...
these don't work with AP_DAL
4 years ago
Andrew Tridgell
9b81c5a1e0
AP_NavEKF3: use dal reference in EKF backends
...
saves a bit of flash space
4 years ago
Andrew Tridgell
096aab9388
AP_NavEKF3: convert to use AP_DAL for new replay structure
...
Co-authored-by: Peter Barker <pbarker@barker.dropbear.id.au>
4 years ago
Paul Riseborough
6c7820dd8d
AP_NavEKF3: Add verification of tiltErrorVariance() method in SITL
4 years ago
Paul Riseborough
16ae75a681
AP_NavEKF3: Consolidate and log tilt error variance calculation
4 years ago
Paul Riseborough
4e41e9f5f1
AP_NavEKF3: Ensure consistent use of 312 or 321 rotation order
...
Prevents possibility of a differnt sequence being used to calculate the yaw and application of the yaw to the quaternions
4 years ago
Paul Riseborough
73d5ca5ad3
AP_NavEKF3: Fix yaw drift after yaw reset at +-90 deg pitch
...
Use existing covariance prediction code to set quaternion state covariances.
Assumes tilt error is 3 deg 1-sigma.
TODO derive and add function that calculates tilt error variance.
4 years ago
Paul Riseborough
a97d030e4b
AP_NavEKF3: Reset earth field variances when starting mag field estimation
4 years ago
Randy Mackay
c3fd82b6d7
AP_NavEKF3: minor spelling fix
4 years ago
Randy Mackay
2f1ae29a9c
AP_NavEKF3: extnav data gets corrected boolean
...
CorrectExtNavForSensorOffset and CorrectExtNavVelForSensorOffset use
corrected boolean to avoid correcting twice
4 years ago
Randy Mackay
71487e9c10
AP_NavEKF3: gps_elements gets corrected flag
...
CorrectGPSForAntennaOffset uses this flag to ensure position and velocity are only corrected once
4 years ago
Peter Barker
0496ecc4f6
AP_NavEKF2: remove unused state variables
4 years ago
Peter Barker
ef505df78a
AP_NavEKF3: remove unused quat2Tbn method
4 years ago
Peter Barker
0eafe4b662
AP_NavEKF3: remove dangling method declarations
4 years ago
Peter Barker
3dc67f09fd
AP_NavEKF3: remove unused innovationIncrement and lastInnovation
4 years ago
Andrew Tridgell
8a4b0f858a
AP_NavEKF3: reset body mag variances at key points
...
we need to reset the body mag variances if we change sensors or if we
are starting 3D fusion. When not doing 3D fusion we zero the
variances, so they must be initialised again when we restart
fusion. This fixes a bug in handling the variances on a 2nd flight
4 years ago
Andrew Tridgell
e9185f526b
AP_NavEKF3: add comments on state indexes
...
make it easier to check indexes
4 years ago
Andrew Tridgell
2bc7d17e7a
AP_NavEKF3: improved core comparison to check alignment
...
this adds coreBetterScore() which takes account of tilt and yaw
alignment when comparing cores for lane switching
this ensures we don't switch to a lane that is unaligned due to the
zero error score for unaligned lanes
4 years ago
Andrew Tridgell
3836b59041
AP_NavEKF3: fixed handling of failed compass
...
when a compass goes unhealthy due to sensor failure we should try
another compass after 10s if another compass is available
5 years ago
Harshit Kumar Sankhla
edc3709653
AP_NavEKF3: implement sensor affinity using EK3_AFFINITY parameter
...
this allows the EKF core index to be used to select a GPS/baro/mag
instance. This is an alternative to GPS blending that allows EKF lane
switching to be used to select the right combination of GPS and IMU
add logging to XKFS message
5 years ago
Randy Mackay
384029d278
AP_NavEKF3: rngBcnDataNew made local
5 years ago
Randy Mackay
8c877f02af
AP_NavEKF3: shrink beacon arrays and add assert
5 years ago
Randy Mackay
4c40d2ac3f
AP_NavEKF3: readyToUseBodyOdm uses delayed imu and wheel encoder timestamps
5 years ago
Randy Mackay
056a2de260
AP_NavEKF3: wheelOdmDataNew member made local
...
this variable is never used outside the writeWheelOdom method
5 years ago
Randy Mackay
3d5161f364
AP_NavEKF3: remove unused usingWheelSensors variable
...
This variable was set but never consumed
5 years ago
Randy Mackay
c4b7a1c41a
AP_NavEKF3: flowDataToFuse moved to local variable
...
This variable is never used outside the SelectFlowFusion() method
This variable is always updated at the top of the function meaning subsequent calls to the function will always overwrite its previous value
5 years ago
chobits
c722367c6c
AP_NavEKF3: move extNavVelNew to local variable
5 years ago
chobits
7ad9e3548f
AP_NavEKF3: make reset source variable local
5 years ago
Peter Barker
6b701ae3b3
AP_NavEKF3: correct includes
5 years ago
Paul Riseborough
c8bbbd2720
AP_NavEKF3: Fix bug preventing setting of mag field states
5 years ago
Paul Riseborough
9a4108f55e
AP_NavEKF3: Simplify setting EK3_MAG_CAL
...
Don't require user to separately set EK3_MAG_CAL to fly without a magnetomer
5 years ago
Paul Riseborough
21ea5d5039
AP_NavEKF3: Require GSF yaw history for reset when not using a yaw sensor
5 years ago
Paul Riseborough
354b551ef0
AP_NavEKF3: Handle repeated FW flight without magnetometer
...
The EKF can build up large yaw errors on ground so it is safer to stop using GPS and re-align after launch as per first launch.
5 years ago
Paul Riseborough
8ff6780323
AP_NavEKF3: Clarify distinct use cases for 'takeoff expected'
5 years ago
Paul Riseborough
d8c2096ad8
AP_NavEKF3: improve angle accuracy during no mag launch
5 years ago
Randy Mackay
ec5f84c926
AP_NavEKF3: remove unused terrainHgtStableSet_ms
5 years ago
Randy Mackay
dccc3e22d9
AP_NavEKF3: minor comment fix re RNG_USE_HGT
5 years ago