gramian/emgr

emgr -- EMpirical GRamian Framework

Christian Himpe

• Personal Website

My Development Interests

• Metadata Lakes
• Metadata Catalogs
• Data-Driven Model Reduction
• System Theory
• Scientific Computing

My Main Repositories

• chicken-arcadedb -- A CHICKEN Scheme driver for ArcadeDB (Scheme)
• matrico -- A flonum matrix module for CHICKEN Scheme (Scheme)
• morgen -- Model Order Reduction for Gas and Energy Networks (MATLAB/Octave)
• hapod -- Hierarchical Approximate Proper Orthogonal Decomposition (MATLAB/Octave)
• emgr -- EMpirical GRamian Framework (MATLAB/Octave)

My Main Gists

• Flamegraphs for Octave (Octave)
• MATLAB and Octave Snippets (MATLAB/Octave)
• versioninfo (MATLAB/Octave)
• Build Octave from Source (Bash)
• Chicken Scheme Dockerfile (Dockerfile)

Contributed To

• arcadedb
• M-M.E.S.S
• FlexiBLAS
• qrupdate-ng
• octave

emgr's Issues

Fix state-weighting

The state weighting closure (emgr.m, line 180) should be:

            wei = @(m) 1.0 ./ max(sqrt(eps),vecnorm(m,2,1));

similarly emgr.py, line 212 should be:

            return 1.0 / np.maximum(math.sqrt(np.spacing(1)),np.linalg.norm(m, 2, axis=0))

This will be fixed in version 5.9

Linearization and Simulink Compatibility

Hi
Model order reduction is widely used by Control System researchers and many system dynamic tests can be applied only on the linearized model of nonlinear systems. Therefore, a function should be added that yields linearized models of the nonlinear systems.
Moreover, a utility which enables the user to use the nonlinear system in simulink for various test can enhance the usability and scope of the software.