Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Extending Affine Recourse Policies to Hybrid Systems


M. Schmitt

Semester Thesis, FS 12 (10138)

This semester project addresses robust model predictive control of hybrid systems re- stricted by hard state and input constraints. In controls, robust optimization has been used to provide control inputs which maximize a certain objective while ensuring hard state and input constraints satisfaction for any possible disturbance from a bounded set. Optimization over "open loop" control inputs in presence of disturbances is very conservative, however. This eect can be reduced by including "recourse" in the problem formulation, i.e. taking into account that inputs are applied in a receding horizon fashion and intelligent decisions will be made in the future. For linear systems with disturbances bounded in polytopic or ellipsoidal sets, the optimization over ane recourse functions emerged as a valid trade o between the computational burden of obtaining a good feedback policy and simplicity of computation. This thesis aims at extending the idea of ane recourse to hybrid systems. A simple hybrid system (buck converter) is chosen as a benchmark for comparison with robust MPC without recourse. A computationally expensive controller based on a mixed-integer formulation is presented, which achieves superior performance. As an alternative, a com- putationally ecient controller is proposed which is based on a relaxed system model in the MPC and a projection of the (relaxed) inputs to a feasible set of mixed integer inputs. iii


Type of Publication:

(13)Semester/Bachelor Thesis

J. Warrington

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2012:IFA_4148
Permanent link