Robotics Demo

1. Kinematics - Forward and Inverse Kinematics

Fn 1.1: Symbolic Calculation using DH Table

using symbolic math package in Python - SymPy - to calculate forward and inverse kinematics with defined dh tables, see 1.1Kinematics_SymPy_DH.ipynb

matlab code is also available in the repository, but requires symbolic math toolbox

Fn 1.2: Dual Quaternion Representation

self-defined dual quaternion class to transfer dh table to dual quaternion and calculate forward and inverse kinematics, see 1.2Kinematics_Robot.ipynb

Eg 1.1: 2D Serial Robot

using example from this website.

2D serial robot with 2 links and 2 joints
workspace visualization (serial #TODO)

Eg 1.2: 3D Serial Robot

#TODO test quaternion and dual quaternion with serial robot
inverse kinematics calculation #TODO
forward kinematics calculation (serial #TODO)
3D serial robot with 3 links and 3 joints

Eg 1.3: 2D Parrallel Robot

2D Parrallel Robot
workspace visualization ( parrallel #TODO)
inverse kinematics calculation
forward kinematics calculation ( parrallel#TODO)

Eg 1.4: 3D Parrallel Robot

2. Visulization

Fn 2.1: URDF Visualization

using yourdfpy to load robot defined in URDF format.

"data/urdf_files": contains urdf files for different robots. Source based on yourdfpy, draw robot in 2D/3D space with matplotlib

Fn 2.2: STL/OBJ Visualization

dual quaternion reference frame visualisation (#TODO) see implementations in 1.2Kinematics_Robot.ipynb

3. Control Method

Eg 3.1: Kalman Filter in Motion Tracking

with traditional Kalman Filter and Unscented Kalman Filter (filterpy package)

Eg 3.2: Unscented Kalman Filter with 1D Pendulum

sigma points visualization {width=200 height=100}
UKF estimation result (parameters sensitive)

Eg 3.3: Particle Filter / Monte Carlo Localization in Motion Tracking

Pariticle initialization {width=100 height=100}
Particle Filter estimation result (parameters sensitive)

Kexin-Wei/Robotics_demo