Study of the Protection and Energy Transmission Modes of One Phase Short Circuit to Ground in Inverters.

Research indicates that phase-to-ground short-circuits in a frequency converter can subject the rectifier diode and IGBT to excessive voltage and current, potentially causing damage if the component selection margin during hardware design is insufficient. In order to solve the above problems, this paper studies the design of the LCL filter and ground short circuit problem of the hundred-kilowatt inverter. Firstly, an analytical method for calculating the DC bus capacitance and reactor of the inverter is proposed. The interaction between the DC bus capacitance and the reactor parameters and performance is considered in the implementation process. The parameters of the DC bus capacitor and reactor are given. Secondly, the one-to-ground short circuit of the inverter is studied, and the energy flow mode and mathematical expression of the double boost circuit, considering the influence of the leakage inductance of the power transformer, are given. Based on the above analysis, a method for determining the rectifier diode and IGBT, considering the one-to-ground short circuit of the inverter, is proposed. Finally, a one-hundred-kilowatt inverter is developed, and the corresponding experiments are carried out. The feasibility of the proposed scheme is verified by simulation and experiment.

DIML: Deep Interpretable Metric Learning via Structural Matching.

In this paper, we present a new framework named DIML to achieve more interpretable deep metric learning. Unlike traditional deep metric learning method that simply produces a global similarity given two images, DIML computes the overall similarity through the weighted sum of multiple local part-wise similarities, making it easier for human to understand the mechanism of how the model distinguish two images. Specifically, we propose a structural matching strategy that explicitly aligns the spatial embeddings by computing an optimal matching flow between feature maps of the two images. We also devise a multi-scale matching strategy, which considers both global and local similarities and can significantly reduce the computational costs in the application of image retrieval. To handle the view variance in some complicated scenarios, we propose to use cross-correlation as the marginal distribution of the optimal transport to leverage semantic information to locate the important region in the images. Our framework is model-agnostic, which can be applied to off-the-shelf backbone networks and metric learning methods. To extend our DIML to more advanced architectures like vision Transformers (ViTs), we further propose truncated attention rollout and partial similarity to overcome the lack of locality in ViTs. We evaluate our method on three major benchmarks of deep metric learning including CUB200-2011, Cars196, and Stanford Online Products, and achieve substantial improvements over popular metric learning methods with better interpretability.

Rapid high-throughput genotyping of HBV DNA using a modified hybridization-extension technique.

China has the highest incidence of hepatitis B virus (HBV) infection worldwide. HBV genotypes have variable impacts on disease pathogenesis and drug tolerance. We have developed a technically simple and accurate method for HBV genotyping that will be applicable to pre-treatment diagnosis and individualized treatment. Multiple sequence alignments of HBV genomes from GenBank were used to design primers and probes for genotyping of HBV A through H. The hybridization was carried out on nitrocellulose (NC) membranes with probes fixed in an array format, which was followed by hybrid amplification by an extension step with DNA polymerase to reinforce the double-stranded DNA hybrids on the NC membrane and subsequent visualization using an avidin-biotin system. Genotyping results were confirmed by DNA sequencing and bioinformatics analysis using the National Center for Biotechnology Information genotyping database, and compared with results from the line probe assay. The data show that multiple sequence alignment defined a 630 bp region in the HBV PreS and S regions that was suitable for genotyping. All genotyping significant single nucleotides in the region were defined. Two-hundred-and-ninety-one HBV-positive serum samples from Northwest Chinese patients were genotyped, and the genotyping rate from the new modified hybridization-extension method was 100% compared with direct sequencing. Compared with line probe assay, the newly developed method is superior, featuring reduced reaction time, lower risk of contamination, and increased accuracy for detecting single nucleotide mutation. In conclusion, a novel hybridization-extension method for HBV genotyping was established, which represents a new tool for accurate and rapid SNP detection that will benefit clinical testing.

GFNet: Global Filter Networks for Visual Recognition.

Recent advances in self-attention and pure multi-layer perceptrons (MLP) models for vision have shown great potential in achieving promising performance with fewer inductive biases. These models are generally based on learning interaction among spatial locations from raw data. The complexity of self-attention and MLP grows quadratically as the image size increases, which makes these models hard to scale up when high-resolution features are required. In this paper, we present the Global Filter Network (GFNet), a conceptually simple yet computationally efficient architecture, that learns long-term spatial dependencies in the frequency domain with log-linear complexity. Our architecture replaces the self-attention layer in vision Transformers with three key operations: a 2D discrete Fourier transform, an element-wise multiplication between frequency-domain features and learnable global filters, and a 2D inverse Fourier transform. Based on this basic design, we develop a series of isotropic models with a Transformer-style simple architecture and CNN-style hierarchical models with better performance. Isotropic GFNet models exhibit favorable accuracy/complexity trade-offs compared to recent vision Transformers and pure MLP models. Hierarchical GFNet models can inherit successful designs in CNNs and be easily scaled up with larger model sizes and more training data, showing strong performance on both image classification (e.g., 85.0% top-1 accuracy on ImageNet-1 k without any extra data or supervision, and 87.4% accuracy with ImageNet-21 k pre-training) and dense prediction tasks (e.g., 54.3 mIoU on ADE20 k val). Our results demonstrate that GFNet can be a very competitive alternative to Transformer-based models and CNNs in terms of efficiency, generalization ability and robustness. Code is available at https://github.com/raoyongming/GFNet.

Dynamic Spatial Sparsification for Efficient Vision Transformers and Convolutional Neural Networks.

In this paper, we present a new approach for model acceleration by exploiting spatial sparsity in visual data. We observe that the final prediction in vision Transformers is only based on a subset of the most informative regions, which is sufficient for accurate image recognition. Based on this observation, we propose a dynamic token sparsification framework to prune redundant tokens progressively and dynamically based on the input to accelerate vision Transformers. Specifically, we devise a lightweight prediction module to estimate the importance of each token given the current features. The module is added to different layers to prune redundant tokens hierarchically. While the framework is inspired by our observation of the sparse attention in vision Transformers, we find that the idea of adaptive and asymmetric computation can be a general solution for accelerating various architectures. We extend our method to hierarchical models including CNNs and hierarchical vision Transformers as well as more complex dense prediction tasks. To handle structured feature maps, we formulate a generic dynamic spatial sparsification framework with progressive sparsification and asymmetric computation for different spatial locations. By applying lightweight fast paths to less informative features and expressive slow paths to important locations, we can maintain the complete structure of feature maps while significantly reducing the overall computations. Extensive experiments on diverse modern architectures and different visual tasks demonstrate the effectiveness of our proposed framework. By hierarchically pruning 66% of the input tokens, our method greatly reduces 31%  âˆ¼  35% FLOPs and improves the throughput by over 40% while the drop of accuracy is within 0.5% for various vision Transformers. By introducing asymmetric computation, a similar acceleration can be achieved on modern CNNs and Swin Transformers. Moreover, our method achieves promising results on more complex tasks including semantic segmentation and object detection. Our results clearly demonstrate that dynamic spatial sparsification offers a new and more effective dimension for model acceleration. Code is available at https://github.com/raoyongming/DynamicViT.

Panoramic analysis of coronaviruses carried by representative bat species in Southern China to better understand the coronavirus sphere.

Bats, recognized as considerable reservoirs for coronaviruses (CoVs), serve as natural hosts for several highly pathogenic CoVs, including SARS-CoV and SARS-CoV-2. Investigating the bat CoV community provides insights into the origin for highly pathogenic CoVs and highlights bat CoVs with potential spillover risks. This study probes the evolution, recombination, host range, geographical distribution, and cross-species transmission characteristics of bat CoVs across China and its associated CoVs in other regions. Through detailed research on 13,064 bat samples from 14 provinces of China, 1141 CoV strains are found across 10 subgenera and one unclassified Alpha-CoV, generating 399 complete genome sequences. Within bat CoVs, 11 new CoV species are identified and 425 recombination events are detected. Bats in southern China, particularly in Yunnan province, exhibit a pronounced diversity of CoVs. Limited sampling and low detection rates exist for CoVs in Myotacovirus, Nyctacovirus, Hibecovirus, Nobecovirus in China. The genus Myotis is highlighted as a potential ancestral host for Alpha-CoV, with the genus Hipposideros suggested as a likely progenitor host for bat-associated Beta-CoV, indicating the complexity of cross-species transmission dynamics. Through the comprehensive analysis, this study enriches the understanding of bat CoVs and offers a valuable resource for future research.

A comprehensive survey of bat sarbecoviruses across China in relation to the origins of SARS-CoV and SARS-CoV-2.

SARS-CoV and SARS-CoV-2 have been thought to originate from bats. In this study, we screened pharyngeal and anal swabs from 13 064 bats collected between 2016 and 2021 at 703 locations across China for sarbecoviruses, covering almost all known southern hotspots, and found 146 new bat sarbecoviruses. Phylogenetic analyses of all available sarbecoviruses show that there are three different lineages-L1 as SARS-CoV-related CoVs (SARSr-CoVs), L2 as SARS-CoV-2-related CoVs (SC2r-CoVs) and novel L-R (recombinants of L1 and L2)-present in Rhinolophus pusillus bats, in the mainland of China. Among the 146 sequences, only four are L-Rs. Importantly, none belong in the L2 lineage, indicating that circulation of SC2r-CoVs in China might be very limited. All remaining 142 sequences belong in the L1 lineage, of which YN2020B-G shares the highest overall sequence identity with SARS-CoV (95.8%). The observation suggests endemic circulations of SARSr-CoVs, but not SC2r-CoVs, in bats in China. Geographic analysis of the collection sites in this study, together with all published reports, indicates that SC2r-CoVs may be mainly present in bats of Southeast Asia, including the southern border of Yunnan province, but absent in all other regions within China. In contrast, SARSr-CoVs appear to have broader geographic distribution, with the highest genetic diversity and sequence identity to human sarbecoviruses along the southwest border of China. Our data provide the rationale for further extensive surveys in broader geographical regions within, and beyond, Southeast Asia in order to find the most recent ancestors of human sarbecoviruses.

VideoABC: A Real-World Video Dataset for Abductive Visual Reasoning.

In this paper, we investigate the problem of abductive visual reasoning (AVR), which requires vision systems to infer the most plausible explanation for visual observations. Unlike previous work which performs visual reasoning on static images or synthesized scenes, we exploit long-term reasoning from instructional videos that contain a wealth of detailed information about the physical world. We conceptualize two tasks for this emerging and challenging topic. The primary task is AVR, which is based on the initial configuration and desired goal from an instructional video, and the model is expected to figure out what is the most plausible sequence of steps to achieve the goal. In order to avoid trivial solutions based on appearance information rather than reasoning, the second task called AVR++ is constructed, which requires the model to answer why the unselected options are less plausible. We introduce a new dataset called VideoABC, which consists of 46,354 unique steps derived from 11,827 instructional videos, formulated as 13,526 abductive reasoning questions with an average reasoning duration of 51 seconds. Through an adversarial hard hypothesis mining algorithm, non-trivial and high-quality problems are generated efficiently and effectively. To achieve human-level reasoning, we propose a Hierarchical Dual Reasoning Network (HDRNet) to capture the long-term dependencies among steps and observations. We establish a benchmark for abductive visual reasoning, and our method set state-of-the-arts on AVR (  âˆ¼ 74 %) and AVR++ (  âˆ¼ 45 %), and humans can easily achieve over 90% accuracy on these two tasks. The large performance gap reveals the limitation of current video understanding models on temporal reasoning and leaves substantial room for future research on this challenging problem. Our dataset and code are available at https://github.com/wl-zhao/VideoABC.

How Does Topography Affect the Value of Ecosystem Services? An Empirical Study from the Qihe Watershed.

Topographic position indices (TPIs) measure essential impacts on ecosystem service supply capacity. The identification of changes in ecosystem services and value metrics under varying TPIs has become a topical subject of global change research. Multidimensional changes in spatiotemporal and geographical aspects of ecosystem service values (ESVs) are assessed in this article using land cover/use data from 2000-2015. Effects of land-use/cover changes and topographic indices on ESVs are explored using the Chinese terrestrial unit area ecosystem service value equivalence table combined with topographic factors. A sensitivity index is introduced to quantify the robustness of total ESV to land-use/cover and topographic indices. The results show that: (1) The total ESV in the Qihe watershed declined with a change in land-use/cover during the period 2000-2015. The maximum ESV was CNY 1.984 billion in 2005 and the minimum was CNY 1.940 billion in 2010; (2) The response of ESV to land/use cover varied greatly across TPIs, with the most significant change in ESV occurring in the 0.6-0.8 TPI range and the greatest change in a single ecosystem service occurred in water areas; (3) The sensitivity indices of ESVs are all less than 1. The sensitivity indices of unused land and water tended to zero. Woodland sensitivity indices were the highest at 0.53, followed by those of arable land and grassland, owing to the large proportion of arable land and grassland areas in the overall area of land-use categories.

A mutation in PHKG1 causes high drip loss and low meat quality in Chinese Ningdu yellow chickens.

With increasing societal development and the concurrent improvement in people's quality of life, meat consumption has gradually changed from a focus on "quantity" to "quality". Broiler production is increasingly used as a means to improve meat quality by altering various characteristics, especially its genetic factors. However, until now, little has been known about the genetic variants related to meat quality traits in Chinese purebred chicken populations. To better understand these genetic underpinnings, a total of 17 traits related to meat quality and carcass were measured in 325 Chinese Ningdu yellow chickens. We performed DNA sequencing to detect nucleotide mutations, after which we conducted association studies between PHKG1 gene polymorphisms and traits related to meat quality and carcass. Results indicated a large phenotypic variation in meat quality traits. More specifically, the single nucleotide polymorphism (SNP) rs15845448 was significantly associated with drip loss at 24 h (P = 8.04 × 10-6) and 48 h (P = 5.47 × 10-6), pH (P = 2.39 × 10-3), and meat color L* (P = 9.88 × 10-3). Moreover, the SNP rs15845448 reduced 24 h and 48 h drip loss by 3.62 and 5.97%, respectively. However, no significant associations were found between rs15845448 and carcass traits (P > 0.05). Furthermore, a haplotype block containing 2 adjacent SNPs (rs15845448 and rs15845450) was identified. This block displayed 4 distinct haplotypes that had significant association with drip loss at 24 h and 48 h, pH, and meat color L*. Collectively, these results provide new insights into the genetic basis of meat quality in Chinese Ningdu yellow chickens. Moreover, the significance of SNP rs15845448 could be incorporated into the selection programs involving this breed.

Adaptive Evolution of Feline Coronavirus Genes Based on Selection Analysis.

PURPOSE: We investigated sequences of the feline coronaviruses (FCoV), which include feline enteric coronavirus (FECV) and feline infectious peritonitis virus (FIPV), from China and other countries to gain insight into the adaptive evolution of this virus. METHODS: Ascites samples from 31 cats with suspected FIP and feces samples from 8 healthy cats were screened for the presence of FCoV. Partial viral genome sequences, including parts of the nsp12-nsp14, S, N, and 7b genes, were obtained and aligned with additional sequences obtained from the GenBank database. Bayesian phylogenetic analysis was conducted, and the possibility of recombination within these sequences was assessed. Analysis of the levels of selection pressure experienced by these sequences was assessed using methods on both the PAML and Datamonkey platforms. RESULTS: Of the 31 cats investigated, two suspected FIP cats and one healthy cat tested positive for FCoV. Phylogenetic analysis showed that all of the sequences from mainland China cluster together with a few sequences from the Netherlands as a distinct clade when analyzed with FCoV sequences from other countries. Fewer than 3 recombination breakpoints were detected in the nsp12-nsp14, S, N, and 7b genes, suggesting that analyses for positive selection could be conducted. A total of 4, 12, 4, and 4 positively selected sites were detected in the nsp12-nsp14, S, N, and 7b genes, respectively, with the previously described site 245 of the S gene, which distinguishes FIPV from FECV, being a positive selection site. Conversely, 106, 168, 25, and 17 negative selection sites in the nsp12-14, S, N, and 7b genes, respectively, were identified. CONCLUSION: Our study provides evidence that the FCoV genes encoding replicative, entry, and virulence proteins potentially experienced adaptive evolution. A greater number of sites in each gene experienced negative rather than positive selection, which suggests that most of the protein sequence must be conservatively maintained for virus survival. A few of the sites showing evidence of positive selection might be associated with the more severe pathology of FIPV or help these viruses survive other harmful conditions.

Reduced Cd, Pb, and As accumulation in rice (Oryza sativa L.) by a combined amendment of calcium sulfate and ferric oxide.

A combined amendment (CF) consisting of 90% calcium sulfate (CaSO4) and 10% ferric oxide (Fe2O3) was used to investigate the feasibility, active principles, and possible mechanisms of the immobilization of heavy metals in paddy soil. A soil incubation experiment, two consecutive pot trials, and a field experiment were conducted to evaluate the effectiveness and persistence of CF on metal(loid) immobilization. Soil incubation experiment results indicated that the application of CF significantly decreased the concentrations of cadmium (Cd), lead (Pb), and arsenic (As) in soil solution. CF treatments simultaneously reduced the accumulation of Cd, Pb, and As in two consecutive pot trials. The total Cd, Pb, and As concentrations in the rice grains were respectively 0.02, 2.08, and 0.62 mg kg-1 in the control treatment in the second year, which exceeded the safety limits of contaminants in food products in China. However, a high amount of CF amendment (CF-H, 0.3%) effectively decreased Cd, Pb, and As by 75.0%, 75.5%, and 46.8%, respectively. Further, with the CF amendment, the bioavailable Cd and Pb in the soil and the accumulation of Cd, Pb, and As in rice grain in the field experiment were also significantly decreased. The concentrations of Cd, Pb, and As in grains were respectively 0.02, 0.03, and 0.39 mg kg-1 in the control treatment in the field experiment, which decreased to 0.01, 0.01, and 0.22 mg kg-1 with CF addition, suggesting that grains produced in the field could pose less health risk. In conclusion, these results implied that CF was an effective and persistent combined amendment to immobilize heavy metals in soil and thereby can reduce the exposure risk of metal(loid)s associated with rice consumption.

Simultaneous immobilization of the cadmium, lead and arsenic in paddy soils amended with titanium gypsum.

In situ immobilization of heavy metals in contaminated soils using industrial by-products is an attractive remediation technique. In this work, titanium gypsum (TG) was applied at two levels (TG-L: 0.15% and TG-H: 0.30%) to simultaneously reduce the uptake of cadmium (Cd), lead (Pb) and arsenic (As) in rice grown in heavy metal contaminated paddy soils. The results showed that the addition of TG significantly decreased the pH and dissolved organic carbon (DOC) in the bulk soil. TG addition significantly improved the rice plants growth and reduced the bioavailability of Cd, Pb and As. Particularly, bioavailable Cd, Pb and As decreased by 35.2%, 38.1% and 38.0% in TG-H treatment during the tillering stage, respectively. Moreover, TG application significantly reduced the accumulation of Cd, Pb and As in brown rice. Real-time PCR analysis demonstrated that the relative abundance of sulfate-reducing bacteria increased with the TG application, but not for the iron-reducing bacteria. In addition, 16S rRNA sequencing analysis revealed that the relative abundances of heavy metal-resistant bacteria such as Bacillus, Sulfuritalea, Clostridium, Sulfuricella, Geobacter, Nocardioides and Sulfuricurvum at the genus level significantly increased with the TG addition. In conclusion, the present study implied that TG is a potential and effective amendment to immobilize metal(loid)s in soil and thereby reduce the exposure risk of metal(loid)s associated with rice consumption.

Genotypes and Hot Spot Mutations of Hepatitis B Virus in Northwest Chinese Population and Its Correlation with Diseases Progression.

Hepatitis B virus (HBV) infection is a critical incentive for chronic hepatitis B (CHB), liver cirrhosis (LC), and hepatocellular carcinoma (HCC). Different genotypes and genome mutations of HBV have been found to be related to the progression of these liver diseases. However, their clinical significance is still under debate. The objective of this study was to determine the association of HBV genotypes and hot spot mutations in the reverse transcriptase (RT) and basal core promoter-precore (BCP-PreC) region with HBV-infected diseases in a northwest Chinese population. HBV genotyping and DNA sequencing were performed in samples of 980 patients. Appropriate statistical methods were adopted to assess HBV genetic features and its clinical association. It was found that the prevalent HBV genotype in northwestern Chinese patients was HBV/C (61.33%), followed by HBV/B (36.63%). In RT region, in addition to the reported nucleoside analogue- (NA-) resistance missense mutations, new silent mutations at rt169 and rt180 were found to raise the risk of HCC in patients with HBV/C. And the heterozygous mutation status of rt169/rt180 was associated with the increased risk of both HCC and NA resistance (OR > 1, P < 0.01) regardless of HBV genotypes. In BCP-PreC region, multiple mutations and combinations, especially at nt 1762/1764 and nt 1896/1899, were characterized to be the causes of spurious HBeAg negativity and liver function injury, as well as the risk factors for HCC progression (P < 0.01). Additionally, a novel mutation at nt1799G>C was likely found to increase the risk of HCC in patients with HBV/B. These findings revealed an association between HBV genotypes and HBV genetic mutations in RT and BCP-PreC region and progression of hepatitis B. It would be helpful for risk evaluation and diagnostic improvement based on these genetic features.

CT diagnosis for metastasis of clear cell renal cell carcinoma to the pancreas: Three case reports.

RATIONALE: Clear cell renal cell carcinoma (CCRCC) metastasis to pancreas is clinically rare. Misdiagnosis for these cases is frequently due to the low incidence, lack of specific clinical symptoms, and laboratory results. PATIENT CONCERNS: Three female patients aged 47 years, 69 years, and 76 years, respectively, were admitted to hospital for routine examination after resection of clear cell carcinoma of kidney for 69 months, 57 months, and 123 months, respectively. All 3 cases had no specific clinical symptoms. Routine laboratory tests and common tumor markers including CEA, AFP, CA19-9, and CA125 showed no obvious abnormality. DIAGNOSIS: All 3 cases were finally diagnosed with CCRCC metastasis to pancreas on the basis of CT and pathological findings. On unenhanced CT, foci of the pancreas showed single or multiple nodules or masses with mildly low or equal density and obscure boundary. On enhanced CT, the enhanced mode of foci was similar to CCRCC and showed "fast in fast out." The main body was confined in the pancreas. The peripheral structure was clear relatively. Obstruction of common bile duct, main pancreatic duct, and local infringement of foci cannot be seen. Additional metastases of right adrenal gland can be seen in one case. INTERVENTIONS: All 3 cases underwent CT examination and surgical treatment, with complete removal of metastatic tumors. OUTCOMES: All 3 cases underwent surgical treatment successfully, and recovered successfully after operation. LESSONS: The manifestations of pancreatic metastases from CCRCC on CT show certain characteristics, which may be useful to assess the histological features of pancreatic metastases from CCRCC and facilitate the preoperative diagnosis.

Analyses of the Genetic Diversities and Mutations of the Hepatitis B Virus Genome BCP/Pre C Region.

We wished to study the genetic diversities and mutations of the basic core promoter (BCP) and pre-C region of the hepatitis B virus (HBV) in liver-cancer tissues. One hundred and ninety-two tissue samples were collected from patients suffering from hepatocellular carcinoma (HCC) and HBV infection in 2015 in Xijing Hospital. Twenty-one cases were selected, of which direct sequencing of the polymerase chain reaction (PCR)-amplified products of BCP/pre-C region was unsuccessful. Cloning and sequencing allowed the DNA sequences of the BCP/pre-C region to be analyzed. Sequencing showed infection with mixed mutants of HBV in 37. 89% of HBV-positive HCC patients, and that HBV DNA in each sample contained 2 ~ 11 types of mutations.. The mutation rate of deletion and insertion was 80. 95%. Other mutations in descending order by mutation rate were a: A1762T/G1764A combined mutation (90. 48%); G1756C/T1803A/A(1757 ~ 1765)/A (1824 ~ 1832) combined mutation (80. 95%); T1753C/A1762T/ G1764A combined mutation (57. 14%); A1762T/G1764A/G1896A combined mutation (42. 86%); G1756C/Δ(1757~176.5) combined mutation,(28. 57%); T1753C/A1762T/G1764A/G1896A combined mutation (23. 81%). The sequencing failure of PCR products may have been correlated directly with the deletion and insertion mutations of HBV DNA. These findings lay the foundation for further studies on HBV mutations, persistent infection, and the mechanism of personalized medicine.