Skip to content

This project aims to develop an AI framework that integrates reinforcement learning (RL) and Central Pattern Generators (CPGs) to generate adaptive gaits optimized for specific terrain types

License

Notifications You must be signed in to change notification settings

MiniroLab/Gait-Optimization

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

image

Adaptive Gait Optimization for Bio-Inspired Turtle Robot

This repository contains code and simulation workflows for adaptive gait control of a sea turtle–inspired robot using Hopf Central Pattern Generators (CPGs) combined with Bayesian Optimization. The project investigates how AI-based optimization can improve locomotion efficiency, adaptability, and robustness across varying terrain conditions.

Features

  • Central Pattern Generator (CPG) Control

    • Implements Hopf oscillator–based rhythmic control for multi-joint locomotion.
    • Supports multiple gait patterns (synchronous, diagonal, turning).
  • Bayesian Optimization Framework

    • Tunes CPG parameters (frequency, amplitude, coupling) for speed, energy efficiency, and cost of transport.
    • Uses Gaussian Process (GP) surrogate models with exploration–exploitation balance.
  • Simulation Environment

    • MuJoCo + PyChrono interfaces for testing gait dynamics.
    • Warm-up phase implementation for oscillator stabilization.
    • Real-time logging of speed, displacement, energy, and cost of transport.
  • Adaptive Terrain Response

    • Integrates terrain classification with gait adaptation.
    • Demonstrates robustness across sand, rock, and aquatic-like surfaces.

About

This project aims to develop an AI framework that integrates reinforcement learning (RL) and Central Pattern Generators (CPGs) to generate adaptive gaits optimized for specific terrain types

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published