The Reinitialized Partial Moment (RPM) approach,
embedded FIR filter models for identifying continuous-time and discrete-time transfer functions
The Matlab routines, proposed here, concern the implementation of the RPM model estimation. A complete presentation of this tool is given in:
R. Ouvrard and J. C. Trigeassou. On embedded FIR filter models for identifying continuous-time and discrete-time transfer functions: the RPM approach. Int. J. Control, 84(3): 616–632, March 2011 (pdf)
For the continuous-time system identification, a document presents the routines (CTRPM_report.pdf). All CTRPM routines can be downloaded here.
The Matlab routine lsctrpm.m computes the LS-estimates of continuous-time (CT) multi input multi output (MIMO) reinitialized partial moment (RPM) models.
The routine ivctrpm.m computes the LS-estimates with instrumental variable (IV) of CT MIMO RPM models.
A demo routine called demo_lsctrpm.m is built as a game for training and for a good use of this CT RPM model in SISO case.
And two test routines test_lsctrpm.m and test_ivctrpm.m show the algorithms with a basic example.
For the discrete-time system identification, the following links give the similar files DTRPM_report.pdf and DTRPM routines.