Benchmarking of system identification methods

The project has two objectives:

  1. Build a system identification benchmarks database, and
  2. compare existing identification methods on the benchmarks.
     

The benchmarks will be selected from practical stands, available at the laboratory of the Control Engineering and System Analysis Department of the ULB and the ELEC department of the VUB. The system identification methods to be compared are a range of subspace and prediction error methods (available in the System identification toolbox of MATLAB), the frequency domain identification toolbox, and a method based on structured low-rank approximation (available from http://slra.github.io).

2018
Emanuele Garone
Back to top