Large Gain Research Articles

Implementation of a Robust PMAE Algorithm to Limit Large Gain Variation in a Fuel-Cell Integrated EV Charging Architecture

This paper proposes an adaptive error convergence technique to improve the charging dynamics of an electric vehicle (EV) charging system in a grid-integrated proton-exchange membrane fuel cell (PEMFC) architecture. The fuel cell is most often characterized by its slow dynamics during multiple load transients and preferably operated as an auxiliary energy source in EV charging architecture. The delayed response affects the overall charging dynamics with further controller phase lag and non-linear system parameters to induce a significant convergence error during reference command tracking. To effectively control multi-source integrated charging dynamics in an EV charging architecture, a projection-modified adaptive estimation (PMAE) algorithm is proposed in this paper. The control algorithm establishes a variable gain adaptation within a bounded error contour and updates the controller gains based on the tracking error dynamics gradient. The fuel cell, supporting other energy resources, provides a stronger charging resiliency with a duty-modified control technique at the EV charging station. To validate the effectiveness of the controller implementation, a grid-integrated multi-objective EV charging prototype is designed and developed for grid-to-vehicle (G2V) and vehicle-to-grid (V2G) at variable operating conditions.

Phonon-assisted upconversion photoluminescence in Mn-doped two-dimensional hybrid metal halides

Phonon-assisted upconversion process is one of the important means to generate upconversion photoluminescence (UCPL), and has important applications in laser refrigeration and other fields. However, until now, the studies on UCPL behavior are still insufficient. Herein, this work synthesizes a series of Mn-doped two-dimensional hybrid metal halide (2D HMH) thin films and investigates their optical properties using various spectral techniques. Raman spectroscopic measurement confirms that Mn-doping is achieved via interstitial doping rather than Pb-substitution. By the measurement of the temperature-dependent PL spectrum, it is found that Mn-doping will weaken the thermal expansion effect of 2D HMHs, but will enhance the electron–phonon coupling strength. More importantly, both two-photon-excited and phonon-assisted UCPL are observed in Mn-doped 2D HMHs, with an extremely large phonon-assisted energy gain (approximately 500[Formula: see text]meV). Our experimental results are of significance for designing phonon-assisted UCPL materials and expanding their applications.

Revealing Novel Mechanisms Underlying Inactivation of Tumor Suppressor Genes on Duplicated Chromosomes in Multiple Myeloma

INTRODUCTION: Multiple myeloma (MM) evolution is complex and heterogeneous. Hyperdiploidy (HRD) and translocations affecting the immunoglobulin heavy chain (IGH) locus are historically considered initiating genomic events of MM. The potential impact of genomic events acquired before known MM initiating events has never been addressed. METHODS: To investigate whether genomic deletions are acquired before and after known MM initiating events we interrogated whole-genome sequencing (WGS) data from patients with newly diagnosed MM (NDMM, n=319) and relapsed MM (RRMM, n=59). Our newly developed analytical and chronological workflow integrating single nucleotide variants (SNV), structural variants (SV), and copy number variants (CNV) comprises two main steps. First, we estimated the molecular time (i.e. corrected ratio between duplicated and non-duplicated clonal SNV) of large clonal chromosomal duplications and identified the earliest set of chromosomal gains in each patient. Second, within each early large chromosomal gain we identified clonal SV mediating CNV loss. A deletion on a gain can generate three possible scenarios: 1) one of the duplicated alleles is lost after the gain (i.e. post-gain) causing a CNV jump from 3:1 to 2:1 (total alleles : minor alleles); 2) there is a deletion before the duplication (i.e., pre-gain) causing a CNV jump from 3:1 to 1:0. 3) the deletion occurs on the minor, non-duplicated allele, causing a CNV jump from 3:1 to 2:0. Timing the deletion in relation with the chromosomal duplication is impossible in this scenario. RESULTS: Molecular time data were successfully generated for 249/319 (78%) NDMM and 51/59 (86%) RRMM patients. Restricting our analysis to NDMM with HRD without canonical IGH translocations, 16/170 (9%) of patients acquired deletions before the earliest multi-chromosomal gains, suggesting MM precursors can acquire deletions before HRD. Investigating the entire series, post-gain deletions were observed in 126/417 (30%) samples considering both early and late time windows. Leveraging a background model, we demonstrated that pre-gain events involved more tumor-suppressor genes (TSG) than expected by chance, including TCF3, ATM and TRAF3. In contrast, oncogenes were involved less than expected. In post-gain deletions, both oncogenes and TSG were involved more than expected by chance . To validate and assess the impact of deletions on driver gene expression we investigated WGS and RNAsequencing data from the MMRF CoMMpass study (n=752 NDMM). Because of the low coverage not allowing for molecular time estimation, we limited our analysis to HRD without IGH translocations. Pre-gain and post-gain deletions were observed in 47/431 (11%) and 225/431 (52%) of patients, respectively. We defined loss and gain of function events based on whether the expression level of a gene affected by the deletion was in the first or fourth quartile, respectively. Pre-gain deletions were mostly associated with downregulation of TSG expression [n=31 events in 13/431 (3%) patients]. In contrast,post-gain deletions had a more heterogenous impact with loss- and gain-of-function events. Gain-of-function events were driven by two main mechanisms: 1) the deletion joined an oncogene with a distal regulatory region inducing its overexpression [n=44 events in 6/431 (1.3%) patients]; 2) the deletion caused a new and expressed fusion [n=231 events in 109/431 (25%) patients], resulting in either loss- or gain-of-function. Surprisingly, post-gain deletions had also a major impact on TSG expression. In 60 patients (14%), we observed 146 post-gain deletions where the affected tumor suppressor gene (TSG) expression was downregulated to the level of cases with monoallelic and biallelic deletions, even though two alleles were retained. Finally, to validate these findings, we investigated 16 RRMM patients with available WGS, scATAC-seq and scRNA-seq and observed additional evidence of TSG downregulation after both pre- and post-gain deletions. CONCLUSION: Leveraging a large cohort of NDMM we show that somatic deletions can be acquired before HRD trisomies that are assumed to be initiating events. Furthermore, post-gain deletions emerged as a new mechanism inducing TSG down-regulation, despite an apparently diploid gene status.

Design of proximity‐coupled antenna array for 5G communication

SummaryMicrostrip patch antennas (MPA) are widely used in wireless communication systems due to their ease of manufacture, low cost, and low profile. This work proposes a proximity‐coupled rectangular microstrip patch antenna array (PRMPA) adaptable for 5G applications. However, the existing research works suffered from low gain, low efficiency, and narrow bandwidth. In addition, 5G communications suffer from path loss in high‐frequency bands. For managing this issue, the large gain antenna is essential. Hence, this research aims to design and fabricate high‐gain compact 2 × 2 slotted PRMPA for 5G applications. The antenna design is simulated on HFSS and operates at 28–37 GHz. Here, the antenna parameters like length, width, and feed line parameters are optimized using a metaheuristic optimization like the mayfly optimizer (MFO). Further, a circular‐shaped metamaterial split‐ring resonator (SRR) reduces the mutual coupling. Then, Rogers RT5880 is used as substrate material, and the design operates at 34.5 GHz. The gain and radiation efficiency achieved by the proposed antenna design are 15.4 dB and 94.4% for the simulated outcomes.

Single frequency pulsed Er,Yb co-doped fiber amplifier with 2 mJ of output energy

High-energy single-frequency pulsed light with low noise is generated from a master oscillator power amplifier (MOPA) system at 1.55 μm eye-safe wavelength using an ultra large mode area Er-Yb gain fiber. At a pulse repetition rate of 2 kHz, the maximum pulse energy without stimulated Brillouin scattering (SBS) reaches 2.07 mJ with 375 ns pulse duration for a pump power of 56.4 W. The pulse energy fluctuation is approximately 0.72 % within 21.5 min. The signal noise ratio (SNR) from the single-frequency fiber amplifier, at the highest pulse energy, centered at 1551.1 nm, is approximately 30 dB. The spectral half width of the signal light is calculated to be 1.28 MHz, which is close to the transform limit.

Analytical modelling of ultra‐small group delay variation of ultra‐broadband RF power amplifier using NSGA‐II algorithm

AbstractThis paper proposes a ± 9.4 ps ultra‐small group delay (GD) variation of fully integrated 65 nm complementary metal oxide semiconductor (CMOS) power amplifier (PA) over 6.5–17 GHz broadband for wireless application. The proposed CMOS PA is realised by using broadband stage, RLC inter‐stage and power stage topologies. The non‐dominated sorting genetic algorithm (NSGA‐II) is employed for PA parameter optimisation to ensure a small GD variation of ±9.4 ps over broadband with an excellent small signal gain flatness of 23.65 ± 1.85 for 6.5–17 GHz. The small GD variation of ±9.4 ps and ± 11.05 ps are attained under two cases of DC supply voltages of 2.4/1.2 V and 1.2/1.2 V, respectively. To the best of author's knowledge, the achieved GD variations are lowest among all CMOS PAs as reported so far. In addition, an analytical modelling of GD is derived to validating the minimum GD variation using zero‐pole compensation. With supply voltages of 2.4/1.2 V at 6.5 GHz, the large signal power gain, Psat and OP1dB are 26 dB, 19.3 dBm and 17.94 dBm, respectively, while peak power added efficiency (PAE) is 38.196%. At reduced supply voltages of 1.2/1.2 V, the PA achieves maximum power gain of 17.7 dB and peak PAE of 35% at 6.5 GHz. The CMOS PA occupies an area of 0.206 mm2.

X-ray timing detectors with HfO2 nanoparticle-loaded plastic scintillator and silicon avalanche photodiode

We successfully developed 20- and 40-wt% HfO2-nanoparticle-loaded plastic scintillators (Hf-PLSs), and subsequently, assembled X-ray timing scintillation detectors using the Hf-PLSs and silicon avalanche photodiodes (Si-APDs) as the photodetector. Two types of Si-APDs were examined for two different detectors. A proportional-mode APD (pAPD) exhibited a nanosecond response, but a relatively low internal gain of less than several hundred. By contrast, a Geiger-mode APD exhibited a high gain of 105−6, but its output pulse width was longer than several tens of nanoseconds; the pixelated device is called as SiPM. A pAPD (Hamamatsu Photonics S8664-4433, 3 mm in diameter) was used as the photodetector and was combined with 20 wt% Hf-PLS (∼3 mm cube) as the X-ray scintillation timing detector. A SiPM (Hamamatsu Photonics S13360-3025, 3 mm × 3 mm) was assembled with 40 wt% Hf-PLS (∼3 mm cube). The characteristics of the two types of scintillation timing detectors were evaluated using a 57.61-keV synchrotron X-ray beam to assess their applicability in nuclear resonant scattering experiments. Both the detectors were cooled down to −35 or −20 °C to reduce noise and obtain a higher APD gain. A good time resolution of (0.31 ± 0.04) ns (FWHM) was easily obtained with the SiPM detector using a large APD gain at −20 °C, although the time resolution of the pAPD detector finally reached (0.39 ± 0.04) ns (FWHM) at −35 °C owing to a small gain of 146. However, a maximum count rate of over 1.4 × 107 s−1 was observed owing to the fast response of the pAPD detector, whereas the SiPM detector exhibited a maximum count rate of 5.9 × 106 s−1 because of the long pulse tail of ∼30 ns.

The Temporal Evolution of Chromosome 1q Gain and Hyperdiploidy and Its Impact on Clinical Outcomes in Multiple Myeloma

INTRODUCTION: Multiple myeloma (MM) evolution is shaped by the acquisition and selection of distinct genomic events over time. While various temporal patterns have been identified, the precise chronological order in which key clonal genomic defining events are acquired, and their specific impact on clinical outcomes remain to be fully elucidated. METHODS: To decipher the genomic life history of MM and its impact on clinical outcomes, we interrogated the high coverage (80X) whole genome sequencing (WGS) of 378 MM patients. Overall, 319 and 59 WGS were generated from samples collected at diagnosis and at relapse, respectively. Newly diagnosed MM (NDMM) patients were treated upfront with proteosome inhibitors, lenalidomide, dexamethasone, +/- monoclonal antibodies. To estimate the relative and absolute temporal relationship among clonal events, we leveraged the molecular time approach, based on the corrected ratio between duplicated and non-duplicated single nucleotide variants (SNV). RESULTS: Overall, 300/378 (79%) patients had at least one large chromosomal gain defined as >1Mb with >50 clonal SNV eligible for molecular time analysis, for a total of 1493 events tested. Consistent with recent findings, using clock-like mutational signatures SBS1-SBS5, we estimated that the first gain in MM is typically acquired around the second or third decade of life, with a median time lag of approximately 33 years between acquisition and diagnosis. In 158/178 (89%) of hyperdiploid (HRD) patients, all trisomies of odd-numbered chromosomes were acquired simultaneously. Among patients with canonical IGH translocations and HRD, the molecular time when trisomies were acquired was usually late compared to HRD without canonical IGH translocations (p<0.0001). Next, we examined the temporal relationship between IGH canonical translocations and gains linked to the same IGH canonical translocation in 29 patients. We consistently observed that when IGH canonical translocations co-occurred with HRD (n=7), and/or 1q gain (n=8) in the same patient, the canonical IGH translocation was always found to be the initiating event, occurring earlier than the chromosomal gains. Interestingly, in 104/178 (58.4%) HRD patients, the earliest multi-gain event was not always restricted to odd chromosomes, but in involved other non-HRD chromosomes such as: 6p (n=43) and 1q (n=32), demonstrating that gain 1q, which is assumed to occur at progression, can also be acquired very early in disease life history. A total of 92 and 21 patients with 1q gain and 1q amp (>3 copies) were identified, respectively. In 14/21 (66%) with 1q amp, the first and second 1q extra copies were acquired in two distinct and independent time windows. Interestingly, patients with early 1q gain showed the same clinical outcomes as 1q amp, and inferior progression free survival outcomes compared to late 1q gain (p=0.01; Figure 1). In 18/21 1q amp the first 1q duplication was acquired early, suggesting that the adverse prognostic effect of 1q gain/amp in NDMM may be more dependent on the timing of its acquisition rather than the number of extra copies gained. To investigate late temporal patterns and differentiate selection versus acquisition of large gains at relapse after autologous stem cell transplant (ASCT), we interrogated 25/59 (42%) RRMM patients exposed to melphalan with detectable SBS-MM1. To establish the temporal order of events, we employed the melphalan mutational signature, SBS-MM1, as a unique temporal barcode (Diamond et al. Blood 2023). Overall, we observed that 21% of all loss of heterozygosity (LOH) events and 15% of all gains were acquired after ASCT (i.e., SBS-MM1 mutations were only detectable in the pre-gain/duplicated SNV). Notably, among these events, 1q gain was acquired after ASCT in 3/4 patients, indicating that the acquisition of 1q gain can occur even after administration of first-line therapy. CONCLUSIONS: In our study, we have uncovered distinct temporal patterns in the acquisition of early genomic drivers in MM. These patterns contribute to the reconstruction and better understanding of the life history of MM and may hold potential clinical and prognostic implications.

On-chip Er-doped Ta2O5 waveguide amplifiers with a high internal net gain.

We designed and fabricated a double-layered structure Er3+:Ta2O5 waveguide and investigated its optical amplification performance in C band. The pump laser threshold for zero gain at 1533 nm was 2.5 mW, and the internal net gain was ∼4.63 dB/cm for a lunched pump power of 36.1 mW at 980 nm and signal input power of -30.0 dBm (1 µW). The relationship between the internal gain and the signal input power was also investigated, and a large internal net gain of 10.58 dB/cm was achieved at a signal input power of ∼-47.1 dBm. The results confirm the potentials of the use of Ta2O5 as a host material for optical waveguide amplification.

A low-power and high-gain frontend for GHz application using trans-impedance amplifier for fast particle detection

In this paper, we present a front-end for gigahertz applications, FGATI, which utilizes a transimpedance amplifier design to amplify the current signal from a diamond particle detector. The transimpedance amplifier design adopts a flipped-voltage-follower-based current-mirror (FVF-CM) topology as the input stage, offering advantages such as low power consumption, large transimpedance gain, gigahertz bandwidth, and reasonable noise levels. The FVF-CM topology was realized to improve noise reduction with a fully differential output configuration. The design was implemented as an ASIC chip using 65 nm CMOS silicon technology. The bandwidth measurement of the FGATI prototype demonstrated a 3-dB bandwidth of 1.2 GHz. Furthermore, the amplifier's power consumption is low, drawing only 7.2 mW/channel from a 1.2 V power supply, including the buffer stage. The measurement of the FGATI output signal indicated an excellent transimpedance gain of 79.2 dBΩ and a noise level of 6.7 mVrms. These findings highlight the feasibility and effectiveness of the proposed front-end design in high-frequency applications.

MADAv2: Advanced Multi-Anchor Based Active Domain Adaptation Segmentation.

Unsupervised domain adaption has been widely adopted in tasks with scarce annotated data. Unfortunately, mapping the target-domain distribution to the source-domain unconditionally may distort the essential structural information of the target-domain data, leading to inferior performance. To address this issue, we firstly propose to introduce active sample selection to assist domain adaptation regarding the semantic segmentation task. By innovatively adopting multiple anchors instead of a single centroid, both source and target domains can be better characterized as multimodal distributions, in which way more complementary and informative samples are selected from the target domain. With only a little workload to manually annotate these active samples, the distortion of the target-domain distribution can be effectively alleviated, achieving a large performance gain. In addition, a powerful semi-supervised domain adaptation strategy is proposed to alleviate the long-tail distribution problem and further improve the segmentation performance. Extensive experiments are conducted on public datasets, and the results demonstrate that the proposed approach outperforms state-of-the-art methods by large margins and achieves similar performance to the fully-supervised upperbound, i.e., 71.4% mIoU on GTA5 and 71.8% mIoU on SYNTHIA. The effectiveness of each component is also verified by thorough ablation studies. Code is available at https://github.com/munanning/MADAv2.

Design of an Ultra-Wideband High-Efficiency Power Amplifier Based on a Novel Impedance Matching Structure With Three Paths

In this paper, a design method for a high-efficiency power amplifier (PA) with a three-octave operation bandwidth is proposed. A unique impedance matching structure with three paths (TPs) and a series of inverse continuous modes (SICMs) are combined to achieve outstanding performance in terms of operating bandwidth as well as efficiency. Especially, a buffer stage of impedance matching is generated by the proposed TPs structure, which further promotes the expansion of target bandwidth. To verify the proposed methodology, an ultra-wideband PA component is implemented with a 10-W GaN HEMT device. The measurement results demonstrate a fractional bandwidth of 155.6 % from 0.5 to 4.0 GHz, with saturated output power of 39.5–42.2 dBm, drain efficiency (DE) of 57 –79 %, and a large signal gain of 9.5–12.2 dB. These results show that the PA realized by the proposed TPs impedance matching design is suitable for small-base station and multiband applications.

Optimization of sub-relativistic co-propagating accelerating structures.

In this paper, novel optimization methodologies of sub-relativistic guided interaction structures for dielectric laser particle acceleration (DLA) are presented. In particular, we focus on co-propagating geometries based on slot waveguides in continuous wave (CW) operation, where the particle flow and the direction of propagation of the accelerating field are co-linear. Since the velocity of sub-relativistic particles varies along the acceleration path, proper tapering of the waveguide geometry is required to achieve an extended acceleration region, and, thus a large energy gain. The design of an optimal taper ensuring particle-wave synchronicity and maximum energy gain is pursued through a physics-based approach, and these results are compared, for validation, with the outcomes of a downhill simplex method searching algorithm. Additionally, the application of a simplified 2D model of the accelerating slot waveguide is investigated and profitably used to get qualitative results useful for fast structure optimization. Indeed, this approach can hold significant potential for the development of novel accelerating structures, as it enables a thorough and fast exploration of the design space.

Distributed observer for LTI systems under stochastic impulsive sequential attacks

In this paper, we consider the distributed state estimation problem of continuous-time linear time-invariant (LTI) systems, where communication channels are attacked by stochastic impulsive sequences. In a networked system, the communication channel between an agent and its neighbor is vulnerable to being attacked. Continuous attacks have been widely studied so far, but with the development of attack techniques, intermittent attacks have been emerging, which are more difficult to be identified and detected, and have not been well portrayed and studied. Therefore, this paper investigates a random intermittent injection attack, which leads to abrupt and random changes of the state information transmitted by the neighbors. First, this attack is modeled as a stochastic impulsive sequential attack, where both its impulsive time and impulsive intensity are random. Then a distributed observer under stochastic impulsive sequential attack is proposed. Since the impulsive attack destroys the stability of the system, in order to attenuate the impact of the impulsive attack, a large feedback gain is necessary. By using the graph theory, matrix theory and stochastic analysis, sufficient conditions for the existence of the distributed observer in the sense of mean square and in the sense of almost sure are presented, respectively. Finally, the effectiveness of the proposed distributed observer is verified by numerical simulations.

Nanoscatterer-Assisted Fluorescence Amplification Technique.

Weak fluorescence signals, which are important in research and applications, are often masked by the background. Different amplification techniques are actively investigated. Here, a broadband, geometry-independent and flexible feedback scheme based on the random scattering of dielectric nanoparticles allows the amplification of a fluorescence signal by partial trapping of the radiation within the sample volume. Amplification of up to a factor of 40 is experimentally demonstrated in ultrapure water with dispersed TiO2 nanoparticles (30 to 50 nm in diameter) and fluorescein dye at 200 μmol concentration (pumped with 5 ns long, 3 mJ laser pulses at 490 nm). The measurements show a measurable reduction in linewidth at the emission peak, indicating that feedback-induced stimulated emission contributes to the large gain observed.

The Wear Behavior of Glass-Ceramic CAD/CAM Blocks against Bovine Enamel

The wear of enamel and crown restorative materials often occur by occlusion. The purpose of this study was to evaluate the wear volume between glass-ceramics used for CAD/CAM blocks (lithium disilicate: Initial LiSi block (LIS), IPS e.max CAD (IPS), zirconia-reinforced lithium silicate glass-ceramics: Celtra DUO (DUO), VITA Suprinity (VITS) and feldspar-based glass-ceramics: Vitablocs Mark II (MAK)) and bovine tooth enamel using a two-body wear test, the hardness, three-point bending strength, micro-structure and the element components of glass-ceramics. The data were analyzed using a one-way analysis of variance and Tukey’s multiple comparison test (α = 0.05). IPS and DUO with relatively large size crystal gain had significantly larger abrader wear volumes. Zirconia-reinforced lithium silicate glass-ceramics (DUO, VITS) caused significantly greater wear volume in antagonist enamel. MAK with scale-shape crystals grains produced distinct scratches after wear tests, both in the material itself and in the enamel. A strong correlation between the mechanical properties (hardness, three-point bending strength) and wear volume could not be confirmed. The type of glass-ceramic, size, and shape of the crystal grains affected the wear behavior of the glass-ceramics for CAD/CAM blocks. Therefore, dentists should consider that wear behavior varies with crystal structure, size, and shape in glass-ceramics for CAD/CAM blocks.

Genomic prediction based on selective linkage disequilibrium pruning of low-coverage whole-genome sequence variants in a pure Duroc population

BackgroundAlthough the accumulation of whole-genome sequencing (WGS) data has accelerated the identification of mutations underlying complex traits, its impact on the accuracy of genomic predictions is limited. Reliable genotyping data and pre-selected beneficial loci can be used to improve prediction accuracy. Previously, we reported a low-coverage sequencing genotyping method that yielded 11.3 million highly accurate single-nucleotide polymorphisms (SNPs) in pigs. Here, we introduce a method termed selective linkage disequilibrium pruning (SLDP), which refines the set of SNPs that show a large gain during prediction of complex traits using whole-genome SNP data.ResultsWe used the SLDP method to identify and select markers among millions of SNPs based on genome-wide association study (GWAS) prior information. We evaluated the performance of SLDP with respect to three real traits and six simulated traits with varying genetic architectures using two representative models (genomic best linear unbiased prediction and BayesR) on samples from 3579 Duroc boars. SLDP was determined by testing 180 combinations of two core parameters (GWAS P-value thresholds and linkage disequilibrium r2). The parameters for each trait were optimized in the training population by five fold cross-validation and then tested in the validation population. Similar to previous GWAS prior-based methods, the performance of SLDP was mainly affected by the genetic architecture of the traits analyzed. Specifically, SLDP performed better for traits controlled by major quantitative trait loci (QTL) or a small number of quantitative trait nucleotides (QTN). Compared with two commercial SNP chips, genotyping-by-sequencing data, and an unselected whole-genome SNP panel, the SLDP strategy led to significant improvements in prediction accuracy, which ranged from 0.84 to 3.22% for real traits controlled by major or moderate QTL and from 1.23 to 11.47% for simulated traits controlled by a small number of QTN.ConclusionsThe SLDP marker selection method can be incorporated into mainstream prediction models to yield accuracy improvements for traits with a relatively simple genetic architecture, however, it has no significant advantage for traits not controlled by major QTL. The main factors that affect its performance are the genetic architecture of traits and the reliability of GWAS prior information. Our findings can facilitate the application of WGS-based genomic selection.

Derivation of Ultra-High Gain Hybrid Converter Families for HASEL Actuators Used in Soft Mobile Robots.

This work proposes, analyzes, designs, and validates superior topologies of UHGH converters that are capable of supporting extremely large conversion ratios up to ∼2000× and output voltage up to ∼4-12 kV for future mobile soft robots from an input voltage as low as the range of a 1-cell battery pack. Thus, the converter makes soft robots standalone systems that can be untethered and mobile. The extremely large voltage gain is enabled by a unique hybrid combination of a high-gain switched magnetic element (HGSME) and a capacitor-based voltage multiplier rectifier (CVMR) that, together, achieve small overall size, efficient operation, and output voltage regulation and shaping with simple duty-cycle modulation. With superior performance, power density, and compact size, the UHGH converters prove to be a promising candidate for future untethered soft robots.

Optical gain and lasing from bulk cadmium sulfide nanocrystals through bandgap renormalization.

Strongly confined colloidal quantum dots have been investigated for low-cost light emission and lasing for nearly two decades. However, known materials struggle to combine technologically relevant metrics of low-threshold and long inverted-state lifetime with a material gain coefficient fit to match cavity losses, particularly under electrical excitation. Here we show that bulk nanocrystals of CdS combine an exceptionally large material gain of 50,000 cm-1 with best-in-class gain thresholds below a single exciton per nanocrystal and 3 ns gain lifetimes not limited by non-radiative Auger processes. We quantitatively account for these findings by invoking a strong bandgap renormalization effect, unobserved in nanocrystals to date, to the best of our knowledge. Next, we demonstrate broadband amplified spontaneous emission and lasing under quasi-continuous-wave conditions. Our results highlight the prospects of bulk nanocrystals for lasing from solution-processable materials.

DeepPatch: Maintaining Deep Learning Model Programs to Retain Standard Accuracy with Substantial Robustness Improvement

Maintaining a deep learning (DL) model by making the model substantially more robust through retraining with plenty of adversarial examples of non-trivial perturbation strength often reduces the model’s standard accuracy. Many existing model repair or maintenance techniques sacrifice standard accuracy to produce a large gain in robustness or vice versa. This article proposes DeepPatch, a novel technique to maintain filter-intensive DL models. To the best of our knowledge, DeepPatch is the first work to address the challenge of standard accuracy retention while substantially improving the robustness of DL models with plenty of adversarial examples of non-trivial and diverse perturbation strengths. Rather than following the conventional wisdom to generalize all the components of a DL model over the union set of clean and adversarial samples, DeepPatch formulates a novel division of labor method to adaptively activate a subset of its inserted processing units to process individual samples. Its produced model can generate the original or replacement feature maps in each forward pass of the patched model, making the patched model carry an intrinsic property of behaving like the model under maintenance on demand. The overall experimental results show that DeepPatch successfully retains the standard accuracy of all pretrained models while improving the robustness accuracy substantially. However, the models produced by the peer techniques suffer from either large standard accuracy loss or small robustness improvement compared with the models under maintenance, rendering them unsuitable in general to replace the latter.