RigidBodyDynamics

RigidBodyDynamics implements various rigid body dynamics and kinematics algorithms.

Design features

Some of the key design features of this package are:

• pure Julia implementation, enabling seamless support for e.g. automatic differentiation using ForwardDiff.jl and symbolic dynamics using SymPy.jl.
• easy creation and modification of general rigid body mechanisms.
• basic parsing of and writing to the URDF file format.
• extensive checks that verify that coordinate systems match before computation, with the goal of making reference frame mistakes impossible
• flexible caching of intermediate results to prevent doing double work
• fairly small codebase and few dependencies
• singularity-free rotation parameterizations

Functionality

Current functionality of RigidBodyDynamics.jl includes:

• kinematics/transforming points and free vectors from one coordinate system to another
• transforming wrenches, momenta (spatial force vectors) and twists and their derivatives (spatial motion vectors) from one coordinate system to another
• relative twists/spatial accelerations between bodies
• kinetic/potential energy
• center of mass
• geometric/basic/spatial Jacobians
• momentum
• momentum matrix
• momentum rate bias (= momentum matrix time derivative multiplied by joint velocity vector)
• mass matrix (composite rigid body algorithm)
• inverse dynamics (recursive Newton-Euler)
• dynamics
• simulation, either using an off-the-shelf ODE integrator or using an included custom Munthe-Kaas integrator that properly handles second-order ODEs defined on a manifold.

Closed-loop systems (parallel mechanisms) are supported, with optional Baumgarte stabilization of the loop joint constraints. Support for contact is very limited (possibly subject to major changes in the future), implemented using penalty methods.

Installation

Installing Julia

Download links and more detailed instructions are available on the Julia website. The latest version of RigidBodyDynamics.jl requires Julia 0.7, but we recommend downloading 1.0 (the latest stable Julia release at the time of writing). Version 0.7 of RigidBodyDynamics.jl is the last to support Julia 0.6.

Installing RigidBodyDynamics

To install the latest tagged release of RigidBodyDynamics, start Julia and enter Pkg mode by pressing ]. Then simply run

add RigidBodyDynamics

To use the latest master version and work on the bleeding edge (generally, not recommended), instead run

add RigidBodyDynamics#master

A third option is to clone the repository (to the directory printed by julia -e 'import Pkg; println(Pkg.devdir())'):

dev RigidBodyDynamics

About

This library was inspired by IHMCRoboticsToolkit and by Drake.

Most of the nomenclature used and algorithms implemented by this package stem from the following resources:

• Murray, Richard M., et al. A mathematical introduction to robotic manipulation. CRC press, 1994.
• Featherstone, Roy. Rigid body dynamics algorithms. Springer, 2008.
• Duindam, Vincent. Port-based modeling and control for efficient bipedal walking robots. Diss. University of Twente, 2006.

Contents

• Quick start guide
• Spatial vector algebra
• Index
• Types
• The @framecheck macro
• Functions
• Joints
• Index
• The Joint type
• Functions
• JointTypes
• Rigid bodies
• Index
• The RigidBody type
• Functions
• Mechanisms
• Index
• The Mechanism type
• Creating and modifying Mechanisms
• Basic functionality
• MechanismState
• Index
• The MechanismState type
• Functions
• Algorithms
• Index
• The DynamicsResult type
• Functions
• StateCache
• Simulation
• Index
• Basic simulation
• Lower level ODE integration interface
• URDF parsing and writing
• Benchmarks

Citing this library

@misc{rigidbodydynamicsjl,
 author = "Twan Koolen and contributors",
 title = "RigidBodyDynamics.jl",
 year = 2016,
 url = "https://github.com/JuliaRobotics/RigidBodyDynamics.jl"
}