# Evolutionary Computation for Power System Planning

Focus: Electrical Engineering

Transmission Expansion Planning (TEP) aims at identifying a set of new pieces of equipment, as transmission lines and transformers, to be installed on the grid over an extended planning horizon. TEP has non-linear and non-convex natures, and it is an NP-hard problem, which means that the computational time required to solve the problem does not grow polynomially within the problem dimension, or in other words, the solution approach is not scalable. For these reasons and for many years, the scientific community has been working to build TEP approximated models in order to make the problem workable. However, there is no guarantee that solutions obtained through approximated models correspond to optimal solutions regarding the original TEP problem. As a matter of fact, there is no guarantee these solutions are even feasible when applied to the complete TEP formulations.
To handle all the mentioned characteristics of TEP problems, in the last years modern heuristic algorithms (as constructive heuristic algorithms, metaheuristics, hyper-heuristics, mateheuristics) have been applied to a variety of TEP approaches. The diversity of models and approximations often lead to discussions on the dichotomy of TEP models that can be summarized in the following question: Which one is better? Approximate the mathematical model until it is possible to reach an optimal solution, or search for a sub-optimal solution in the original and complete problem? In the first approach, it is always necessary to prove that the obtained optimal solution for the modified problem is at least feasible in the original one. In the second case, search procedures lack mathematical proofs for optimality, even it is always possible to use the original model, and therefore, ensuring feasibility.
In this way, this course aims at presenting:
1) A brief introduction to Transmission Expansion Planning: AC and DC mathematical models.
2) Modern heuristics algorithms: Application on TEP problem.
Each module consists of 2 hours, resulting in a 4-hour course. As dealing with general concepts, the course is open to Graduate and Undergraduates, as well as to anyone interested in participating and discussing the topic.

Professor: Phillipe Vilaça (Universidad Rey Juan Carlos, Espanha) and Bruno Henriques Dias (UFJF)

Language: English

Mode of instruction: online