Commit Graph

11413 Commits

Author SHA1 Message Date
Gautier Hattenberger d8e590dead [settings] remove settings from automatically unloaded modules 2014-10-08 23:29:28 +02:00
Gautier Hattenberger 85e79a4018 [conf] update all example airframe for new modules' settings 2014-10-08 15:28:15 +02:00
Gautier Hattenberger 5e9e6de411 [ocaml] remove leading white space when get tree values 2014-10-08 15:22:55 +02:00
Gautier Hattenberger e2e97aa709 [settings] select new modules' settings by default 2014-10-08 15:05:24 +02:00
Gautier Hattenberger f0c3dfb578 remove another unneeded settings file 2014-10-01 11:26:34 +02:00
Gautier Hattenberger f81295dccb fix conf_example 2014-10-01 10:39:40 +02:00
Gautier Hattenberger ba82c0419f [settings] fix example conf 2014-10-01 10:22:35 +02:00
Gautier Hattenberger 83403a0ba3 [settings] move settings to module files and remove old files 2014-09-30 19:19:47 +02:00
Gautier Hattenberger 170e5b6843 [settings] allow several groups of settings in modules 2014-09-30 18:49:28 +02:00
Gautier Hattenberger 7230aa4a92 [settings] add settings directly inside modules files 2014-09-30 13:57:17 +02:00
Gautier Hattenberger 5605864d75 [settings] variable size modules settings area 2014-09-17 17:29:44 +02:00
Gautier Hattenberger 03eafced58 [settings] add a few more settings in modules 2014-09-16 22:37:17 +02:00
Gautier Hattenberger e39292c77b [settings] improve settings in pprzcenter
- check box for faster enable/disable
- loading from modules
2014-09-16 22:17:35 +02:00
Freek van Tienen 9e3510e51e [loggers] Direct memory logger 2014-09-16 17:14:19 +02:00
Felix Ruess 42192213e5 Merge branch 'math_functions'
Convert most type specific macros to functions and general cleanup.
This should make the code easier to use, debug and wrap externally. Additionally code size probably reduces since it's not all inline anymore...

Also cleaned up all the headers and added extern "C" to all the header files the math lib can easier be used in C++ applications.

So far the macros still exist and call the respective functions for backwards compatibility.

Merging pull request #819

* math_functions: (32 commits)
  [math][ins] rewrite/move some math to ins_float_invariant
  [airborne] use math functions instead of macros
  [math] some dox updates
  [math] more cleanup, e.g. rmat_[transp]_vmult
  [math] rmat|quat_identity functions
  [math] add double versions for vect3/quat norm
  [math] VECTx_NORM updates
  [math] int32_rmat_*mult functions
  [math] algebra: remove some more type specific macros
  [math] more x_integrate_fi functions
  [math] fix code style
  [math] geodetic: RMat instead of Mat33 for ltp_of_ecef
  [math] remove some unused or type specifc VECTx macros
  [math] add float_quat_integrate_fi
  [math] set default PAPARAZZI_SRC
  [test] add math c files
  [math] add missing alias
  [math] wgs84_ellipsoid_to_geoid as function
  [math] use math functions in orientation conversion
  [misc] fix natnet2ivy compilation
  ...
2014-09-15 17:18:33 +02:00
Felix Ruess 005327a516 Merge pull request #818 from paparazzi/propagate_dt
Pass dt to ahrs/ins propagation and updates instead of using the defines in the implementations.

The fixedwing ins_alt_float now uses the real dt for alt_kalman update from baro or gps.

For the AHRS: If the frequencies are not defines or USE_AUTO_AHRS_FREQ is TRUE, the dt is calculated from the system time.
Should be cleaner like this and less prone to misconfiguration, possible drawback however is that you get some more jitter on the DT.

The complementary filters now don't use the AHRS_x_FREQUENCY defines for calculation of the gains anymore and instead count the number of propagations since last update.
2014-09-15 16:17:16 +02:00
Felix Ruess 0502f631ed [airborne] calc dt with usec resolution 2014-09-15 15:52:44 +02:00
Felix Ruess 12d0eca18d [ins] alt_float, bound dt and used fixed GPS_DT if defined 2014-09-15 15:46:50 +02:00
Felix Ruess 20ee0bb0d4 [fixedwing] minor cleanup for ahrs_sim 2014-09-15 15:46:49 +02:00
Felix Ruess 1cec1e9db9 [rotorcraft] only calc mag dt if actually using mag 2014-09-15 15:46:49 +02:00
Felix Ruess e3abe48436 [ahrs] also pass dt to update_accel and update_mag
Not so sure passing dt to the update functions is the best way to go...
Is only used to update the gains of the cmpl filters right now.
But it makes it possible to properly compute the gains and avoids having a dependency on sys_time in the AHRS.
2014-09-15 15:46:49 +02:00
Felix Ruess 38873ac4fc [test] update some mostly unused ahrs tests 2014-09-15 15:46:49 +02:00
Felix Ruess df023584f9 [fixedwing] remove reduced propagation rate stuff
Also calculate dt if USE_AUTO_AHRS_FREQ or !defined(AHRS_PROPAGATE_FREQUENCY)
2014-09-15 15:46:49 +02:00
Felix Ruess 9a0bee8be8 [rotorcraft] ahrs dt: only calc if USE_AUTO_AHRS_FREQ or !defined(AHRS_PROPAGATE_FREQUENCY) 2014-09-15 15:46:49 +02:00
Felix Ruess e0afb70909 [test] fix test_settings 2014-09-15 15:04:19 +02:00
Felix Ruess ac6ef1b6c7 [conf] update order in conf_example.xml 2014-09-15 14:59:50 +02:00
Felix Ruess f02bc5ffa7 [settings] fix includes 2014-09-15 13:28:27 +02:00
Felix Ruess 7f316988a3 [settings] firmware specific StoreSettings handlers 2014-09-15 11:10:00 +02:00
gtoonstra 41865cf729 radio change for 8 channels and a script to start flightgear 2014-09-15 11:09:07 +02:00
gtoonstra 5a6b13f849 Cridea airframes 2014-09-15 11:08:55 +02:00
Felix Ruess dd1746bd83 [math][ins] rewrite/move some math to ins_float_invariant
- use float_quat_vmult instead of FLOAT_QUAT_RMAT_B2N|N2B
- use float_quat_derivative instead of float_quat_vmul_left
2014-09-12 18:05:27 +02:00
Felix Ruess 56457d0801 [airborne] use math functions instead of macros
sed to the rescue:
find sw -path sw/airborne/math -prune -o -name "*.[ch]" -exec sed -i 's/\([ ]*[A-Z0-9_]*QUAT_OF[A-Z0-9_]*(\)\([a-zA-Z0-9_]*,[ ]*\)\([a-zA-Z0-9_]*)\)/\L\1\E\&\2\&\3/g' {} +
and so on...
2014-09-12 18:01:11 +02:00
Felix Ruess a3e860ccb7 [math] some dox updates 2014-09-12 17:58:53 +02:00
Felix Ruess 8d6d84376d [math] more cleanup, e.g. rmat_[transp]_vmult
for consistency always call the fuctions x_rmat_[transp]_vmult,
FLOAT_RMAT_VECT3_MUL and FLOAT_RMAT_VECT3_TRANSP_MUL are removed.
2014-09-12 15:44:56 +02:00
Felix Ruess c7fcf10745 [math] rmat|quat_identity functions 2014-09-12 14:12:57 +02:00
Felix Ruess b8349ce252 [math] add double versions for vect3/quat norm 2014-09-12 00:32:53 +02:00
Felix Ruess 336defae49 [math] VECTx_NORM updates
- norm functions/macros return the scalar value, don't take norm as arg
- remove some type specific macros where the generic ones can be used
2014-09-11 19:52:12 +02:00
Felix Ruess 3f984923f2 [math] int32_rmat_*mult functions 2014-09-11 16:44:13 +02:00
Felix Ruess 1a8be652fc [math] algebra: remove some more type specific macros 2014-09-11 15:13:39 +02:00
Felix Ruess aa10dbe2a8 [math] more x_integrate_fi functions 2014-09-11 15:13:39 +02:00
Felix Ruess 8cf9da0f1b update fix_code_style.sh script 2014-09-10 23:21:32 +02:00
Felix Ruess 63d67a322f [math] fix code style 2014-09-10 23:20:48 +02:00
Felix Ruess 43d563f9e2 [math] geodetic: RMat instead of Mat33 for ltp_of_ecef 2014-09-10 21:39:08 +02:00
Felix Ruess 8d52fef350 [math] remove some unused or type specifc VECTx macros
find sw -path sw/airborne/math -prune -o -name "*.[ch]" -exec sed -ri 's/FLOAT_(VECT[23]_(ASSIGN|COPY|SMUL|DIFF|SUM|ADD|SUB|DOT_PRODUCT|CROSS_PRODUCT))/\1/g' {} +
2014-09-10 20:52:06 +02:00
Felix Ruess 630f2e4ab4 update code style, pointer to type 2014-09-10 17:55:02 +02:00
Felix Ruess 70d36ec583 update eclipse code style
eclipse still sucks at formatting code though
2014-09-10 17:06:58 +02:00
karlito139 9a1a811597 [modules] new high_speed_logger_direct_memory using SPI
Added a tool to log messages by connecting a simple flash chip directly to the SPI of the autopilot

closes #815
2014-09-10 15:53:02 +02:00
Felix Ruess 04bd833e76 add initial fix_code_syle.sh script
You need to have astyle installed.
Run with ./fix_code_style.sh <FILENAME>
2014-09-10 15:23:55 +02:00
Felix Ruess 5b7ef47585 add sublime project file 2014-09-10 15:09:16 +02:00
Gautier Hattenberger 557c0112fc [ocaml] uint32 now fits into a Int64
close #831
2014-09-10 14:31:25 +02:00