Distribution Circuits Research Articles

Memory-Computing Decoupling: A DNN Multitasking Accelerator With Adaptive Data Arrangement

Multiple deep neural networks (DNNs) are increasingly used in real-world intelligent applications, such as intelligent robotics and autonomous vehicles to collectively complete complicated tasks running on edge devices. Because each layer of the subtasks prefers a distinct dataflow due to the heterogeneity in shape and scale of the network layers, a variable dataflow approach on the DNN accelerators is urgently required. On DNN accelerators that enable multiple dataflows, however, we detect a dimension mismatch between parallel processing under the dataflow approach and linear data memory arrangement. When multiple DNN tasks share partial features or weights, the issue is further exacerbated. During processing, this mismatch causes a sluggish data supply from both off-chip and on-chip memory. Consequently, the overall throughput, performance, and energy efficiency suffer since DNN models are sensitive to data density. In this work, we reveal the mechanism behind this data dimension mismatch and present a series of metrics that quantify the influence on system performance. On this foundation, we offer a framework that tracks the data tensor dimension conversion and employs a flexible data arrangement over multi-DNN computation to adapt to dataflow variability. An accelerator architecture named data arrangement multi-DNN accelerator (DARMA) that features a data arrangement and distribution circuit and hierarchical memory for data dimension conversion is also presented. Since the mismatch is mitigated, the suggested accelerator outperforms current accelerators in terms of bandwidth and processing unit utilization. Through tests on VR/AR, MLperf, and other multitask applications, the evaluation results show that the proposed architecture provides both energy-efficiency and throughput improvements.

Delineating the transcriptional landscape and clonal diversity of virus-specific CD4+ T cells during chronic viral infection.

Although recent evidence indicates that CD4+ T cells responding to chronic viral infection are functionally heterogenous, our understanding of the developmental relationships between these subsets, and a determination of how their transcriptional landscape compares to their acute infection counterparts remains unclear. Additionally, whether cell-intrinsic factors such as TCR usage influence CD4+ T cell fate commitment during persistent infection has not previously been studied. Herein, we perform single-cell RNA sequencing (scRNA-seq) combined with single-cell T cell receptor sequencing (scTCR-seq) on virus-specific CD4+ T cells isolated from mice infected with chronic lymphocytic choriomeningitis virus (LCMV) infection. We identify several transcriptionally distinct states among the Th1, Tfh, and memory-like T cell subsets that form at the peak of infection, including the presence of a previously unrecognized Slamf7+ subset with cytolytic features. We further show that the relative distribution of these populations differs substantially between acute and persistent LCMV infection. Moreover, while the progeny of most T cell clones displays membership within each of these transcriptionally unique populations, overall supporting a one cell-multiple fate model, a small fraction of clones display a biased cell fate decision, suggesting that TCR usage may impact CD4+ T cell development during chronic infection. Importantly, comparative analyses further reveal both subset-specific and core gene expression programs that are differentially regulated between CD4+ T cells responding to acute and chronic LCMV infection. Together, these data may serve as a useful framework and allow for a detailed interrogation into the clonal distribution and transcriptional circuits underlying CD4+ T cell differentiation during chronic viral infection.

High-sensitivity metamaterial sensor based on electromagnetically induced transparency (EIT) effect

We propose a metamaterial sensor based on electromagnetically induced transparency (EIT) effect, which is composed of two elliptical structures with splits. Based on the coupling of bright and dark modes, a low-loss transparent window is realized at 1.88 THz. Surface current distribution and equivalent circuit model are used to analyze the physical mechanism of the EIT effect. In addition, the sensitivity of the sensor is investigated for different thickness and refractive index for analyte. The simulation results show that when the analyte thickness is 15 μm and the refractive index is between 1 and 1.5, the sensitivity of the sensor is as high as 450.7 GHz/RIU, and the figure of merit (FOM) value is 7.7. And when the refractive index of analyte is 1.5 and its thickness is between 0 and 5 μm, the sensor can also be used for thickness sensing with a sensitivity of 40.8 GHz μm−1. In addition, the proposed metamaterial sensor can be used to simulate the detection of sodium chloride solution, and the sensitivity is 66.2 GHz/. These above features indicate that the proposed structure paves the way for designing high-sensitivity EIT sensors in the terahertz region, which plays important roles in promoting terahertz sensing and detection technology.

Synthetic data generation for machine learning model training for energy theft scenarios using cosimulation

AbstractTechnical and non‐technical losses in distribution circuits result in significant economic costs to power utilities. One type of non‐technical loss is energy theft by various means including illegal tapping of feeders, bypassing the meter, and billing fraud. These losses are usually hard to detect, and can remain undetected for long periods of time. Machine learning models have been proven effective in detecting these conditions, but rely on the availability of large, good‐quality training data sets. The problem is exacerbated by the imbalanced nature of data related to these conditions—energy theft, though costly, is very rare. The available data sets generally have very few samples of theft with most of the data pertaining to normal operation. Such data sets are generally not suitable to train machine learning models. In this paper, an overview of energy theft detection techniques, the challenges with their data needs, and the limitations of current techniques to bridge such data limitations is presented. A co‐simulation framework is proposed to generate reliable training data for machine learning algorithms for theft detection. An example scenario is presented and a machine learning model is built to detect certain kinds of energy theft.

Probabilistic optimization of preventive maintenance inspection rates by considering correlations among maintenance costs, duration, and states transition probabilities

Power system degradation has highly motivated utilities to perform preventive maintenance. Inadequate inspection and maintenance could lead to early failures. However, too much of them could be very costly; thus, many studies have been conducted to achieve the optimal inspection and maintenance rates. Previous studies have only considered the transition from a deterioration, repair, or inspection state to another one probabilistically. However, in practice, repair and maintenance costs and duration also vary randomly and are correlated with each other. This study proposes a probabilistic approach by considering all the aforementioned correlations and uncertainties to find the optimal inspection rates. The approach can incorporate the dependent deterioration phenomenon among different system/equipment components in the maintenance process. Additionally, a simple judgment matrix is assumed to estimate the present deterioration state before performing inspections or maintenance activities. The matrix makes classical approaches more usable from a practical point of view. A semi-Markov chain based on Monte Carlo simulations employing 95 percentiles of the total cost is utilized to determine the optimal inspection rates. Finally, the proposed method is applied to the distribution circuit breakers (CBs) of the Roy Billinton test system (RBTS) for illustration. The numerical results show that considering the correlations and uncertainties can change the optimal inspection rates and reduce the calculated costs and/or unavailability in comparison to the conventional approaches. Furthermore, the results indicate that the method is simple and accurate and can be integrated into asset management tools for the maintenance decision-making process.

Enhanced Performance of Single-Walled Carbon Nanotube-Germanium Near-Infrared Photodetector by Doping with Au Nanoparticles

This paper presents a near-infrared (near-IR) photodetector based on a gold nanoparticles-doped (AuNPs-doped), single-walled carbon nanotube–germanium (SWCNT/Ge) heterojunction. The responsivity, detectivity, and response time of the AuNPs-doped, SWCNT/Ge heterojunction photodetector measured 476 mA W−1 (a 291% improvement), 1.0 × 1012 cm Hz1/2 W−1 (a 208% improvement), and 8 μs, respectively. The mechanism of the enhanced performance originated from the surface modification by gold doping, which effectively improved the work function of the carbon nanotube films and thus increased the barrier height between the heterojunctions, as measured by the contact potential distribution (CPD) and open circuit voltage (Voc) of the SWCNT/Ge interface. In addition, we investigated the effect of various particle sizes on the performance and stability of the photodetector. The results demonstrate the promising prospects of the presented heterojunction photodetector for infrared detection applications.

Electronic transport in doped and dielectric inserted MLGNR interconnects: Crosstalk induced delay and stability analyses at sub-threshold regime

A simplified equivalent distributed temperature-dependent RLC circuit model of coupled multilayer graphene nanoribbon (MLGNR) interconnects has been discussed by considering the relaxation time induced mean free path (MFP), and its impact on line resistance (R) has been examined in detail. Further, the performance of copper, undoped, doped and dielectric inserted MLGNRs has been compared in terms of crosstalk induced delay and frequency response at sub-threshold operating conditions. It is revealed that the various scattering mechanisms (acoustic scatterings, edge roughness etc.) control the overall MFP in undoped and doped MLGNRs, however, the electron MFP in dielectric inserted MLGNRs depends on surrounding dielectric environment. Out of the five different cases of dielectric inserted MLGNRs, the insertion of hafnium oxide (HfO2) between GNR layers, along with silicon dioxide (SiO2) provides the best performance characteristics. To obtain minimum crosstalk delay at the subthreshold operating mode, copper, undoped, doped, and dielectric inserted MLGNRs have been optimized for thicknesses of 36, 38.25, 41.4, and 43.65 nm.

A Machine-Learning Pipeline for Large-Scale Power-Quality Forecasting in the Mexican Distribution Grid

Electric power distribution networks face increasing factors for power-quality (PQ) deterioration, such as distributed, renewable-energy generation units and countless high-end electronic devices loaded as controllers or in standalone mode. Consequently, government regulations are issued worldwide to set up strict PQ distribution standards; the distribution grids must comply with those regulations. This situation drives research towards PQ forecasting as a crucial part of early-warning systems. However, most of the approaches in the literature disregard the big-data nature of the problem by working on small datasets. These datasets come from short-scale off-grid configurations or selected portions of a larger power grid. This article addresses a study case from a region-sized state-owned Mexican distribution grid, where the company must preserve essential PQ standards in approximately 700 distribution circuits and 150 quality-control nodes. We implemented a machine-learning pipeline with nearly 4000 univariate forecasting models to address this challenge. The system executes a weekly forecasting pipeline and daily data ingestion and preprocessing pipeline, processing massive amounts of data ingested. The implemented system, MIRD (an acronym for Monitoreo Inteligente de Redes de Distribution—Intelligent Monitoring of Distribution Networks), is an unprecedented effort in the production, deployment, and continuous use of forecasting models for PQ indices monitoring. To the extent of the authors’ best knowledge, there is no similar work of this type in any other Latin-American distribution grid.

Line Surge Arresters Applications On The Multi Circuit Overhead Lines

This paper presents application of line surge arresters (LSA) on the different voltage level multi circuit overhead lines. Double circuit shielded compact line with and without distribution circuit on the same tower is analyzed. Distribution circuit has lower insulation level, meaning that almost all flashovers will happen on that circuit. Flashovers on the distribution circuit help to improve lightning performance of the transmission circuits. Flashovers on the distribution circuit diverts fraction of the lightning current along its phase conductors, improving at the same time coupling between distribution and transmission circuits. All software simulations are performed using sigma slp software package. A short description of the modeling for multi circuit flashover rate determination is given. In order to prevent flashovers on the distribution circuit LSA are installed on this circuit only. The improvement in the transmission circuit lightning performance is similar to that obtained without LSA. LSA installed on the distribution circuit are much cheaper than transmission LSA.

Simulation of Electro-Thermal Runaway and Thermal Limits of a Loaded HVDC Cable

This paper presents simultaneous simulation of interdependent electric and thermal fields inside a dc cable under loaded conditions using analogous distributed circuit models for thermal and electrical phenomena. Certain nonlinearities are integrated in such circuit models, leading to simulation of electro-thermal runaway, for the first time, in a dc cable. Using the suggested, nonlinear circuit models, complete dynamics of the temperature profile and electric field inside the cable dielectric can also be obtained under various test and operating conditions until electro-thermal runaway limits in time domain. The results are useful for design and development of dc cables with safety factors, apart from understanding the electro-thermal limits. The models are validated with experimental results, found to be in excellent agreement.

A Novel Technique for Dynamic Analysis of an Electromagnetic Rail Launcher using FEM Coupled with Simplorer

The performance of a rail gun depends on the current density distribution over the rail and armature as it determines the force that accelerates the projectile of the rail gun. A finite element method (FEM) coupled with Simplorer was developed to model and study the performance of the rail gun. The rail gun was modeled using an ANSYS eddy current field solver to determine the current density distribution and equivalent rail gun circuit for the given rail gun geometry. The armature velocity was then calculated using Simplorer by coupling the obtained equivalent rail gun circuit and exciting the rails using a capacitor-based pulsed power supply (PPS) system. The FEM coupled with Simplorer method was verified by numerical calculations for the rectangular rails and also with other researchers’ value, and that showed a good agreement between the results. Further, the current density distribution over rails and armature and velocity of the armature was calculated for different rail cross sections such as circular concave, circular convex, rectangular concave, rectangular convex, T-shaped concave, and T-shaped convex with a C-shaped armature. It was observed that the circular convex rail gun with C-shaped armature showed minimum current density distribution and gives a higher value of armature velocity compared with other rail gun structures. Thus, the circular convex armature was found to be suitable for the electromagnetic (EM) rail gun launchingsystem.

A Novel Design of a Compact Frequency-Selective Surface With High Selectivity and Angular Stability

A novel bandpass frequency-selective surface (FSS) with high selectivity and angular stability is presented. In the skirts of the narrow passband, two transmission zeros (TZs) are generated, leading to the steep roll-offs at both left and right skirts. The realization of two TZs using a single piece of a printed circuit board facilitates the fabrication. The current distribution and equivalent circuit model are given to illustrate the generation mechanism of the TZs. The proposed FSS can achieve good stabilities of polarization and angle. The size of this FSS is <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.15\lambda $ </tex-math></inline-formula> . The prototypes of the proposed FSS are fabricated, and good performance is obtained.

Analysis and Modeling of a Directional Microwave Coupler as SMD Component

The article describes the calculation, modeling and layout of the directional coupler. In the course of the work the market analysis was carried out, the model of directional coupler in the frequency range of 2.9—4.5 GHz, the layout of the device and measuring equipment were developed. Directional couplers with different transient attenuation are widely used in the distribution circuits of antenna arrays, as well as to control the power of the radio signal propagating in the transmission line, in power meters, automatic gain control systems, radio receivers, etc. The choice of circuit solutions used for the construction of directional couplers is carried out depending on the specific technical requirements for electrical parameters, the range of operating frequencies and the capabilities of the technological base. Practically all variety of structures of directional couplers realized on elements is a result of modification of the simplest structures. The main types of directional taps include stub, single-stage and tandem-based. During the simulation, it turned out that the directional coupler based on a tandem connection of two connected transmissions with the location of the transmission lines on different layers of the multilayer printed section was suitable for this work. This is due to the fact that when modeling the directional coupler with transmission lines on one printed circuit board, The gap between these lines is too small (&amp;10 nm), which is extremely difficult to perform on a microstrip structure.

Defect‐induced current coupling in multi‐junction solar cells revealed by absolute electroluminescence imaging

AbstractAn electroluminescence (EL) anomaly singular spot was observed in an industry‐standard InGaP/GaAs multi‐junction solar cell (MJSC). Affected by this singular spot, the spatially resolved subcell current distributions were found to exhibit unique opposite characteristics, which we call “defect‐induced current coupling.” Herein, we conducted systematic investigations to reveal the defect‐induced current coupling phenomenon, for the first time, through the absolute EL imaging technique. Specifically, a modified carrier‐balance model was first proposed to describe the subcell optoelectronic distribution around the singular spot. Then, optical and electrical properties including subcell non‐radiative recombination distributions, J–V characteristics, essential photovoltaic parameters, and energy losses were quantitatively extracted for characterizing the performance inhomogeneity caused by the defect‐induced coupling. Moreover, a three‐dimensional distributed circuit model was established to numerically simulate the absolute EL emissions and extract local electrical parameters around the singular spot. Simulation results revealed that the defect‐induced current coupling phenomenon is attributed to the deterioration of weak‐diode and shunt parameters at the GaAs bottom cell, and the efficiency of the MJSC was estimated to be reduced by 0.408% compared with the defect‐free MJSC model.

Distributed Nonlinear Model and Fast Analysis for In-Band IMD3 Prediction of Surface Acoustic Wave Resonators.

The motivation of this work is to analyze the in-band intermodulation distortion (IMD) occurring in surface acoustic wave (SAW) devices, using a recently developed fast method based on the input-output equivalent sources (IOES). The method calculates the equivalent current sources of a given harmonic (H) or IMD, which when applied at the boundaries of any uniform nonlinear region produce the same nonlinearities as the full distributed circuit. The accuracy of the method is validated with a very simplified SAW resonator with ten digits, which is modeled by a discretized Mason-based circuit. The IOES method provides equal results to the ones obtained through harmonic balance (HB) simulations, performed by means of commercial software, being the first 1000 times faster. Once the accuracy of the method is guaranteed, it is used to analyze the measured in-band IMD3 of several lithium tantalite 42° cut leaky SAW (LSAW) resonators with different pitches and duty factors at the B66 long term evolution (LTE) frequency band. Those resonators are comprised of 100 and 20 electrode pairs for the active region and each of the reflectors, respectively, which implies the analysis of a very large distributed nonlinear problem with thousands of nonlinear local sources. The IOES method takes 35.4 s in simulating 51 frequency points, whereas this simulation is not possible using a commercial HB simulator on a general-purpose computer.

Design of Periodic End-Coupled Leaky-Wave Antenna With Asymmetrically Loading Unit

In this letter, a periodic end-coupled leaky-wave antenna is designed based on an asymmetrically loading unit. A simple balancing condition in the equivalent distributed circuit model is developed for the open stopband suppression by introducing a complex effective resonant frequency. The proposed balancing condition naturally satisfies both the frequency-balancing condition and Q-balancing condition in the lumped circuit model. Based on the derived formula, the physical parameters could be quantitatively determined. Subsequently, two typical periodic end-coupled leaky-wave antennas are designed, implemented, and verified experimentally. The measured reflection coefficient and the radiation results show that the open stopband at the broadside frequency is successfully suppressed.

Voltage Regulation With High Penetration of Low-Carbon Energy in Distribution Networks: A Source–Grid–Load-Collaboration-Based Perspective

In this article, a source–grid–load-collabora tion-based control framework is proposed to improve the power quality of active distribution networks (ADNs) with high penetration of low-carbon energy. First, hybrid dynamics of ADNs are characterized by addressing the voltage regulation and operation economics in each operation mode, and the mode switching control is designed in line with the operation principle of the on-load tap changer, where voltage security events are used to build the event-triggered functions. Second, multiobjective optimization is formulated with consideration of the system-wide operation cost and distribution circuit loss of the ADN in a relatively slow time scale, while in the fast time scale, all the inverter-based distributed generators, energy storages, and static var compensator devices are coordinated at the source–load side, through which multiple voltage issues, including voltage profile issue and voltage increment issue, can be addressed in a fully distributed manner. Finally, simulation results validate the effectiveness and robustness of the proposed method based on the modified IEEE 33-bus system.

Optimal Battery Sizing for Frequency Regulation and Energy Arbitrage

This paper proposes an optimization methodology for sizing and operating battery energy storage systems (BESS) in distribution networks. A BESS optimal operation for both frequency regulation and energy arbitrage, constrained by battery state-of-charge (SoC) requirements, is considered in the proposed optimization algorithm. We use utility historical data as input in a case study on a real-world distribution circuit. The results indicate that a BESS generally brings high profits by participating in the frequency regulation market and should be sized to its highest allowable power capacity rather than its energy capacity.

Computer Interface for Remote Teleoperated Mobile Robot Based on LabVIEW and Open Source Hardware

A remote controlled mobile robot utilizing National Instruments LabVIEW software, communicating over a low-rate serial ISM radio link. The robot is modular in design with system modules connected via serial interfaces to a system controller. A custom power supply and distribution circuit board provides power to the robot using sealed lead acid batteries as the source. The mobile robot streams system voltage and current status, JPEG images, ambient temperature and humidity readings, and GPS location. The project demonstrates the successful use of LabVIEW to develop a system using communication over lowrate serial radio links.

Compact Wideband Pattern Reconfigurable Antennas Inspired by End-Fire Structure for 5G Vehicular Communication

Aiming for 5G-NR vehicular applications, this paper proposes two miniaturized wideband pattern reconfigurable antennas with different pattern steering capabilities. They are inspired by an end-fire structure that could be easily mounted on cars. This newly proposed planar pattern reconfigurable end-fire structure is firstly presented and studied by current distribution and equivalent circuit analysis. By turning the ON / OFF state of the PIN diodes on the passive dipole, a planar reconfigurable antenna with one omnidirectional and two end-fire beams is implemented. Additionally, a low-profile pattern diversity antenna with a compact size of 0.311 λ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> × 0.244 λ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> × 0.106 λ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> is implemented by loading a reactive metasurface, its main lobe can be steered in the broadside direction (35°, 0°, and −35°). Both antennas achieve a wideband response over 3.30 to 4.20 GHz (24.0%) range and good radiation efficiency of 84.0%. They feature the advantages of a miniaturized size (low-profile), structurally simple and fabrication friendly design, wideband, good radiation efficiency (low-loss), and flexible multi-beam switching capability, which are well suited for the coming sub 6 GHz vehicular applications in different scenarios.