autumnday / ltl_sampling

Sampling base path planning with LTL specifications for multi-vehicles

Sampling base path planning with LTL specifications

1. Development Environment

• OS: Ubuntu 14.04,
• Compiler: gcc 4.7+
• Building System: CMake

2. Install Dependencies

• Building tools, Git, CMake

$ sudo apt-get install build-essential
$ sudo apt-get install git
$ sudo apt-get install cmake

• lcm

Download the source code from website.

Follow the build instructions website.

$ git clone https://github.com/lcm-proj/lcm lcm
$ cd lcm
$ ./configure
$ make
$ sudo make install
$ sudo ldconfig

Post install

$ export LCM_INSTALL_DIR=/usr/local/lib
$ echo $LCM_INSTALL_DIR > /etc/ld.so.conf.d/lcm.conf

• Spot

Download the released package from the website. Extract the files into a folder you prefer.

$ cd <path-to-spot-source-folder>
$ ./configure
$ make
$ sudo make install

By default the library-related files are install at "/usr/local". You need to add this path to "LD_LIBRARY_PATH" if you haven't done so. Otherwise the system may not be able to find the shared library when trying to run programs linked against the spot library. Add the following line to your ~/.bashrc:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib/

• Eigen

$ sudo apt-get install libeigen3-dev

2. Set Up Workspace

You can set up your workspace at any location you prefer. Here I'm using "~/Workspace/ltl_sampling" as an example.

$ mkdir -p ~/Workspace/ltl_sampling
$ cd ~/Workspace/ltl_sampling
$ git init
$ git remote add origin https://github.com/zetian/ltl_sampling
$ git pull origin master

3. Build Project

$ cd ~/Workspace/ltl_sampling
$ mkdir build
$ cd build
$ cmake ..
$ make

Edit the path variable for lcm, edit ~/.bashrc

export LTLSAMPLING=$HOME/Workspace/ltl_sampling
export PYTHONPATH=$LTLSAMPLING/src/lcmtypes/python:$LTLSAMPLING/src/lcmtypes/python/sampling:$PYTHONPATH

4. Test the example

Single UAV case

Follow the test example in src/tests/test_sample_dubins.cpp. Once you set up all the paremeters, tasks and environment, run src/python_vis/sampling_vis.py for visualization first, then run the test example, the result would look like:

Comments about the test example

1. Orange parts are the region of interests(ROI), grey parts are obtacles.
2. The task for this example is to visit all ROIs without considering the order.
3. The size of the map in this example is 100*100, the minimun turning radius is 15 for the UAV.
4. Set a large enough iterations to ensure a feasible solution, the solution towards optimal when iterations -> infinite.

Multiple UAV case

Follow the test example in src/tests/test_cbba.cpp, the result looks like:

Comments about the test example

1. There are 2 UAVs in this set up, both of them have a minimun turning radius 15.
2. The environment and task is the same as before: visit all ROIs in the map without considering the order.