Search

The growth of electric vehicles (EVs) significantly increases the residential electricity demand and potentially leads to the overload of the transformer in the distribution grid. Aiming to coordinate the charging control of EVs, this article formulates the EVs charging problem as a Markov game with an unknown transition function and proposes a cooperative charging control strategy based on the multiagent deep reinforcement learning. The uncertainties from the dynamic electricity price, non-EV residential load consumption and drivers’ individual behaviors are considered to construct the dynamic charging environment. Each agent contains a collective-policy model and an independent learner. The collective-policy model is introduced to model other agent's behaviors by approximating their power consumption. The independent learner is used to learn the optimal charging strategy by interacting with the environment. The soft-actor-critic framework is adopted to train the independent learner, enabling the proposed method to address the continuous state and action. Agents are trained with only the local observation and approximation, indicating that the proposed approach is fully decentralized and scalable to the problem with multiple agents. Finally, several numerical studies constructed based on the real-world data demonstrate the effectiveness and scalability of the proposed approach. [ABSTRACT FROM AUTHOR]

The advance of nonvolatile memory in storage technology has presented challenges in redefining the ways in handling the main memory and the storage. This work is motivated by the strong demands in effective handling of page faults over ultralow-latency storage devices. In particular, we propose synchronous and asynchronous prefetching strategies to satisfy process executions with different memory demands in supporting of synchronous page fault handling. An adaptive CPU scheduling strategy is also proposed to cope with the needs of processes in maintaining their working sets in the main memory. Six representative benchmarks and applications were evaluated. It was shown that our strategy can effectively save 12.33% of the total execution time and reduce 13.33% of page faults, compared to the conventional demand paging strategy with nearly no sacrificing of process fairness. [ABSTRACT FROM AUTHOR]

The uncertainties from the distributed energy resources (DERs) and the prosumer’s heterogeneous characteristics bring great challenges to the operation of the community energy market. In this paper, the real-time energy sharing and management in the community market are decomposed into two-layered sub-problems: the household appliance scheduling problem and the internal energy trading pricing problem. A hierarchical deep reinforcement learning (HDRL) based scheme is proposed for the community energy trading with multiple households, containing a two-stage learning process. In the inner stage, a multi-agent deep reinforcement learning (MADRL) based approach is developed to learn the real-time appliances scheduling policy based on the local observations and given internal electricity price in a decentralized way. In the outer stage, a deep reinforcement learning (DRL) based pricing approach is proposed to determine the real-time internal electricity prices based on the participants’ historical net power and external energy supplier’s electricity prices. The scheduling policy and households’ state are not required in the pricing process. Finally, the simulations with real-world datasets demonstrate that the internal energy trading can significantly reduce the prosumers’ daily cost and the control performance of the proposed scheme is superior to the existing studies. [ABSTRACT FROM AUTHOR]

In this brief, an automatic power amplifier (PA) design system based on simulated annealing particle swarm optimization (SA-PSO) algorithm is proposed to perform multi-objective optimization in designing broadband Doherty like load-modulated balanced amplifier (LMBA). Specifically, by combining the programming environment with commercial electronic design automation (EDA) software, the optimization process can be implemented automatically. Then, complex and rigorous impedance trajectory aligning at different power levels can be easily obtained using the proposed optimization algorithm, demonstrating the automatic PA design system’s superiority over the built-in optimizer in commercial EDA software for designing broadband Doherty like LMBA. A 2.5-3.5 GHz dual-input Doherty like LMBA is simulated and measured, showing a saturated output power of 45.4-48.2 dBm, a saturation drain efficiency of 60.7% - 69.7%, a 6 dB power back-off DE of 59.2% - 65.8%, and a 9 dB power back-off DE of 52.7% - 66.9%. [ABSTRACT FROM AUTHOR]

Based on the variation trend of the depletion layer of GaAs-PIN photodiode (PD) under different optical power and reverse bias voltages, an extensive large-signal equivalent circuit model that characterizes PD in non-punch-through and punch-through states is proposed in this article. According to the analysis of the PD’s depletion layer variation, the nonlinear parameters of the model are described by exponential function and Gaussian function. To verify the reliability of equivalent circuit model, on-chip measurements were performed on PIN-PD fabricated by commercial GaAs process, and the measured ${I}$ - ${V}$ and reflection coefficient agree with the proposed model simulation results well. Furthermore, using the proposed model for simulation with CAD tools shows that reducing junction capacitance has potential in improving PD output power and gain flatness. At the same time, the model provides a certain reference for selecting bias condition in the design of the PD matching network and fabricating PD with high power output. [ABSTRACT FROM AUTHOR]

Gallium arsenide (GaAs) has shown potential for its wide range of applications in radio frequency integrated circuit (RF-IC) packaging due to its unique properties over the traditional Si wafers. We reported the fabrication of high-density Ag nanotwinned films deposited on (400)-oriented GaAs substrates by magnetron sputtering and electron beam evaporation. In contrast to Si, the Ti precoating on GaAs is no longer necessary to provide better adhesion of the Ag nanotwinned film on the substrates. The results indicate that both the sputtered and evaporated Ag thin films exhibit high-density columnar Ag grains vertically growing on the substrates with the coincidence site lattice (CSL)- $\Sigma 3$ twin boundaries fraction of 35.6%. The electron backscatter diffraction (EBSD) analysis reveals that the twin boundaries within the sputtered Ag columns consist of a strongly preferential orientation of (111). The successful fabrication of Ag nanotwinned film via evaporation deposition will allow the applications in the backside metallization industry to be practical. [ABSTRACT FROM AUTHOR]

This letter presents a fully integrated dual power mode Doherty power amplifier (PA) in the Nanjing Electronic Devices Institute (NEDI) 250-nm GaN process. Due to the advantages of high power (HP) and gain, the cascode architecture is introduced to both carrier and peaking PAs. To the author’s best knowledge, this is the first time cascode architecture is applied in GaN process for sub-6 GHz Doherty PA design. The equivalent parasitic network and interstage impedance matching network of cascode transistors are constructed and extracted within the frequency range of 1–15 GHz. High efficiency voltage–current waveform at the current generator plane of carrier PA is achieved. The gate and drain bias of the common gate transistor are adjusted to attain low power (LP)/HP mode. A Doherty PA prototype operating in 5G NR FR1 n78 is presented with a saturated power of 36.5–37.2 dBm, a peaking power added efficiency (PAE) of 49.6%–52.0%, a 6-dB back-off PAE of 34.4%–39.6%, and a gain of 12.2–12.9 dB for LP mode. The realized high-gain saturated power is 39.8–40.5 dBm with a relatively HP density of 3.96 W/mm, a peaking PAE of 42.5%–47.7%, a 6-dB back-off PAE of 32.0%–37.8%, and a gain of 16.0–16.4 dB for HP mode. [ABSTRACT FROM AUTHOR]