Newton methods for k-order Markov Constrained Motion Problems
This provides an incremental improvement for robotics researchers and engineers working on motion optimization problems.
The authors developed a framework for robot motion optimization that applies classical constrained optimization methods (Gauss-Newton, Augmented Lagrangian, log-barrier) to a general class of motion problems, with the main novelty being an any-time version of Augmented Lagrangian. The result is an efficient implementation through abstractions that mirror the source code API.
This is a documentation of a framework for robot motion optimization that aims to draw on classical constrained optimization methods. With one exception the underlying algorithms are classical ones: Gauss-Newton (with adaptive step size and damping), Augmented Lagrangian, log-barrier, etc. The exception is a novel any-time version of the Augmented Lagrangian. The contribution of this framework is to frame motion optimization problems in a way that makes the application of these methods efficient, especially by defining a very general class of robot motion problems while at the same time introducing abstractions that directly reflect the API of the source code.