Gait and Trajectory Optimization for Bipeds

Status: Completed Tools: MATLAB, Python, CasADi Language: Python
Biped Optimization

Abstract

This project implements trajectory optimization on a 5-link kneed bipedal walker. Generated both periodic and non-periodic gaits for flat and uneven surfaces including sinusoidal, staired, and sloped terrains. Also developed understanding of non-linear dynamics and chaos theory for passive walking.

Implemented trajectory optimization using direct collocation methods. Developed a two-link passive walker and simulated it on MATLAB to understand chaotic dynamics and bifurcation behavior.

Note: This project is no longer maintained.

5-Link Biped Model

Trajectory Optimization

Using direct collocation methods with CasADi for optimal control:

Terrain Adaptation

Successfully generated stable gaits for various terrains:

Passive Walking & Chaos

Explored the fascinating dynamics of 2-link passive walkers — robots that walk down a slope with no actuators, powered only by gravity:

Basic Systems

The repository also includes trajectory optimization examples for fundamental systems:

Results

Generated stable periodic gaits for walking speeds of 0.3-1.0 m/s. Demonstrated chaotic behavior in passive walkers and analyzed bifurcations in the gait family. Human and ostrich-style gaits were both achieved with the 5-link model.