Pyomo — The Complete Guide: A Free Optimisation Modelling Language in Python
Pyomo is an algebraic modelling language for optimisation, written in Python and developed at Sandia National Laboratories with other contributors. You define variables, constraints and the objective in a form close to mathematical notation, and Pyomo passes the problem to a suitable solver.
Pyomo
Sandia National Laboratories and Pyomo contributors
An optimisation modelling language in Python — linear, nonlinear and mixed-integer problems for processes, energy and scheduling.
Get it from the vendorWhat you get
- Notation close to the maths — a model that is easy to read and check.
- Problem types — linear, nonlinear and mixed-integer.
- Freedom of solver — the same model with free or commercial solvers.
- Engineering uses — process design, production scheduling, energy system planning and allocation.
- A foundation for other tools — for example IDAES and oemof.solph.
What to know
Pyomo does not include solvers — you install a free or commercial solver separately, depending on the problem type. Start with a small linear problem and build up from there. It is licensed under BSD-3-Clause, commercial use included.
Related tools
For chemical process modelling on top of Pyomo — IDAES. For energy system models — oemof.solph and PyPSA. For multidisciplinary optimisation of engineering systems — OpenMDAO.
IDAES
IDAES consortium (U.S. DOE, NETL and partners)
A platform for process modelling and optimisation — flowsheets, unit models and property packages, from the U.S. DOE.
Get it from the vendor
oemof.solph
oemof developer group
A model generator for energy system optimisation (LP/MILP) — combined heat and power, storage, heat pumps and renewables.
Get it from the vendor
PyPSA
PyPSA community (TU Berlin)
Optimisation of power and energy systems — economic dispatch, capacity expansion, storage and sector coupling.
Get it from the vendor
OpenMDAO
NASA Glenn Research Center
A framework for multidisciplinary optimisation, from NASA.
Get it from the vendorThe bottom line
Pyomo is a natural choice for an engineer who wants to formulate an optimisation problem in Python — free, flexible and not tied to one solver.