Hybrid Symbolic-Numeric Framework for Power System Modeling and Analysis

With the recent proliferation of open-source packages for computing, power system differential-algebraic equation (DAE) modeling and simulation are being revisited to reduce programming efforts. Existing tools require manual efforts develop code numerical equations, sparse Jacobians, discontinuous components. This paper proposes a hybrid symbolic-numeric framework, exemplified by an Python-based library ANDES, which consists symbolic layer descriptive numeric vector-based computation. method enables implementation DAE models mixing matching components, through described. In rich set components standard transfer function blocks provided besides essential elements rapid modeling. ANDES can automatically generate robust fast code, as well high-quality documentation. Case studies present a) two implementations turbine governor model TGOV1, b) flow computation time break down MATPOWER systems, c) validation time-domain with commercial software using three test systems variety models, d) full eigenvalue analysis Kundur's system. Validation shows that closely matches tool DSATools flow, simulation, analysis.