Around the beginning of 1996, a cross-disciplinary group of researchers at Cornell University began developing and using a web-based electric power market testing platform called PowerWeb to study the effects of various market designs on the behavior of the participants and on the operation of the power system. PowerWeb required an underlying power system simulation and optimization software to clear its markets, so a set of power flow (PF) and optimal power flow (OPF) tools were developed in Matlab for that purpose and then released as a free, open source package called MATPOWER.

Now, some 20 years later, MATPOWER has grown into a popular tool for power systems research and education, currently with roughly 20,000 downloads per year. Because of its library of ready-to-run case files, its efficient primal-dual interior point OPF solver, interfaces to 3rd party solvers, and its OPF customization features, it has become a powerful research-enabling tool and a type of de facto standard benchmark against which other results, tools and methods are routinely compared. Version 6, to be released soon, includes the MATPOWER Optimal Scheduling Tool, or MOST, a generalized optimization tool for solving problems ranging from a simple deterministic economic dispatch problem to a secure, stochastic, multiperiod, hybrid unit commitment/OPF problem with storage modeling and more.

This talk will describe the evolution of the MATPOWER software and the growing complexity and difficulty of the problems it addresses.