Review of Hybrid Data-driven and Physics-based Modeling for the Operation of New-type Power Systems
Constructing new-generation power systems dominated by renewable energy is crucial to achieve China's carbon neutrality goal,but this will inevitably bring significant changes and challenges to the existing power grids.Hybrid data-driven and physics-based modeling(hybrid modeling for short)is an emerging technique to combine the advantages of physic laws and data,showing great potential to serve as an important analysis tool for the new-generation power systems.To this end,this paper clarifies the relevant concepts and use cases at first,and then discusses the research trends and hotspots in recent literature.A general framework to evaluate the performance of hybrid modeling from an aspect of technical features and hybrid patterns is also proposed.In addition,this paper is focused on the operation of new-generation power systems,and has fully summarized the pros and cons of hybrid modeling in addressing the existing technical challenges.Further suggestions for future research work and pilot projects are discussed at last.
new-type power systemhigh renewable energy penetrationdata-drivenknowledge-drivenartificial intelligencemachine learning
国家科技期刊平台
NSTL
万方数据
