Hybridization Toolbox for Model Predictive Control

Datacite citation style:
Gharavi, Leila (2023): Hybridization Toolbox for Model Predictive Control. Version 1. 4TU.ResearchData. software. https://doi.org/10.4121/2a4a7bed-63b9-43d9-a4d2-192bc9163dd1.v1
Other citation styles (APA, Harvard, MLA, Vancouver, Chicago, IEEE) available at Datacite

Software

choose version: version 2 - 2025-01-30 (latest)
version 1 - 2023-09-21

This toolbox can be used to hybridize any nonlinear function given as its input argument, which can be either a nonlinear prediction model or the nonlinear function expressing the boundary of the feasible region, i.e. the nonlinear constraints.


A grid is generated on the function domain and the toolbox returns the hybridized form of the nonlinear function. The user can select the type and form of approximation based on the problem type:


  • For model approximation, the options are
  1. selecting the grid type and
  2. specifying the number of affine modes in the MMPS formulation.
  • For constraint approximation, the options are
  1. specifying the number of subregions,
  2. selecting between polytopic (MMPS-based) or ellipsoidal approximation, and
  3. choosing between boundary-based or region-based approximation.

History

  • 2023-09-21 first online, published, posted

Publisher

4TU.ResearchData

Format

ZIP file including MATLAB codes

Funding

  • Control of Evasive Manoeuvres for Automated Driving: Solving the Edge Cases (EVOLVE) (grant code 18484) NWO Open Technology Programme

Organizations

Delft University of Technology, Faculty Mechanical, Maritime and Materials Engineering (3ME), Delft Center for Systems and Control

DATA