Paul Riseborough
4fbdab27ff
AP_NavEKF: Use range finder for primary hgt ref in opt flow mode
...
Falls back to baro if range finder is unavailable
Adds parameter enabling user to select which height source (baro or range finder) will be used during optical flow nav.
2015-04-10 11:08:07 +09:00
Paul Riseborough
ff9917d338
AP_AHRS: Fix bug preventing EKF from initialising
2015-04-10 11:07:42 +09:00
Paul Riseborough
d618c55e2f
AP_NavEKF: Improved handling of noisy GPS speed accuracy data
2015-04-10 11:07:34 +09:00
Paul Riseborough
53358a4e10
AP_OpticalFlow: Add parameter to compensate for flow sensor yaw angle
...
AP_OpticalFlow: update parameter name
2015-04-10 11:07:31 +09:00
Paul Riseborough
cab171b580
AP_InertialNav: Add method to return EKF height above ground estimate
...
AP_InertialNav: Add validity flag to height above ground estimate
2015-04-10 11:07:27 +09:00
Paul Riseborough
e98edaa6cb
AP_NavEKF: Return more accurate validity status for height above ground
2015-04-10 11:07:24 +09:00
Paul Riseborough
586e4a7d2b
AP_NavEKF: Add Matlab derivations and simulations behind small EKF
2015-04-10 11:07:21 +09:00
Jonathan Challinger
dc5bdd5ad8
AP_AHRS: use filtered INS output to compute _accel_ef_blended
2015-04-09 20:19:56 +09:00
Randy Mackay
a095a8c3a1
AC_PID: more protection against NaN and Inf
2015-04-09 20:19:55 +09:00
Randy Mackay
bdfe8bbc54
AC_AttControl_Heli: remove RATE_RP_MAX, RATE_Y_MAX
2015-04-09 20:19:54 +09:00
Leonard Hall
afcd1c6ec3
AC_AttitudeControl: sqrt controller on Stab
2015-04-09 20:19:43 +09:00
Leonard Hall
ae77c18a19
AC_PID: Protect against NaN and INF
2015-04-09 18:30:12 +09:00
Andrzej Zaborowski
132cdc4916
Motors: Use _motor_to_channel_map in tricopter consistently.
...
In AP_MotorsTri.cpp the AP_MOTORS_MOT_1, _2 and _4 constants are
always mapped to actual output channels through _motor_to_channel_map
while the _CH_TRI_YAW is not, but there were a few inconsistencies
in this that could lead to things like PWM min and max values being
set on wrong channels.
It looks like all in all _motor_to_channel_map being in PROGMEM
probably doesn't help save memory and I'm not sure how useful it is
in the first place but regardless the usage should be consistent.
2015-04-09 11:43:22 +09:00
Randy Mackay
558018d839
InertialNav: add get_location method
2015-04-09 10:56:57 +09:00
Tom Pittenger
9ecbd0e30f
AP_L1_Control: fixed uTurn-wobble
...
When performing a u-turn in AUTO (waypoints are 180deg turn from each other) sometimes the aircraft can't decide to turn left or right and wobbled back and forth a couple times. There was existing code to solve this but it was not executing all the time like when in LOITER mode. Frankly, I don't understand the criteria that was required to make it execute and i suspect there is still a gremlin in that logic but just executing the check all the time makes prevents the wobble behavior.
2015-04-08 15:21:02 -07:00
Jonathan Challinger
6e8008ab69
AP_SerialManager: use protocol_match in find_baudrate
2015-04-08 09:14:31 -07:00
Randy Mackay
f260536373
AC_PID: fix example sketch compile error
2015-04-08 15:05:29 +09:00
Jonathan Challinger
6baec4952b
AP_Math: add conversions to and from 3-1-2 euler angles
2015-04-07 21:20:51 -07:00
Jonathan Challinger
a1d4f40c16
AP_Math: change quaternion operator* and operator/ to const
2015-04-07 21:20:51 -07:00
Jonathan Challinger
07735fefa6
AP_Math: fix div by zero in quaternion
2015-04-07 21:20:51 -07:00
Paul Riseborough
c57e25142c
AP_NavEKF: Update optical flow fusion maths to reduce height errors
2015-04-07 20:51:18 -07:00
Andrew Tridgell
35bac3ab4f
GCS_MAVLink: regenerate without mavlink_frame_char changes
2015-04-06 16:43:02 -07:00
Andrew Tridgell
df948e7358
AP_Mount: fixed for new GIMBAL_REPORT message
2015-04-06 16:42:23 -07:00
Andrew Tridgell
33824e0b2f
GCS_MAVLink: re-ran generate.sh
2015-04-06 16:30:16 -07:00
Jonathan Challinger
73b8890bd3
GCS_MAVLink: move GIMBAL_REPORT and GIMBAL_CONTROL and reserve messages
2015-04-06 16:29:02 -07:00
myly10
55befdc345
AP_NavEKF: Typo correction for EAS_NOISE description
2015-04-06 15:52:29 -07:00
mirkix
445f358839
AP_HAL_Linux: Fix compiler warning of MAX_PWMS redefinition (previously defined)
2015-04-06 15:34:50 -07:00
Andrew Tridgell
5284211c96
AP_SerialManager: fixed matching bug in protocol_match()
...
thanks to Mike Clement for noticing this
2015-04-06 15:28:03 -07:00
Mike Clement
c5c40adb76
GCS_MAVLink: use instance in find_baudrate
2015-04-06 14:47:51 -07:00
Mike Clement
89a50b99bb
SerialManager: use instance in find_baudrate
2015-04-06 14:47:11 -07:00
Paul Riseborough
9b3656e77c
AP_NavEKF: Fix bug introduced in Y axis flow fusion
2015-04-05 21:17:20 -07:00
Andrew Tridgell
6e52ef2a74
GCS_MAVLink: added send_local_position()
2015-04-05 09:16:41 -07:00
Andrew Tridgell
baf292def1
AP_NavEKF: prevent float exception on startup
2015-04-05 09:16:14 -07:00
Paul Riseborough
7fc0f026d2
AP_NavEKF: Fix bug in optical flow innovation variance integrity check
...
The check allowed negative innovation variances to pass. If this did occur, the filter would diverge.
2015-04-04 17:03:06 -07:00
Andrew Tridgell
389148ac54
AP_RangeFinder: use RNGFND_OFFSET for distance offset in PWM and I2C
2015-04-04 17:03:00 -07:00
Andrew Tridgell
55019abd71
AP_HAL: allow for default parameter override for PX4 and SITL
2015-04-04 09:01:40 -07:00
Andrew Tridgell
5a755c0400
AP_Param: allow a HAL_PARAM_DEFAULTS_PATH to override defaults
...
this allows a HAL_PARAM_DEFAULTS_PATH to be specified for a build to
override the default parameters for a build. This is useful to build a
firmware that has different default parameters
2015-04-04 09:01:40 -07:00
Andrew Tridgell
82a51e8791
GCS_MAVLink: use AP_Param::set_param_by_name()
2015-04-04 09:01:40 -07:00
Andrew Tridgell
5ca38e3d75
AP_Param: added set_param_by_name()
...
this simplifies the GCS_MAVLink code
2015-04-04 09:01:39 -07:00
Andrew Tridgell
c1a0375562
AP_NavEKF: prevent divide by zero in EKF logging
2015-04-04 07:09:02 -07:00
Paul Riseborough
d44cf14178
AP_AHRS : Publish EKF learned compass offsets
2015-04-03 15:24:47 -07:00
Paul Riseborough
10f050c53b
AP_NavEKF: Prevent baro drift causing toilet bowling
...
The magnetic field states are reset once at 1.5 metres and again at 5 metres. This height check was using the height at the first arm event as the reference. In the situation where there is baro drift and extgended time between the first arm event and flight, this can cause the magnetic field state to be reset when on the ground. If this happens when flying off a metallic surface, the resultant heading errors can cause sever toilet bowling.
2015-04-03 15:18:42 -07:00
Paul Riseborough
d4c60ca956
AP_NavEKF: Fix bug preventing reset of velocity after OF fusion timeout
2015-04-03 15:18:39 -07:00
Jonathan Challinger
95cd3480ec
AP_NavEKF: review all uses of dtIMU and use dtIMUactual where necessary
...
pair-programmed-with: Paul Riseborough <p_riseborough@live.com.au>
2015-04-03 15:18:09 -07:00
Andrew Tridgell
79b44d3988
AP_NavEKF: initialise gndEffectMode
2015-04-03 15:15:11 -07:00
Andrew Tridgell
3165c43dfe
AP_NavEKF: initialise gpsSpdAccuracy
2015-04-03 15:15:11 -07:00
Paul Riseborough
98c32012fa
AP_NavEKF: remove accel bias rate limit when disarmed
2015-04-03 15:15:11 -07:00
Paul Riseborough
fe76cb4c0b
AP_NavEKF : Make initial height variance consistent with baro noise
...
This makes sense to do because we initialise the state to the instantaneous baro reading
2015-04-03 15:15:11 -07:00
Paul Riseborough
a976e9dad2
AP_NavEKF : Fix bug in scaling of initial Z accel bias state variance
2015-04-03 15:15:11 -07:00
Paul Riseborough
92df3adb5e
AP_NavEKF : Fix bug in Z accel bias update for IMU1
2015-04-03 15:15:10 -07:00