Chu, Tianshu | Engineering Papers

DOE OSTI2022

Unconventional Photoconversion from In‐Plane 2D Heterostructures of 2D Transition Metal Carbides/Semiconductors

2D transition metal carbides/nitrides (MXenes) are thought of promising photothermal materials due to their broadband localized surface plasma resonance (LSPR) and attractive electromagnetic interference (EMI) properties. However, the weak oxidation resistance and unclear photothermal mechanism of MXenes impede their photothermal modulations. Herein, the in‐plane 2D Ti 3 C 2 /TiO 2 heterostructures are fabricated by partial oxidation and an unconventional photothermal effect under near infrared irradiation, even if the oxidation compromises the EMI and LSPR capacities, is demonstrated. Such performance is ascribed to the faster transfer of photoexcited electrons across the seamless in‐plane Ti 3 C 2 /TiO 2 heterointerface than the reported out‐of‐plane one, and the 1–2 orders of magnitude faster relaxation processes of electrons than the ground bleaching in the Original MXene. These experimental results are well‐supported by theoretical calculations. The conceptual advances broaden the fundamental understanding of the photoconversion of MXenes, which would be extended into a variety of applications, such as biomedical therapy, photosynthesis, and photovoltaics.

Energy & Fuels↗

DOE OSTI2022

PowerNet: Multi-agent Deep Reinforcement Learning for Scalable Powergrid Control

This paper develops an efficient multi-agent deep reinforcement learning algorithm for cooperative controls in powergrids. Specifically, we consider the decentralized inverter-based secondary voltage control problem in distributed generators (DGs), which is first formulated as a cooperative multi-agent reinforcement learning (MARL) problem. We then propose a novel on-policy MARL algorithm, PowerNet, in which each agent (DG) learns a control policy based on (sub-)global reward but local states and encoded communication messages from its neighbors. Motivated by the fact that a local control from one agent has limited impact on agents distant from it, we exploit a novel spatial discount factor to reduce the effect from remote agents, to expedite the training process and improve scalability. Furthermore, a differentiable, learning-based communication protocol is employed to foster the collaborations among neighboring agents. In addition, to mitigate the effects of system uncertainty and random noise introduced during on-policy learning, we utilize an action smoothing factor to stabilize the policy execution. To facilitate training and evaluation, we develop PGSim, an efficient, high-fidelity powergrid simulation platform. Here, experimental results in two microgrid setups show that the developed PowerNet outperforms the conventional model-based control method, as well as several state-of-the-art MARL algorithms. The decentralized learning scheme and high sample efficiency also make it viable to large-scale power grids.

24 POWER TRANSMISSION AND DISTRIBUTION↗