As featured in the Journal of Hydraulic Research Paper "Implications of the selection of a particular modal decomposition technique for the analysis of shallow flows". This algorithm connects the spatially orthogonal Proper Orthogonal Decomposition with the temporally orthogonal Dynamic Mode Decomposition. From flow visualisation images, PIV or CFD vector fields it is possible to extract coherent structures which are quasi-spatially and temporally orthogonal.
Two example cases are included with the script. Please cite:
Higham, J.E., Brevis, W & Keylock, C.J. (2018) - Implications of the selection of a particular modal decomposition technique for the analysis of shallow flows , Journal of Hydraulic Research, DOI:10.1080/00221686.2017.1419990
Jonathan Higham (2020). POD_DMD (https://github.com/jonnyhigham/POD_DMD), GitHub. Retrieved .
Hello thank you for the great work. Can we apply this method on 1-D/time series signals?
I got it, Jonathan, I totally missed unzipping the data_cylinder zip file.
Hi Jonathan, Do you have any tutorial to work with this program? I appreciate you if you can prepare a tutorial on youtube or anywhere.
Thanks Yonatan, I have changed the script as per your suggestion. There is no change in functionality. It was just me being lazy and not checking the manual for findpeaks better!
I think there is a bug in the script for Matlab 2013: X is the vector index of the peaks (not a frequency value), and x is a frequency value.
[Y, X]=findpeaks(px,fx); [Y, X]=findpeaks(px);
[v, l]=min(abs(X-x)); [v, l]=min(abs(fx(X)-x));
% Find the matching frequency from POD
Added to github
Check for 2018a - works fine.
Change to description and code edit. No change in functionality.
Small change to script. For this case no change in functionality.
Change to syntax as suggested by Yonatan Cadavid. There is no affect to the functionality.
Change to text