Power system emergency control is generally regarded as the last safety net for grid security and resiliency. Existing emergency control schemes are usually designed off-line based on either the conceived "worst" case scenario or a few typical operation scenarios. These schemes are facing significant adaptiveness and robustness issues as increasing uncertainties and variations occur in modern electrical grids. To address these challenges, this paper developed novel adaptive emergency control schemes using deep reinforcement learning (DRL) by leveraging the high-dimensional feature extraction and non-linear generalization capabilities of DRL for complex power systems.Furthermore, an open-source platform named Reinforcement Learning for Grid Control (RLGC) has been designed for the first time to assist the development and benchmarking of DRL algorithms for power system control. Details of the platform and DRL-based emergency control schemes for generator dynamic braking and under-voltage load shedding are presented. Robustness of the developed DRL method to different simulation scenarios, model parameter uncertainty and noise in the observations is investigated. Extensive case studies performed in both the two-area, four-machine system and the IEEE 39-bus system have demonstrated excellent performance and robustness of the proposed schemes.
Subsampled image acquisition followed by image inpainting in a scanning transmission electron microscope is a novel approach to control dose and increase the image frame rate during experiments, thereby allowing independent control of the spatial and temporal dose envelope during image acquisition. Here, subsampled imaging is shown to permit precise in situ observations of the fundamental kinetic processes behind nucleation and growth of silver (Ag) nanoparticles from an aqueous solution. At high sampling-levels, nanoparticles can be observed with morphologies that are consistent with strong interface interactions, i.e., rafts and pillars, whereas at low sampling-levels, the particles exhibit regular spherical morphologies. The relative numbers of rafts/pillars and regular nanoparticles, their sizes, and their incubation times can be attributed to local changes in the molar concentration of the Ag ions in the aqueous solution; higher sampling-levels significantly increase the reactants in the vicinity of the window, leading to rapid supersaturation and the precipitation on the window surface. These precisely controlled kinetics highlight subsampled imaging as a method by which the driving force for nucleation and growth (i.e., the electron beam) can be disentangled from the spatial/temporal resolution of the observation in all in situ experiments, providing a pathway to identify and quantify the importance of individual kinetic factors behind nucleation and growth in a wide variety of complex materials systems and architectures.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.