Modeling Concepts#

Foundational concepts for understanding the ANDES modeling framework.

Key Ideas#

Hybrid Symbolic-Numeric Framework#

ANDES uses SymPy to define model equations symbolically, then generates optimized numerical code automatically. This enables:

  • Automatic Jacobian derivation

  • Self-documenting models with LaTeX equations

  • Rapid prototyping of new models

Atomic Types#

Models are built from three fundamental atom types: parameters, variables, and services. These share a common v-provider interface (the v attribute) that enables interoperability in equations. Variables additionally serve as e-providers, contributing equation residuals to the DAE system.

DAE Formulation#

Power system dynamics are modeled as differential-algebraic equations:

  • Differential equations (f): Generator dynamics, controller states

  • Algebraic equations (g): Network power balance, algebraic constraints

System Architecture#

The System class orchestrates models, routines, and the DAE arrays.