COMANDO: A Next-Generation Open-Source Framework for Energy Systems\n Optimization

Abstract

Existing open-source modeling frameworks dedicated to energy systems\noptimization typically utilize (mixed-integer) linear programming ((MI)LP)\nformulations, which lack modeling freedom for technical system design and\noperation. We present COMANDO, an open-source Python package for\ncomponent-oriented modeling and optimization for nonlinear design and operation\nof integrated energy systems. COMANDO allows to assemble system models from\ncomponent models including nonlinear, dynamic and discrete characteristics.\nBased on a single system model, different deterministic and stochastic problem\nformulations can be obtained by varying objective function and underlying data,\nand by applying automatic or manual reformulations. The flexible open-source\nimplementation allows for the integration of customized routines required to\nsolve challenging problems, e.g., initialization, problem decomposition, or\nsequential solution strategies. We demonstrate features of COMANDO via case\nstudies, including automated linearization, dynamic optimization, stochastic\nprogramming, and the use of nonlinear artificial neural networks as surrogate\nmodels in a reduced-space formulation for deterministic global optimization.\n