This paper presents a framework to alleviate the Deep Reinforcement Learning (DRL) training data sparsity problem that is present in challenging domains by creating a DRL agent training and vehicle integration methodology. The methodology leverages accessible domains to train an agent to solve navigational problems such as obstacle avoidance and allows the agent to generalize to challenging and inaccessible domains such as those present in marine environments with minimal further training. This is done by integrating a DRL agent at a high level of vehicle control and leveraging existing path planning and proven low-level control methodologies that are utilized in multiple domains.
View Article and Find Full Text PDFA new family of transition-metal monosilicides (MSi, M = Ti, Mn, Fe, Ru, Ni, Pd, Co, and Rh) electrocatalysts with superior electrocatalytic performance of hydrogen evolution is reported, based on the computational and experimental results. It is proposed that these MSi can be synthesized within several minutes by adopting the arc-melting method. The previously reported RuSi is not only fabricated more readily but eventually explored 8 MSi that can be good hydrogen evolution reaction catalysts.
View Article and Find Full Text PDFWe comprehensively investigated the hydrogen evolution reaction (HER) activity of a series of transition metal phosphides (MPs) (M = Cr, Mn, Fe, Co, and Ni) using first-principles calculations. The free energy difference was calculated for possible sites on the surface to pinpoint the reactive sites and the associated catalytic activities. We found that the chemical properties of these considered MPs are different from those of WP, including CrP which has the same electronic configuration as WP but was shown not to be a good electrocatalyst.
View Article and Find Full Text PDF