About

libBipedal contains various algorithms related to bipedal walking. This library was originally developed as part of my Bachelor's Thesis at H2T.

This library is written in C++11, but still uses boost::shared_ptr to integrate better with Simox.

libBipedal is licensed under the simplified BSD 2-clause license.

How to build

mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make

Structure

src/ and include/ are subdived into several components:

1. pattern algorithms related to generating dynamically stable walking pattern. Currently the approach based on ZMP Preview control [1] is implemented.

2. stabilizer algorithms related stabilizers for bipedal walking. Currently the approach based on [2] is implemented. Also, a heuristic approach that requires the global pose of several reference frames.

3. recovery algorithms for push recovery. The implemented approach is based on the Future Immediate Capture Point [3].

Todo

• Evalutae interface: We could make Simox optional for most code.

• MMM is only needed for TrajectoryPlayer and TrajectoryExporter make it optional

• The push recovery is very basic and is mostely a hack: Do this properly. I think the best way would be to integrate a force-feedback to check when the recovery foot has actually reached the floor. This is a problem because we are actually falling thus the pre-planed foot trajectory will much likely not hit the floor when expected.

• Add in-depth documentation for the public interfaces

• bipedal.h should be split up into Pattern.h, Stabilizer.h etc. to avoid rebuilding the whole project if a predefine was changed.

Dependencies

• Simox build with SimDynamics

• Bullet 2 build with double precision

• MMMcore for trajectory playback

References

[1] Kajita, Shuuji, et al. "Biped walking pattern generation by using preview control of zero-moment point." Robotics and Automation, 2003. Proceedings. ICRA'03. IEEE International Conference on. Vol. 2. IEEE, 2003.
[2] Kajita, Shuuji, et al. "Biped walking stabilization based on linear inverted pendulum tracking." Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on. IEEE, 2010.
[3] Koolen, Twan, et al. "Capturability-based analysis and control of legged locomotion, Part 1: Theory and application to three simple gait models." The International Journal of Robotics Research 31.9 (2012): 1094-1113.