Installation

Table of Contents

• Requirements:
  • System Requirements
  • Required Dependencies
• Building the CT
  • For ROS or catkin users
    • Compiling the Library
    • Building and running examples
    • Building and running the unit tests
    • Building the documentation
  • Using plain cmake.
    • Compiling the Library
    • Building and running examples
    • Building and running the unit tests
    • Building the documentation
    • Clean up
• Optional Dependencies and Bindings
• build arguments

Requirements:

System Requirements

• Ubuntu 16.04 or 18.04
• a compiler with C++14 support, e.g. gcc or clang

Required Dependencies

• install Eigen 3 via

  sudo apt-get update 
  sudo apt-get install libeigen3-dev 

• install boost

  sudo apt-get install libboost-all-dev

• clone the kindr library into your workspace

  cd ~/your_workspace/src
  git clone https://github.com/ANYbotics/kindr.git

• finally, clone the control toolbox into your workspace

  cd ~/your_workspace/src
  git clone https://github.com/ethz-adrl/control-toolbox.git

Building the CT

For ROS or catkin users

Compiling the Library

It is assumed that you have a valid catkin-workspace set up, e.g. using catkin tools. Then, execute the following steps.

cd ~/your_workspace
catkin build -DCMAKE_BUILD_TYPE=Release

Building and running examples

To build an run the examples, e.g. the optimal control and optimal filtering examples:

cd ~/your_workspace
catkin build -DCMAKE_BUILD_TYPE=Release -DBUILD_EXAMPLES=true
source devel/setup.bash
rosrun ct_optcon ex_NLOC
rosrun ct_optcon ex_KalmanFiltering

Building and running the unit tests

To build an run the unit tests:

cd ~/your_workspace
catkin build -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=true

Building the documentation

To build the documentation execute

catkin build ct_doc -v --make-args doc

This will build the documentation. The index.html can be found in ct_doc/doc/html

Using plain cmake.

Compiling the Library

First, make sure you build and install the kindr-library, see here. Then go to your local clone of the control toolbox and run the build script:

cd your_workspace/control-toolbox/ct
chmod 775 build_ct.sh
./build_ct.sh -DCMAKE_BUILD_TYPE=Release

Building and running examples

Then go to your local clone of the control toolbox and run the build script with the following flags.

cd your_workspace/control-toolbox/ct
./build_ct.sh -DCMAKE_BUILD_TYPE=Release -DBUILD_EXAMPLES=true

This creates a /build subdirectory in every single CT-project, where the examples will be located. For running the nonlinear optimal control or filtering examples, for instance, go to your workspace an run

./src/control-toolbox/ct_optcon/build/examples/ex_NLOC
./src/control-toolbox/ct_optcon/build/examples/ex_KalmanFiltering

Building and running the unit tests

To build an run the unit tests:

cd your_workspace/control-toolbox/ct
./build_ct.sh -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=true

Building the documentation

cd your_workspace/control-toolbox/ct
chmod 775 build_doc.sh
./build_doc.sh

The index.html can be found in ct_doc/doc/html

Clean up

In order to clean up the build workspaces, run the clean script

cd your_workspace/control-toolbox/ct
chmod 775 clean_ct.sh
./clean_ct.sh

Optional Dependencies and Bindings

The following dependencies and bindings are optional but can help to enhance the user-experience or performance of the CT

• lapack (enables Schur method as an alternative to iterative method in ct::optcon::LQR)

  $ sudo apt-get install liblapack-dev 

• clang compiler (faster compilation for large linear models), install from command-line via

  $ sudo apt-get install clang 

  and possibly replace clang with a more specific version, e.g. clang-3.8. Please make sure that the clang and clang++ commands are in your path. You may need to create a symbolic link in /usr/bin or use a bash alias.
• IPOPT or SNOPT (for ct::optcon::SnoptSolver and ct::optcon::IpoptSolver as used by ct::optcon::DMS)
• blasfeo linear algebra and <a href=https://github.com/giaf/hpipm">HPIPM solver as alternative high-performance (constrained) linear-quadratic optimal control solver. The CT currently supports hpipm version 0.1.1. Detailed reference about blasefeo can be found here. Both packages were developed by Gianluca Frison, University of Freiburg.
• install the catkin_tools build system

  sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu `lsb_release -sc` main" > /etc/apt/sources.list.d/ros-latest.list'
  wget http://packages.ros.org/ros.key -O - | sudo apt-key add -
  sudo apt-get update
  sudo apt-get install python-catkin-tools

• Qwt (for basic plotting)
• MATLAB (for Matlab logging)
• clang-format, install via

  $ sudo apt-get install clang-format

build arguments

The following additional build flags are available, which can be appended to the build command above

Build Flag	Default value	Description
-DBUILD_EXAMPLES=<BOOL>	FALSE	Builds examples for all packages
-DBUILD_HYQ_FULL=<FALSE>	FALSE	Build HyQ examples and executables (warning: slow with GCC!)
-DUSE_CLANG=<BOOL>	FALSE	Use CLANG instead of the default compiler
-DCLANG_CXX_COMPILER=<clang-bin>	"/usr/bin/clang++"	Set clang C++ compiler binary
-DCLANG_C_COMPILER=<clang-bin>	"/usr/bin/clang"	Set clang C compiler binary
-DUSE_INTEL=<BOOL>	FALSE	Use Intel compiler instead of the default compiler
-DINTEL_CXX_COMPILER=<clang-bin>	"/opt/intel/bin/icc"	Set Intel C++ compiler binary
-DINTEL_C_COMPILER=<clang-bin>	"/opt/intel/bin/icpc"	Set Intel C compiler binary