last_letter
last_letter copied to clipboard
Investigate the need for more precise physics integrators
What do other types of integration methods offer in terms of accuracy and speed, such as second-order, trapezoidal etc?
There may not be any need for anything else more than a simple Forward Euler (the current integrator), as indicated here (Chapter 7): [1] K. Krigsman and J. Nilsson, “Code generation for efficient real-time execution of diagnostic residual generators,” University of Linkoppings, 2004. This integrator offers significantly high stability margins.
This will be closed once Gazebo is employed as the Physics engine.