As modern power systems become increasingly complex—integrating diverse inverter-based resources and advanced control schemes—there is a growing need for analysis tools that combine analytical rigor with computational flexibility. Traditional purely numerical simulations often lack transparency. SyNAPS bridges this gap by unifying symbolic computation with numerical simulation, enabling scalable modeling, analysis, and control design for DC and AC power systems.
It has the following key capabilities:
Integrates SymPy and SciPy to connect symbolic derivations with numerical evaluation. Symbolic models capture system structure and parameter dependencies, while numerical routines reduce the modeling complexity and enable scalable evaluation and simulation.
Automatically derives Jacobian matrices and linearized state-space models from nonlinear differential algebraic equations (DAEs). This enables transparent analysis of how operating points, grid conditions, and control strategies influence small-signal dynamics.
Leverages SciPy for numerical linear algebra, eigenvalue analysis, and time-domain simulation.
Defines parameter spaces (e.g., droop gains, PLL gains) and certifies stability using algebraic methods such as Linear Matrix Inequalities (LMIs) and convex optimization. Facilitates control co-design and exploration of stability boundaries.
SyNAPS is developed using both Python and MATLAB. The symbolic–numerical modeling and eigenvalue/sensitivity analysis modules are implemented in Python, leveraging libraries such as Sympy, Scipy, and Numpy. The LMI-based stability boundary and region identification module is developed in MATLAB.
To get started:
- Fork the repository and navigate to the
examplefolder. - Open the main Jupyter Notebook, which generates the parametric A-matrix and exports it as a MATLAB
.mfile. - After exporting, go to the
stabilityfolder to perform stability boundary and region identification.
- Buxin She (PI) – Conceptualization and framework design; DC network module development
- Ramij R. Hossain (Co-PI) – Symbolic–numerical modeling module development; AC network and DC-AC coupling module development
- Soumya Kundu (Co-PI) – Small-signal stability boundary and region identification module development
- Marcelo Elizondo (Thrust Lead) – Advisor
- Veronica Adetola (Initiative Lead) – Advisor
This work is funded by the Laboratory Directed Research and Development (LDRD) at the Pacific Northwest National Laboratory (PNNL) as part of the E-COMP initiative. PNNL is operated by Battelle for the DOE under Contract DOE-AC05-76RL01830
To be released.
Released under the GPL3 license.