Special Session on Power System Optimization and Machine Learning Models (PSO-ML)

Information

Scope

Technological developments and environmental concerns have led to an unprecedented evolution of energy systems. In the new paradigm, advanced energy resource management is needed to enable the intensive integration of distributed energy resources, taking advantage of its full potential. Furthermore, modern power systems require a satisfactory level of flexibility over different time horizons to maintain operation at a reasonable cost and satisfy restrictions imposed by the distribution system operator. Despite the improvement in forecasting techniques, it is a fact that power systems were not designed to effectively deal with the increased levels of uncertainty associated with distributed energy resources.

Due to memory requirements and limited runtimes, current works are often restricted to small cases in terms of problem size and are rarely able to consider all sources of uncertainty together, thus lacking realism. Computational intelligence approaches are proving to be very effective in dealing with large problems with nonlinearities, and new metaheuristics have been continually proposed in the literature to mitigate these problems. Moreover, recent advances in the field of machine learning are enabling reaching more accurate predictions of generation and demand, consequently reducing the uncertainty, and contributing to a more efficient power system management and operation.

PSO-ML aims at providing an advanced discussion forum on recent and innovative work in the fields of machine learning and optimization as solutions to address power system related problems. Special relevance is indorsed to solutions involving the application of artificial intelligence approaches, including metaheuristic optimization, automated machine learning, , forecasting methodologies and agent-based systems.

Keywords

Artificial Intelligence; Energy Resources Management; Electricity Markets; Machine Learning; Metaheuristic Optimization; Multi-Agent Systems; Renewable Energy;