Data-driven comparison of federated learning and model personalization for electric load forecasting

Residential short-term electric load forecasting is essential in modern decentralized power systems.Load forecasting methods mostly rely on neural networks and require access to private and sensitive electric load data for model training.Conventional neural network training aggregates all data on a centralized server to train one global model.However,the aggregation of user data introduces security and data privacy risks.In contrast,this study investigates the modern neural network training methods of federated learning and model personalization as potential solutions to security and data privacy problems.Within an extensive simulation approach,the investigated methods are compared to the conventional centralized method and a pre-trained baseline predictor to compare their respective performances.This study identifies that the underlying data structure of electric load data has a significant influence on the loss of a model.We therefore conclude that a comparison of loss distributions will in fact be considered a comparison of data structures,rather than a comparison of the model performance.As an alternative method of comparison of loss values,this study develops the"differential comparison".The method allows for the isolated comparison of model loss differences by only comparing the losses of two models generated by the same data sample to build a distribution of differences.The differential comparison method was then used to identify model personalization as the best performing model training method for load forecasting among all analyzed methods,with a superior performance in 59.1%of all cases.

Reviewing 40 years of artificial intelligence applied to power systems-A taxonomic perspective

Artificial intelligence(AI)as a multi-purpose technology is gaining increased attention and is now widely used across all sectors of the economy.The growing complexity of planning and operating power systems makes AI extremely valuable for the power industry.Until now,there has been a lack of clarity regarding the specific points along the power system supply chain where AI applications demonstrate significant value,as well as which AI domains are best suited for such applications.This study employs an AI taxonomy and automated web search to qualitatively and quantitatively unveil the biggest potentials of AI in the power industry.Our analysis,based on a review of 258’919 publications between 1982 and 2022,reveals where AI applications are particularly promising.We consider six AI domains(reasoning,planning,learning,communication,perception,integration&interaction)and 19 use cases from the power supply chain(i.e.,generation,transmission networks,distribution networks,isolated grids/microgrids,market operations and retail).Our findings indicate that,as of now,the focus is predominantly on AI applications in power retail(55%),transmission(14%)and generation(13%).Most analyzed works describe applications built on algorithms of the AI domains“learning”(45%)and“planning”(14%).Results also suggest that the current definition of AI domains is ambiguous,and they highlight missing information on the actual use and successful implementation of AI in power system use cases.