17 May 2019
Release of OpenOCL v4.3325 Apr 2019
Modeling for Reinforcement Learning and Optimal Control: Double pendulum on a cart17 Apr 2019
Release of OpenOCL v4.2010 Apr 2019
Release of OpenOCL v4.00Model Predictive Control, short MPC, is the application of optimal control theory to actual physical systems. Because many dynamical systems need to be controlled with a high frequency (for example in the millisecond range), in the implementation of a (model) predictive controller the computation time for calculating the control inputs needs to be considered. In the best case the computation time is deterministic (realtime), but often a fast enough average computation time is sufficient.
Due to this strict demand on the computational efficiency, linear system models, and quadratic cost models are preferred, but there is ongoing research on realizing predictive controllers with non-linear models.