Evolution of Short Term Unit Scheduling Solutions in Power Systems Operations - Product Detail page title (Slides)

This webinar provides a high level brief overview of the solutions and algorithms evolved over the last several decades to solve the short term unit scheduling problem. The focus is on the practically most used solutions to power systems operations. The discussion will include priority merit list, dynamic programming, Linear Programming, Lagrangian relaxation, Augmented Lagrangian relaxation, and the game changer Mixed Integer Program (MIP). The mathematical breakthroughs and efficient search algorithms used in recent MIP commercial solvers made MIP the algorithm of choice widely adopted by the power systems industry to solve such highly complex, non-convex, and combinatorial optimization problem. The huge reduction in the MIP optimization solution time, the ease of mathematical constraints modeling, and the increased solution quality transformed the real-time applications and made it possible to model large number of complex resources’ types like combined cycle units, startup times and cost, transition time and cost, dynamic ramp rates as functions of resources MW. These advances in MIP made it possible to also include unprecedented large set of cross-resources and cross-time intervals coupling constraints such as transmission, and environment constraints in optimizing operations in day-ahead and real-time applications. As illustrative example, the presentation will highlight the current large-scale mathematical MIP model sizes, solution times, and the optimal MIP integer gaps achieved in both day-ahead and real-time market applications at the California ISO’s production system.