While on newer firmware there is no version.txt anymore the script would give "Checking Bebop firmware version... 0.0.0.0" and continue. So not s strictly needed change. we should then check from "/data/FVT6.txt" there you find :FWVersion= or last resort eeprom.dat.txt contains build info like : date Feb 14 2018 time 09:36:04, we should not force max versionb allowed only give a warning "non tested firmware, proceed with caution" or some tinh along that line. As Soon as we testflown a newer firmware well still highen this number anyhow.
* Bump rustlink
* Changes in GEC, now communication and key exchange works. TODO is key renegotiation and some cleanup
* Bump pprzlink
* Update in GEC and related control panel configs
* Updated name
* Bump rustlink
* Changes in GEC, now communication and key exchange works. TODO is key renegotiation and some cleanup
* Bump pprzlink
* Update in GEC and related control panel configs
* Update atomic settings
* Minor changes, style fix
* Code style fix
* More code style fix
* Add nps target to cv_opencvdemo module
Also re-enabled the canny edge detector example, as the result
is easier to see than the blurring example.
Added the LDFLAGS produced by the current version of opencv_bebop.
* Ignore uncommitted changes to opencv_bebop submodule
Added an "ignore = dirty" flag to .gitmodules to prevent opencv_bebop
from showing up in the git status with uncommited changes once it is
installed.
* First attempt at bebop fisheye camera
* Try setting raw camera output size during initialization, does not work
SetImageSize() is ignored and SetEnvTextureSize() causes an
exception. Currently the camera output is too large for real-time
operation.
* FIX apply MT9F002_OUTPUT_SCALER to gazebo front camera
* Add MT9F002 camera cropping to fdm_gazebo
* Apply MT9F002_TARGET_FPS to bebop front cam in gazebo
* Set up bottom camera for bebop model
* Minor fixes and cleanup
* Fix MT9F002 partially applied to all cameras
* Remove sensors_params_bebop, introduce NPS_SIMULATE_MT9F002
Reduces duplicate code for sensor parameters, and follows the
same pattern as NPS_SIMULATE_LASER_RANGE_ARRAY.
Also does not require the sensor params to be changed in every
bebop airframe.
* Add gazebo model for Parrot Bebop 1
* Add first-order high pass filter
* First implementation of actuator dynamics and spinup torque
Flies ok with PID control
Does not fly ok with INDI, the required G1 R is 0.159 instead of
0.0022 (estimated by flying in adaptive mode). The other parameters
are close to those specified in the airframe file.
* FIX incorrect yaw behavior
Fixed two bugs:
1. gazebo_actuators.torques was set to NPS_ACTUATOR_THRUSTS
2. spinup torque direction is now also controlled by motor mixing
* Clean up debug code and file logger
* Clean up gazebo example airframes
* Minor cleanup
* FIX warnings about missing initializer in fdm_gazebo
* Update documentation
* Remove modules section, move to firmware
* Minor fixes to fdm_gazebo and high_pass_filter
* Move actuator dynamics to included airframe files
Should prevent duplicate actuator defines spread accross all
airframe files.
* Fix FILTER_ROLL_RATE in bebop2_indi airframe
Moved to firmware targets so it will be disabled in gazebo.
Is this actually working in jsbsim??
* Update documentation
* Fix newlines
* Minor fixes in nps_fdm_gazebo.cpp
* Re-enable filter roll rate
Caution: causes oscillations around roll axis in gazebo
* Disable roll filter and cleanup
* Fixed string related build warnings
* Changed Failures to accept all strings
* Updated pprzlink for v405 fix
* updated use of lowercase, uppercase and capitalize
Some warnings related to the use of the Label() macro are still there.
It seems that the only way would be generate the labels instead of
calling the macros.
See issue #2207
PPRZLINK Proxy is a tool to allow multiple aircraft simulations
with air-to-air communication using the PPRZLINK v2 protocol
using UDP.
As aircarft can't be identified by their IP addresses (most of the
time, all simulators are running on the same computer), each agent (GCS,
aircraft) are using different ports and the mapping between ports and IDs
is done by this tool.
Applying the modification proposed in #2138 after testing it.
The version check is for now disabled if the correct valid cannot be
retrive from the drone. This should be improved later.
close#2138
- add a callback to receive all the regid bodies of a frame in a single
list together with the correct timestamp
- use the timestamp instead of system time to compute dt
- remove the padding offset (yes, there is a bug in the code provided by
optitrack...)
- tested and working in real fligts
The new version of the Optitrack software (Motive 2.0) comes with NatNet
version 3. A Python module is provided with NatNet for
serialization/deserialization of the data streamed over the network.
This module is slightly modified here for a better handling of the
initialization and for stopping the threads.
The PPRZLINK message sent is REMOTE_GPS_LOCAL, so the transformation to
global coordinate system is left to the airborne part.
Michal Podhradsky
Hector Garcia de Marina
OpenUAS
Ewoud Smeur
matejkarasek
Kirk Scheper
Michal Podhradsky
Gautier Hattenberger
Tom van Dijk
Rijesh Augustine
Titus
Tom van Dijk
Fabien-B
Felix Ruess
Hector Garcia de Marina