Search

Isothermal oxidation behaviour of two nickel base superalloys with different Cr contents was investigated in the temperature range of 700–800 °C. Weight–change curves show that weight loss increases with increasing temperature and a relatively lower Cr ratio corresponds to slow weight gain rate. Microscopic observations uncover that oxidation penetration attacks the alloys primarily along grain boundary especially for the lower-Cr alloy, and the penetration depth is strongly dependent on the oxidation temperature. The underlying oxidation mechanisms of the two alloys are explained using energy-dispersive x-ray spectroscopy and X-ray photoelectron spectroscopy.

Most existing works consider direct perturbations of victim's state/action or the underlying transition dynamics to show vulnerability of reinforcement learning agents under adversarial attacks. However, such direct manipulation may not always be feasible in practice. In this paper, we consider another common and realistic attack setup: in a multi-agent RL setting with well-trained agents, during deployment time, the victim agent $\nu$ is exploited by an attacker who controls another agent $\alpha$ to act adversarially against the victim using an \textit{adversarial policy}. Prior attack models under such setup do not consider that the attacker can confront resistance and thus can only take partial control of the agent $\alpha$, as well as introducing perceivable ``abnormal'' behaviors that are easily detectable. A provable defense against these adversarial policies is also lacking. To resolve these issues, we introduce a more general attack formulation that models to what extent the adversary is able to control the agent to produce the adversarial policy. Based on such a generalized attack framework, the attacker can also regulate the state distribution shift caused by the attack through an attack budget, and thus produce stealthy adversarial policies that can exploit the victim agent. Furthermore, we provide the first provably robust defenses with convergence guarantee to the most robust victim policy via adversarial training with timescale separation, in sharp contrast to adversarial training in supervised learning which may only provide {\it empirical} defenses.

Reliable localization is crucial for autonomous robots to navigate efficiently and safely. Some navigation methods can plan paths with high localizability (which describes the capability of acquiring reliable localization). By following these paths, the robot can access the sensor streams that facilitate more accurate location estimation results by the localization algorithms. However, most of these methods require prior knowledge and struggle to adapt to unseen scenarios or dynamic changes. To overcome these limitations, we propose a novel approach for localizability-enhanced navigation via deep reinforcement learning in dynamic human environments. Our proposed planner automatically extracts geometric features from 2D laser data that are helpful for localization. The planner learns to assign different importance to the geometric features and encourages the robot to navigate through areas that are helpful for laser localization. To facilitate the learning of the planner, we suggest two techniques: (1) an augmented state representation that considers the dynamic changes and the confidence of the localization results, which provides more information and allows the robot to make better decisions, (2) a reward metric that is capable to offer both sparse and dense feedback on behaviors that affect localization accuracy. Our method exhibits significant improvements in lost rate and arrival rate when tested in previously unseen environments.

Deep learning-based recommender systems (DRSs) are increasingly and widely deployed in the industry, which brings significant convenience to people's daily life in different ways. However, recommender systems are also shown to suffer from multiple issues,e.g., the echo chamber and the Matthew effect, of which the notation of "fairness" plays a core role.While many fairness notations and corresponding fairness testing approaches have been developed for traditional deep classification models, they are essentially hardly applicable to DRSs. One major difficulty is that there still lacks a systematic understanding and mapping between the existing fairness notations and the diverse testing requirements for deep recommender systems, not to mention further testing or debugging activities. To address the gap, we propose FairRec, a unified framework that supports fairness testing of DRSs from multiple customized perspectives, e.g., model utility, item diversity, item popularity, etc. We also propose a novel, efficient search-based testing approach to tackle the new challenge, i.e., double-ended discrete particle swarm optimization (DPSO) algorithm, to effectively search for hidden fairness issues in the form of certain disadvantaged groups from a vast number of candidate groups. Given the testing report, by adopting a simple re-ranking mitigation strategy on these identified disadvantaged groups, we show that the fairness of DRSs can be significantly improved. We conducted extensive experiments on multiple industry-level DRSs adopted by leading companies. The results confirm that FairRec is effective and efficient in identifying the deeply hidden fairness issues, e.g., achieving 95% testing accuracy with half to 1/8 time.

Despite of the superb performance on a wide range of tasks, pre-trained language models (e.g., BERT) have been proved vulnerable to adversarial texts. In this paper, we present RoChBERT, a framework to build more Robust BERT-based models by utilizing a more comprehensive adversarial graph to fuse Chinese phonetic and glyph features into pre-trained representations during fine-tuning. Inspired by curriculum learning, we further propose to augment the training dataset with adversarial texts in combination with intermediate samples. Extensive experiments demonstrate that RoChBERT outperforms previous methods in significant ways: (i) robust -- RoChBERT greatly improves the model robustness without sacrificing accuracy on benign texts. Specifically, the defense lowers the success rates of unlimited and limited attacks by 59.43% and 39.33% respectively, while remaining accuracy of 93.30%; (ii) flexible -- RoChBERT can easily extend to various language models to solve different downstream tasks with excellent performance; and (iii) efficient -- RoChBERT can be directly applied to the fine-tuning stage without pre-training language model from scratch, and the proposed data augmentation method is also low-cost., Accepted by Findings of EMNLP 2022

Objectives: We studied the consequences of parental domestic smoking behavior on children’s digital entertainment and mental health during COVID-19, and put forward the initiative of family tobacco control toprotect children’s physical and mental health. Methods: From February 20 to February 23, 2021, using Sojump survey method, 800 smoking parents who live with children aged 6∼12 answered the questionnaire online about basic information, children’s digital entertainment status, strengths and difficulties (SDQ). Results: 60.8% of the children exposed to secondhand smoke spend more than 1 h on digital entertainment every day; abnormal detection rates of emotional symptoms, conduct problems, hyperactivity and inattention, peer interaction, pro-social behavior, and difficulty total scores are 29.2%, 38.2%, 6.5%, 80.3%, 22.0%, 40.0%, respectively. Single factor analysis suggests that there are statistical differences in the degree of impact of household monthly income and children’s life under smoking parents on digital entertainment duration (χ= 14.76, 10.88; P

The marine internet of things (MIoT), an increasingly important foundation for ocean development and protection, consists of a variety of marine distributed sensors under water. These sensors of the MIoT have always been highly dependent on batteries. To realize in situ power supply, a flexible seaweed-like triboelectric nanogenerator (S-TENG) capable of harvesting wave energy is proposed in this study. The flexible structure, designed with inspiration from the seaweed structure, processes extensive marine application scenarios. The bending and recovering of the S-TENG structure under wave excitations are converted to electricity. As the output performance increases with the number of parallel connected S-TENG units, an S-TENG system with multiple units could serve for floating buoys, coastal power stations, and even submerged devices. Through the demonstration experiments performed in this study, the flexible, low-cost S-TENG could become an effective approach to achieve a battery independent MIoT.

Endometrial cancer (EC) is the gynecological tumor with the highest incidence. In recent years, it has been proved that necroptosis is a method of cell death related to EC. However, the expression of necroptosis-related miRNA in EC and its correlation with prognosis still ill-defined. Use the Cancer Genome Atlas (TCGA) cohort to obtain prognostic data and related clinical data for ECs and normal endometrium tissues. In this study, we identified three necroptotic regulatory miRNAs that are necroptosis-related and survival-related miRNAs (DENSMs) between normal endometrium tissues and EC from 13 necroptosis-related miRNAs. The three DENSMs signature was built to develop prognostic model and classified all EC patients into a high or low risk group. EC patients in the low-risk group showed significantly higher survival possibilities than those in the high-risk group (p = 0.0242), and the risk score was found to be an independent prognosis factor for predicting the OS of EC patients (p = 0.0254) in multivariate Cox regression. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed dephosphorylation, microtubule, protein serine/threonine kinase activity, PI3K-Akt signaling pathway and MAPK signaling pathway are closely related to it. In conclusion, the risk prediction model based on necroptosis-related miRNAs can effectively predict the prognosis of EC patients.

For the scalability of industrial online advertising systems, a two-stage auction architecture is widely used to enable efficient ad allocation on a large set of corpus within a limited response time. The current deployed two-stage ad auction usually retrieves an ad subset by a coarse ad quality metric in a pre-auction stage, and then determines the auction outcome by a refined metric in the subsequent stage. However, this simple and greedy solution suffers from serious performance degradation, as it regards the decision in each stage separately, leading to an improper ad selection metric for the pre-auction stage. In this work, we explicitly investigate the relation between the coarse and refined ad quality metrics, and design a two-stage ad auction by taking the decision interaction between the two stages into account. We decouple the design of the two-stage auction by solving a stochastic subset selection problem in the pre-auction stage and conducting a general second price (GSP) auction in the second stage. We demonstrate that this decouple still preserves the incentive compatibility of the auction mechanism. As the proposed formulation of the pre-auction stage is an NP-hard problem, we propose a scalable approximation solution by defining a new subset selection metric, namely Pre-Auction Score (PAS). Experiment results on both public and industrial dataset demonstrate the significant improvement on social welfare and revenue of the proposed two-stage ad auction, than the intuitive greedy two-stage auction and other baselines.

In recent years, with the development of modern wireless communication technology, it has become a new communication technology. Compared with the traditional communication technology, modern wireless communication is relatively combined with more advanced science and technology. When people communicate on wireless communication, they can get more convenience at a faster speed, which has a significant impact on people’s lives and is also recognized by the general public. Modern wireless communication technology makes people’s life and communication more convenient, but how to improve this technology in the future is also the focus of the majority of scholars. In this paper, the author first briefly expounds the modern wireless communication technology, then studies the problems of modern wireless communication technology, and finally expounds its future development trend.

Gynostemma longipes contains an abundance of dammarane-type ginsenosides and gypenosides that exhibit extensive pharmacological activities. Increasing attention has been paid to the elucidation of cytochrome P450 monooxygenases (CYPs) and UDP-dependent glycosyltransferases (UGTs) that participate downstream of ginsenoside biosynthesis in the Panax genus. However, information on oxidosqualene cyclases (OSCs), the upstream genes responsible for the biosynthesis of different skeletons of ginsenoside and gypenosides, is rarely reported. Here, an integrative study of the metabolome and the transcriptome in the leaf, stolon, and rattan was conducted and the function of GlOSC1 was demonstrated. In total, 46 triterpenes were detected and found to be highly abundant in the stolon, whereas gene expression analysis indicated that the upstream OSC genes responsible for saponin skeleton biosynthesis were highly expressed in the leaf. These findings indicated that the saponin skeletons were mainly biosynthesized in the leaf by OSCs, and subsequently transferred to the stolon via CYPs and UGTs biosynthesis to form various ginsenoside and gypenosides. Additionally, a new dammarane-II synthase (DDS), GlOSC1, was identified by bioinformatics analysis, yeast expression assay, and enzyme assays. The results of the liquid chromatography–mass spectrometry (LC–MS) analysis proved that GlOSC1 could catalyze 2,3-oxidosqualene to form dammarenediol-II via cyclization. This work uncovered the biosynthetic mechanism of dammarenediol-II, an important starting substrate for ginsenoside and gypenosides biosynthesis, and may achieve the increased yield of valuable ginsenosides and gypenosides produced under excess substrate in a yeast cell factory through synthetic biology strategy.

In online advertising, auto-bidding has become an essential tool for advertisers to optimize their preferred ad performance metrics by simply expressing high-level campaign objectives and constraints. Previous works designed auto-bidding tools from the view of single-agent, without modeling the mutual influence between agents. In this paper, we instead consider this problem from a distributed multi-agent perspective, and propose a general $\underline{M}$ulti-$\underline{A}$gent reinforcement learning framework for $\underline{A}$uto-$\underline{B}$idding, namely MAAB, to learn the auto-bidding strategies. First, we investigate the competition and cooperation relation among auto-bidding agents, and propose a temperature-regularized credit assignment to establish a mixed cooperative-competitive paradigm. By carefully making a competition and cooperation trade-off among agents, we can reach an equilibrium state that guarantees not only individual advertiser's utility but also the system performance (i.e., social welfare). Second, to avoid the potential collusion behaviors of bidding low prices underlying the cooperation, we further propose bar agents to set a personalized bidding bar for each agent, and then alleviate the revenue degradation due to the cooperation. Third, to deploy MAAB in the large-scale advertising system with millions of advertisers, we propose a mean-field approach. By grouping advertisers with the same objective as a mean auto-bidding agent, the interactions among the large-scale advertisers are greatly simplified, making it practical to train MAAB efficiently. Extensive experiments on the offline industrial dataset and Alibaba advertising platform demonstrate that our approach outperforms several baseline methods in terms of social welfare and revenue., Accepted by WSDM 2022

Natural Language Processing (NLP) has been revolutionized by the use of Pre-trained Language Models (PLMs) such as BERT. Despite setting new records in nearly every NLP task, PLMs still face a number of challenges including poor interpretability, weak reasoning capability, and the need for a lot of expensive annotated data when applied to downstream tasks. By integrating external knowledge into PLMs, \textit{\underline{K}nowledge-\underline{E}nhanced \underline{P}re-trained \underline{L}anguage \underline{M}odels} (KEPLMs) have the potential to overcome the above-mentioned limitations. In this paper, we examine KEPLMs systematically through a series of studies. Specifically, we outline the common types and different formats of knowledge to be integrated into KEPLMs, detail the existing methods for building and evaluating KEPLMS, present the applications of KEPLMs in downstream tasks, and discuss the future research directions. Researchers will benefit from this survey by gaining a quick and comprehensive overview of the latest developments in this field., Comment: 19 pages, 12 figures, 192 references

In this study, we investigated the effect of fat aggregate size and percentage on fat network formation and, ultimately, on the melting properties of ice cream. To control fat destabilization degree and fat network formation, we varied fat and emulsifier content, and blended coconut oil with milk fat, obtaining three different sample series varying in: 1) fat content, 2) fat aggregate size, and 3) fat aggregate percentage. The degree of fat destabilization in terms of aggregate size and the percentage was measured by light scattering techniques. The distribution of the fat around the air cells and in the unfrozen serum phase was calculated based on the measured overrun. Overall, a similar overrun was found in the three series of ice cream. The fat percentage in the remaining phase was measured to verify how much fat and what type of fat aggregates were present in the fat network. The results show that fat destabilization degree is relatively more important than fat content in determining the melting behavior of ice cream with low overrun. Clear relations between different fat destabilization parameters and ice cream melting behavior were established, indicating that the melting behavior of ice cream is related to specific properties of the fat network. Controlling fat destabilization in the unfrozen ice cream phase may be used to alter the properties of ice cream, and could contribute as a fat reduction strategy in ice cream.

Recent studies show that pre-trained language models (LMs) are vulnerable to textual adversarial attacks. However, existing attack methods either suffer from low attack success rates or fail to search efficiently in the exponentially large perturbation space. We propose an efficient and effective framework SemAttack to generate natural adversarial text by constructing different semantic perturbation functions. In particular, SemAttack optimizes the generated perturbations constrained on generic semantic spaces, including typo space, knowledge space (e.g., WordNet), contextualized semantic space (e.g., the embedding space of BERT clusterings), or the combination of these spaces. Thus, the generated adversarial texts are more semantically close to the original inputs. Extensive experiments reveal that state-of-the-art (SOTA) large-scale LMs (e.g., DeBERTa-v2) and defense strategies (e.g., FreeLB) are still vulnerable to SemAttack. We further demonstrate that SemAttack is general and able to generate natural adversarial texts for different languages (e.g., English and Chinese) with high attack success rates. Human evaluations also confirm that our generated adversarial texts are natural and barely affect human performance. Our code is publicly available at https://github.com/AI-secure/SemAttack., Comment: Published at Findings of NAACL 2022

Generations of epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) can significantly improve the outcome of EGFR-positive NSCLC patients. However, acquired TKIs-resistant mutations are inevitable. Except the common EGFR alterations, more and more rare mutations are revealed by next-generation sequencing (NGS), the clinical significance of which are still unclear. Here, we report an advanced lung adenocarcinoma patient who harbored two novel EGFR exon 19 deletions (750_758del and I759S) at the beginning and exhibited a short response to icotinib for 7.0 months. Then, secondary resistance EGFR T751_I759delinsS occurred. Chemotherapy combined with bevacizumab and erlotinib was administered in turn but failed. Standard-dose osimertinib (80 mg daily) obtained durable clinical remission for 16 months, and high-dose osimertinib (160 mg daily) further prolonged the survival of 9 months after leptomeningeal metastases (LM) occurring. This study presented the first case of intractable terminal NSCLC in a patient with EGFR 750_758del, I759S and T751_I759delinsS mutations, who responded positively to osimertinib and achieved a prolonged OS of 52 months, providing a potential therapeutic option for the patients harboring these particular EGFR mutations.

Microplastics (MPs) pollution has drawn increasing concern due to its widespread occurrence and potential risks in the environment. The reliable methods and instruments for fast analysis of microplastics (MPs) less than 5 mm are urgently needed. In this study, a new method based on custom-made portable pyrolysis-mass spectrometry (Pyr-MS) is developed, which enables rapid identification and mass related quantification of MPs. MPs are decomposed in the compact pyrolyzer and then directly analyzed in the portable MS by the chemical fingerprints of polymers including characteristic ions and their special ratio. It avoids the complex extraction and separation procedures of the pyrolysis/thermogravimetric-gas chromatography-mass spectrometry (Pyr/TGA-GC-MS), realizes the rapid analysis of MPs in 5 min, and thus can practically apply to a large number of MPs samples. In comparison to Fourier transform infrared spectroscopy (FT-IR) and Raman, this method is not limited by the shape, size, and color of MPs. Four common plastics including polyethylene (PE), polypropylene (PP), polystyrene (PS), and poly(methyl methacrylate) (PMMA) were investigated to verify the feasibility of this method. The environmental MPs samples collected from a beach were successfully identified and quantified, demonstrating the simplicity and practicality of this approach. The influence of plastics aging on the chemical fingerprints and the potential of mixed plastics detection by Pyr-MS are also assessed. The portable Pyr-MS could provide a promising tool for in-field analysis of MPs such as ship-based marine MPs surveys.

Environmental or human-induced disturbances, such as compaction and drought commonly caused by modern agricultural machinery and extreme climate events, have drawn great attention towards their impacts on soil microbial activities and associated declines in soil health and crop yield. To be specific, soil compaction combined with local extreme weather events, such as intensive rainfall or prolonged drought, may cause a significant decrease in soil microbial activity and threaten sustainable crop production. The use of organic amendment and crop rotation are commonly adopted as improved field management strategies to mitigate the disturbance-induced (compaction and drought) declines in crop yield, soil carbon, and soil health. Although widespread interest exists in the assessment of the detrimental impacts of environmental disturbances on agricultural systems, the response of belowground nutrient turnover processes to compaction and/or drought and how such responses are mediated by the microbial communities are less explored. In addition, there is a paucity of information on soil microbial responses to compaction and drought stresses and their recovery from the stressed conditions, particularly the differences in soils with conventional and improved management history. Furthermore, a robust, cheap, and easy-to-understand index needs to be developed in monitoring soil microbial community responses to compaction and drought stresses as current mainstream soil health indicators are expensive and complicated for end-users.This thesis aimed to investigate microbial responses to soil constraints as a measure of soil health in sugarcane and grain systems. The specific obj ctives of this thesis were to a) determine the mechanisms responsible for soils��� microbial functional response and its thresholds to environmental disturbance (compaction and drought); b) adopt this mechanism of the responses of soil microbial community following soil compaction (or drought) stress into both conventional and improved soil management practices contexts; c) generate soil resistance and resilience index based on the responses of soil microbes to compaction and/or drought ; d) enrich the knowledge of soil microbial activity and nutrient pools responses to different practices of removing compaction stress, such as deep trenching mill-mud application and shallow furrow application, in fields condition. Three incubation experiments and a field trial were designed to test following hypotheses: 1) Soil biological functions such as respiration and microbial biomass and activity are reliable indices for assessment of soil health and resilience to compaction and moisture (waterlogging or drought) stresses; 2) Soils with improved management practice history have higher resistance against compaction stress and higher resilience following removal of the stress; 3) Response pattern of soil microbial community to drought stress is highly related to the applied stress levels rather than the history of field management practices; 4) Application of organic amendments (e.g., crop residues, mill-mud) and removal of soil compaction stress would increase the size of soil microbial community and microbial activity (indicating soil resistance and resilience), which further improves soil health condition and increases crop yield. In the first incubation experiment, two contrasting sugarcane soils (Planosols at Rocky Point and Nitisols at Ingham) were exposed to a combination of compaction and drought stresses, and microbial functions were investigated going into the stress as well as their response coming out of the stress. We artificially applied a gradient of bulk densities (0.9 - 1.5 g cm-3) and water fill pore space (WFPS; 21% to 100%). Under high water content (i.e., WFPS of 47% to 100%), low compaction levels (1.1 and 1.2 g cm-3) increased soil cumulative microbial respiration in Planosols treatments by 18% in comparison to nil compaction treatment (0.9 g cm-3), while high compaction levels (1.3 - 1.5 g cm-3) decreased soil cumulative microbial respiration by 25% with increasing of compaction stress in 74 days incubation experiments. In contrast, in Nitisols with high water content, the highest compaction treatment (1.5 g cm-3) significantly increased soil cumulative microbial respiration by 12% in comparison to nil treatment (0.9 g cm-3). Our data revealed that the Nitisols had a higher resistance index of microbial C use efficiency, while the Planosols had a higher resilience index of microbial C use efficiency to compaction and moisture stresses. This could be attributed to greater aggregation potential (associated with more finely textured particles), and higher diversity of the microbial community in the Nitisols, which provide higher functional stability to resist environmental disturbance, while fast drainage and higher adaptation of microbial communities in the Planosols would provide a faster recovery from the applied stresses. This also indicates that soil microbial responses to compaction stress are governed by soil texture and moisture status. In the second incubation experiment, Nitisols with conventional and improved management history from adjacent sugarcane farms at Foresthome, North Queensland, Australia, were exposed to compaction stress (1.4 g cm-3), and microbial functions were investigated during compaction and ploughing cycle as well as their response to surface and mixing plant residue application methods over 70 days. The improved management block was managed with minimum tillage, mound planting, and legume plantation in comparison to conventional furrow management. Compaction was applied at the commencement of the experiment and removed at day 28 via plant residue application combined with ploughing. Overall, the concentration of labile C was 42% higher in treatments with improved management history. Within treatments with the same management history, the treatments with ploughing practice following compaction stress generated the highest cumulative and net cumulative CO2 emissions, followed by compaction-only and ploughing-only treatments. Interestingly, we found that legume residue incorporation along with improved land management minimized fluctuation of net cumulative CO2 emissions between compacted and non-compacted treatments by the end of experiment; however, we did not find the same pattern in treatments with conventional land management. Our results revealed that different field management practices might not alter soil microbial response patterns to compaction stress as microbial activity is mainly governed by water content and water fill pore space. Improved field management practices can improve soil resistance and resilience to compaction stress, which is shown as a faster stabilization of microbial properties after compaction stress and its removal due to their higher organic matter content and complexity of microbial community. The present study confirmed that application of legume residue increased the supply of organic C and N for soil microbial community, enhanced the nutrient cycling processes, and improved soil health status. In the third incubation experiment, Planosols collected from Wickepin, Western Australia (one soil with conventional management history and one soil with improved management history) were exposed to severe, moderate and nil drought stress and microbial functions were investigated going into stress as well as their response coming out of stress via rewetting and plant residue input. In general, treatments with improved management history had higher (15% - 53%) cumulative CO2 respiration in comparison to treatments with conventional management throughout the incubation period without plant residue amendment. In addition, drought stress significantly decreased cumulative CO2 respiration regardless of soil management history. We found that the differences of cumulative CO2 emissions decreased at day 56 but increased at the end of experiment (day 70) in both drought stressed treatments with conventional management history. The hot water extractable organic C pool was highly related (r = 0.56 - 0.78) to the umulative CO2 respiration and microbial activity. The results showed that different field management practices may not change soil microbial response patterns to drought, as microbial activity is mainly governed by soil texture when water content is limited. Improved field management could help to build soil resilience to drought, which is shown as the tolerance to moderate drought and resistance to severe drought due to their high organic matter content and complexity of microbial community. However, soil under conventional field management would have a lower resistance, but higher recovery to drought stress. The plant residue application increased the concentration of microbial biomass and enzyme activity via increasing labile C and nutrients contents regardless of soil management history. A field trial was conducted at Burdekin, Australia, to investigate the effects of different decompaction field managements on soil nutrient cycling, associated biological activities, and sugarcane yield. This experiment included four treatments comprising: control (CK, without mill-mud), mill-mud shallow furrow (MS), deep trench without mill-mud (DT), and deep trench mill-mud application (MD). Overall, the application of mill-mud significantly improved soil organic matter and nutrient content. Surface millmud application increased the concentration of soil Colwell P six-fold in comparison to the control. Deep trench application of mill-mud increased concentrations of hot water extractable organic C by 30% - 70% and hot water extractable total N by 30% - 90% at the application depth (ca. 20 cm depth). Soil microbial biomass C and N were also higher in mill-mud applied layers (ca. 20 cm depth). As expected, in comparison to the control, mill-mud applied treatments increased plant cane yield by 7% (MS treatment) and 14% (MD treatment). The deep trench without mill-mud (DT) practice also increased the plant cane yield by 11% compared to the control, which may be due to the release of native organic C and improved soil health. Deep trench combined with the mill-mud application (MD) increased the supply of organic C and N and nutrients to the microbial community within the entire soil profile, enhanced nutrient cycling processes and soil health for sugarcane growth, and thus increased sugarcane productivity. In the presented thesis, soil microbial responses to two main soil constraints (compaction and drought) under different soil textures, vegetation type and field management history were investigated. Based on the presented result, soil texture and WFPS played vital roles in regulating microbial responses to compaction and drought. The finer soil texture provided a better and stable microbial habitat and WFPS governed the diffusion of soil labile C and nutrients for microbial growth. In addition, field managements such as organic matter amendment and/or ploughing would stimulate microbial responses to soil constraints as organic amendment would directly increase soil organic matter content and ploughing would adjust soil WFPS. Also, these two factors are intensively linked to crop yield as increased sugarcane yield was found in organic matter applied and ploughed treatments, particularly in ploughed only treatment. Additionally, the presented study found that the soil microbial community had higher tolerance to soil disturbances under improved field management history than conventional management history, which could be attributed to the complexity of microbial community structure established through long term improved field management. However, microbial community structure and key functional gene analysis are needed to confirm this statement. In conclusion, application of molecular biology approaches would further increase the potential of using microbial responses to soil constraints as a measure of soil health and generate a comprehensive index in monitoring soil health in the long run.

Domain generalization (DG) aims to learn from multiple source domains a model that can generalize well on unseen target domains. Existing DG methods mainly learn the representations with invariant marginal distribution of the input features, however, the invariance of the conditional distribution of the labels given the input features is more essential for unknown domain prediction. Meanwhile, the existing of unobserved confounders which affect the input features and labels simultaneously cause spurious correlation and hinder the learning of the invariant relationship contained in the conditional distribution. Interestingly, with a causal view on the data generating process, we find that the input features of one domain are valid instrumental variables for other domains. Inspired by this finding, we propose an instrumental variable-driven DG method (IV-DG) by removing the bias of the unobserved confounders with two-stage learning. In the first stage, it learns the conditional distribution of the input features of one domain given input features of another domain. In the second stage, it estimates the relationship by predicting labels with the learned conditional distribution. Theoretical analyses and simulation experiments show that it accurately captures the invariant relationship. Extensive experiments on real-world datasets demonstrate that IV-DG method yields state-of-the-art results., Accepted by ACM Transactions on Knowledge Discovery from Data (TKDD)

Measuring and promoting policy diversity is critical for solving games with strong non-transitive dynamics where strategic cycles exist, and there is no consistent winner (e.g., Rock-Paper-Scissors). With that in mind, maintaining a pool of diverse policies via open-ended learning is an attractive solution, which can generate auto-curricula to avoid being exploited. However, in conventional open-ended learning algorithms, there are no widely accepted definitions for diversity, making it hard to construct and evaluate the diverse policies. In this work, we summarize previous concepts of diversity and work towards offering a unified measure of diversity in multi-agent open-ended learning to include all elements in Markov games, based on both Behavioral Diversity (BD) and Response Diversity (RD). At the trajectory distribution level, we re-define BD in the state-action space as the discrepancies of occupancy measures. For the reward dynamics, we propose RD to characterize diversity through the responses of policies when encountering different opponents. We also show that many current diversity measures fall in one of the categories of BD or RD but not both. With this unified diversity measure, we design the corresponding diversity-promoting objective and population effectivity when seeking the best responses in open-ended learning. We validate our methods in both relatively simple games like matrix game, non-transitive mixture model, and the complex \textit{Google Research Football} environment. The population found by our methods reveals the lowest exploitability, highest population effectivity in matrix game and non-transitive mixture model, as well as the largest goal difference when interacting with opponents of various levels in \textit{Google Research Football}., We investigate a new perspective on unifying diversity measures for open-ended learning in zero-sum games, which shapes an auto-curriculum to induce diverse yet effective behaviors

The data generated by the launch vehicle test launch system has the characteristics of big data and small samples. Using small sample learning to start from a small amount of faulty data, and learning and processing sample data through data enhancement and metric learning strategies can greatly improve the effect of data mining. In view of the current situation that the deep learning method in the launch vehicle test system has not been used on a large scale, the characteristics of the data generated by the launch vehicle test launch are analyzed, and the method of small sample learning based on the deep neural network is combined with the data analysis of the launch vehicle test. And expounded the potential value and significance of this method to the development of launch vehicle test system.

Beta adrenergic receptors (βARs) mediate physiologic responses to the catecholamines epinephrine and norepinephrine released by the sympathetic nervous system. While the hormone epinephrine binds β1AR and β2AR with similar affinity, the smaller neurotransmitter norepinephrine is approximately tenfold selective for the β1AR. To understand the structural basis for this physiologically important selectivity, we solved the crystal structures of the human β1AR bound to an antagonist carazolol and different agonists including norepinephrine, epinephrine and BI-167107. Structural comparison revealed that the catecholamine-binding pockets are identical between β1AR and β2AR, but the extracellular vestibules have different shapes and electrostatic properties. Metadynamics simulations and mutagenesis studies revealed that these differences influence the path norepinephrine takes to the orthosteric pocket and contribute to the different association rates and thus different affinities.

The purpose of this study was to quantitatively assess the vertical force distribution (VFD) of subject-specific healthy blue sheep while walking on different slopes using a pressure-sensing walkway. The blue sheep was trained to walk over the pressure-sensing walkway by choosing a comfortable walking speed, and the slope angle increased from 0° to 25°. The sheep's hooves were divided into four quadrants, namely, the cranio-lateral, cranio-medial, caudo-lateral, and caudo-medial quadrants, to investigate the VFD of the peak vertical force (PVF), vertical impulse (VI) and occurrence time of the PVF during the stance phase (TPVF). This study demonstrates that the main stressed quadrant of the front hoof changes from the caudo-medial quadrant to the cranio-medial quadrant with increasing slope. The main stressed quadrant of the rear hoof is the cranio-medial quadrant and does not change with the increasing slope. For all the slopes, the vertical force shifted from the lateral quadrant to the medial quadrant and from the caudal quadrant to the cranial quadrant. All the results obtained in the study suggest the feasibility of detecting gait changes in blue sheep, which has potential for the diagnosis of lower limb musculoskeletal diseases in quadrupeds.

This study investigated the effect of activated silica on both the polymerization mechanism and strength development of municipal solid waste incineration (MSWI) bottom ash alkali-activated mortars, with a specific focus on increasing the silica content by adding either solid or liquid sodium silicate. The results show that such additives increase the content of active silicon and improve the resulting compressive strength of mortars, which is more effectively enhanced by liquid solid sodium silicate addition. However, a significant decline of these effects was found with increased addition of sodium silicate. Solid sodium silicate is difficult to dissolve; therefore, increasing its content not only fails to improve the content of active silicon but also causes precipitation. Using liquid sodium silicate could further improve the content of active silicon; however, excessive addition will result in a large surplus of activated silicon and thus, causing sodium and magadiite precipitation.

High-quality low Sn-content GeSn film has been grown on tensile strained Ge/Si (100) substrate using a physical vapor deposition reactor and no subsequent annealing was carried out after the deposition of the film. High Resolution X-ray diffraction (HR-XRD) measurement was performed to estimate the crystallinity and the residual strain in the Ge layer. Secondary-ion mass spectrometry (SIMS) and X-ray photoelectron spectrometer (XPS) show Sn incorporation of the film is approximately 3%. High-resolution transmission electron microscopy (TEM) and atomic force microscopy (AFM) results show that the crystallinity of the film is quite good and it shows a smooth surface (RMS = 0.744 nm) in 5 µm × 5 µm scan. Raman spectroscopy and photoluminescence (PL) measurements were carried out to study the structural and optical properties of the film. Room-temperature PL of the film was observed and it results in a decrease in the difference between Γ and L valleys of tensile strained Ge (0.21%) from 87 meV to 76 meV, which makes this method promising for obtaining Pseudo-direct bandgap group IV films.

Power factor correction (PFC) converter is widely utilized in grid-connected equipment and power adapter, etc. Conventional PFC Buck converter is not proper to maintain good performance at low output voltage with a wide input range. In the bridge-based PFC converter, the forward voltage of rectifier diodes is an important source of power loss. For overcoming the above-mentioned issues, a bridgeless PFC converter based on switched-capacitor dual-resonance tank is proposed in this paper. The switched-capacitor dual-resonance tank is meaningful to improve voltage regulating capacity and the bridgeless structure is helpful to eliminate rectifier diode loss. Additionally, only small inductors are utilized in the proposed PFC converter, which is useful to improve power density.

With the rapid development of smart cities, white light-emitting diodes (LEDs) are widely used in indoor lighting due to their characteristics of energy-saving, long lifetimes, and low cost. At the same time, the high-frequency modulated LEDs allow visible light to be used for indoor communications, positioning, and security. Visible light applications have many advantages, such as avoiding electromagnetic interference, high communication speed, great privacy, and high-precise positioning. In this paper, we will survey the application prospects and research results of visible light from scenarios such as high-speed communication, privacy security, and navigation localization and focus on analyzing and discussing the challenges and development trends encountered in visible light positioning.

Aiming at the problems existing in the current fault location methods of distribution grid, such as precise calculation of wave velocity and the finite precision of fault location, a precise fault location method of traveling wave in the distribution grid based on multiple measuring points is proposed. In this method, the line-mode traveling wave of a specific frequency is selected as the basis for the time calibration of the initial wave head, and the circuit is divided into fault zone and non-fault zone. The traveling wave velocity is real time calculated by using the traveling wave of the non-fault zone. The fault branch identification matrix (FBIM) is constructed and compared with the fault branch identification array line by line to identify the fault branch (FB). On this basis, the fault distance is solved precisely by using redundant end-point traveling wave and least square method. This method only requires the placement of measuring points at the end of each branch, which to some extent solves the problem of limitation of measuring point allocation. Simulation results show that the method is highly reliable and has high fault location accuracy in different fault locations.

Huiying Li,1 Tingting Yu,1 Yongjuan Lin,1 Yu Xie,1 Jie Feng,1 Mingmin Huang,1 Aibin Guo,1 Xiangyu Liu,2 Zhenyu Yin1 1Department of Geriatric Oncology, Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School, Nanjing, People’s Republic of China; 2Department of Neurosurgery, Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School, Nanjing, People’s Republic of ChinaCorrespondence: Zhenyu Yin Department of Geriatric OncologyAffiliated Nanjing Drum Tower Hospital of Nanjing University Medical School Tel +8613913904579Email zhenyuyin68@163.comXiangyu LiuDepartment of Neurosurgery, Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School Tel +8613851672018Email liuxiangyumail@163.comAbstract: Generations of epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) can significantly improve the outcome of EGFR-positive NSCLC patients. However, acquired TKIs-resistant mutations are inevitable. Except the common EGFR alterations, more and more rare mutations are revealed by next-generation sequencing (NGS), the clinical significance of which are still unclear. Here, we report an advanced lung adenocarcinoma patient who harbored two novel EGFR exon 19 deletions (750_758del and I759S) at the beginning and exhibited a short response to icotinib for 7.0 months. Then, secondary resistance EGFR T751_I759delinsS occurred. Chemotherapy combined with bevacizumab and erlotinib was administered in turn but failed. Standard-dose osimertinib (80 mg daily) obtained durable clinical remission for 16 months, and high-dose osimertinib (160 mg daily) further prolonged the survival of 9 months after leptomeningeal metastases (LM) occurring. This study presented the first case of intractable terminal NSCLC in a patient with EGFR 750_758del, I759S and T751_I759delinsS mutations, who responded positively to osimertinib and achieved a prolonged OS of 52 months, providing a potential therapeutic option for the patients harboring these particular EGFR mutations.Keywords: non-small cell lung cancer, EGFR, 750_758del, I759S, T751_I759delinsS, osimertinib

Humans can systematically generalize to novel compositions of existing concepts. Recent studies argue that neural networks appear inherently ineffective in such cognitive capacity, leading to a pessimistic view and a lack of attention to optimistic results. We revisit this controversial topic from the perspective of meaningful learning, an exceptional capability of humans to learn novel concepts by connecting them with known ones. We reassess the compositional skills of sequence-to-sequence models conditioned on the semantic links between new and old concepts. Our observations suggest that models can successfully one-shot generalize to novel concepts and compositions through semantic linking, either inductively or deductively. We demonstrate that prior knowledge plays a key role as well. In addition to synthetic tests, we further conduct proof-of-concept experiments in machine translation and semantic parsing, showing the benefits of meaningful learning in applications. We hope our positive findings will encourage excavating modern neural networks' potential in systematic generalization through more advanced learning schemes., Accepted to the Fifth BlackboxNLP in EMNLP 2022

In this work, electromagnetic heating was firstly explored as sample introduction approach in portable microplasma-atomic emission spectrometer to achieve the direct, rapid analysis of soil sample. The device primarily consists of an electromagnetic heating unit, a dielectric barrier discharge (DBD) excitation source and an optical signal acquisition unit. A W-boat was used as an electromagnetic heating medium and sample carrier, and copper coil spiraled around the tube was used as magnetic induction coil. With applying a voltage on copper coil, W-boat was electromagnetically heated to vaporize analyte-containing species for sample introduction into the microplasma. The portable battery-powered device is controlled by a miniature touch screen computer with the main advantages of small size (40.5 cm (l) × 30 cm (w) × 15 cm (h).), light weight (less than 7 kg), low-power consumption (the average power consumption is 118 W). By this method, Hg, Cd and Pb in soil were simultaneously analyzed within 4 min. Under the optimal conditions, the limits of detection for Hg, Cd and Pb in soils were 8.0 μg/kg, 17.8 μg/kg and 3.5 mg/kg, respectively, meeting the requirements for environmental quality standards for soils of China. Different types of CRM soils were analyzed, demonstrating good accuracy, stability and utility of this method. This technique could be a promising and powerful tool for on-site, rapid analysis of heavy metals in soil even other solid samples. Electromagnetic heating mode provides a good alternative for solid sampling to develop portable, miniaturized atomic spectrometers for solid sample analysis.

Surface charge accumulation on spacer is an important factor to cause surface flashover. An experimental set-up is established to measure surface potential over a truncated cone-type spacer under DC voltage in SF 6 gas at atmospheric pressure using a Kelvin vibrating electrostatic probe. The surface charge density is calibrated based on multipoint measurement technique, which takes the influence of electrostatic probe into account. The experimental results show that surface charge distribution patterns can be divided into three classes including centrosymmetric distribution, spot-like distribution, and cloud-like distribution. The symmetrically distributed charges are injected by electrode, which migrate along spacer surface through the surface conductivity. While the charges showing spot-like and cloud-like distribution are generated by the ionisation of SF 6 due to local electrical field concentrated. Besides, the hetero-charges are more easily to be saturated with different charging durations and the polarity of the net charge is the same as that of applied voltage.

In this paper, a novel fully-depleted Ge1-xSnx n-Tunneling FET (FD Ge1-xSnx nTFET) with field plate is investigated theoretically based on the experiment previously published. The energy band structures of Ge1-xSnx are calculated by EMP and the band-to-band tunneling (BTBT) parameters of Ge1-xSnx are calculated by Kane's model. The electrical characteristics of FD Ge1-xSnx nTFET and FD Ge1-xSnx nTFET with field plate (FD-FP Ge1-xSnx nTFET) having various Sn compositions are investigated and simulated with quantum confinement model. The results indicated that the GIDL effect is serious in FD Ge1-xSnx nTFET. By employing the field plate structure, the GIDL effect of FD-FP Ge1-xSnx nTFET is suppressed and the off-state current Ioff is decreased more than 2 orders of magnitude having Sn compositions from 0 to 0.06 compared with FD Ge1-xSnx nTFET. The impact of the difference of work function between field plate metal and channel Φfps is also studied. With the optimized Φfps = 0.0 eV, the on-state current Ion = 4.6 × 10−5 A/μm, the off-state current Ioff = 1.6 × 10−13 A/μm and the maximum on/off ration Ion/Ioff = 2.9 × 108 are achieved.

Ocean current energy harvester is a promising infrastructure to achieve self-powered marine wireless sensing system. This study proposes and investigates an underwater flag-like Triboelectric Nanogenerator (UF-TENG) based on the flow-induced vibration. The UF-TENG consists of two conductive ink-coated polyethylene terephthalate membranes and one strip of poly tetra fluoroethylene membrane with their edges sealed up by a waterproof PTFE tape. In this way, the triboelectric layers are prevented from contacting water. It is verified that the vortex street induced by a cylinder enhances the vibration of the UF-TENG. The low velocity startup enables the UF-TENG to harvest extremely low-velocity ocean current energy, e.g., presently achieving 0.133 m/s as the critical startup velocity. The proposed UF-TENG has a simpler structure and better low velocity performance when comparing with electromagnetic generators and piezoelectric generators. Parametric studies are conducted to evaluate the influence of geometrical parameters of the UF-TENG on the vibration behavior and output performance. It is demonstrated that the underwater electrical appliance can be powered by parallel UF-TENG units. The present UF-TENG is a more cost-effective and more accessible hydroelectric technology to utilize the renewable ocean current for powering sensors or micro electric appliances in Internet of Things.

In this paper, a novel structure of the fully-depleted Ge-on-Insulator (GOI) n-channel MOSFET with field plate is proposed and studied by two-dimensional numerical simulation. The results indicate that the gated-induced drain leakage (GIDL) of the fully-depleted GOI NMOSFET with field plate (FD-FP NMOSFET) is suppressed effectively. Besides, off-state current Ioff decreases by 2 orders of magnitude compared with the device without field plate. The impacts of the distance of field plate from the drain Lcb and the difference of work function between field plate metal and channel material Φfps on the electrical characteristics of FD-FP NMOSFET are investigated. FD-FP NMOSFET with a Φfps of 0 eV and a Lcb of 5 nm, demonstrates a reduction in Ioff by more than 3 orders of magnitude as compared with the control device. An on/off ratio Ion/Ioff = 2.57 × 107, and a sub-threshold swing of 76.8mV/decade are achieved.

This paper proposes a parameter calculation method for the mixed initiation circuit with both nonel detonators (NDs) and electronic detonators (EDs) to increase the detonator delay for reducing the blast-induced vibration in tunnel blasting and to ensure safe circuit operation. The circuit involves the outside-hole delay for increasing the delay numbers under the premise of safe circuit operation. The time calculation and vibration-reduction effects of the circuit were quantitatively determined through mathematical modeling and field experimental studies. A calculation model of mixed detonator groups was established based on three constraint conditions: no delay irregularly, no misfire, and an increased delay number. The main parameters obtained were the initiation sequence of the EDs, the delay number of the NDs and the time interval of the EDs inside the blastholes. The factor considered in the mathematical modeling was the delay error of the NDs. The calculation results agreed well with experimental results, the latter confirming that the mixed initiation circuit was feasible and safe when using a 90 ms delay outside the blasthole. In addition, the delay numbers of the mixed-detonator group were increased by 40% relative to those of NDs. Compared with conventional blasting that reduces the advance per round, blasting with the mixed initiation circuit using an outside-hole delay produced 82.3% less blast-induced vibration. It was also found that the blasting cost could be reduced by 54.3% compared with that of using EDs only. Overall, the mixed initiation method had a positive effect on tunnel controlled-blasting. The research methods and results are expected to be useful in guiding related engineering efforts.

In this paper, a novel junctionless Ge n-Tunneling Field-Effect Transistors (TFET) structure is proposed. The simulation results show that Ion = 5.5 × 10−5A/μm is achieved. The junctionless device structure enhances Ion effectively and increases the region where significant BTBT occurs, comparing with the normal Ge-nTEFT. The impact of the lightly doped drain (LDD) region is investigated. A comparison of Ion and Ioff of the junctionless Ge n-TFET with different channel doping concentration ND and LDD doping concentration NLDD is studied. Ioff is reduced 1 order of magnitude with the optimized ND and NLDD are 1 × 1018cm−3 and 1 × 1017 cm−3, respectively. To reduce the gate induced drain leakage (GIDL) current, the impact of the sloped gate oxide structure is also studied. By employing the sloped gate oxide structure, the below 60 mV/decade subthreshold swing S = 46.2 mV/decade is achieved at Ion = 4.05 × 10−5A/μm and Ion/Ioff = 5.7 × 106.

In this paper, the structure of fully-depleted Ge NMOS with P-substrate (FDP-NMOS) is proposed. The DIBL effect of fully depleted GeOI NMOS (FD-NMOS) and FDP-NMOS has been studied based on the simulation results. It is demonstrated that DIBL of FD-NMOS occurs at the deep bulk of channel before the surface. From the current-voltage (I D -V G ) characteristics, the DIBL effect, the leakage current and the on-state current are determined. The results indicate that with the doping concentration of P-substrate (N B ) increasing, the DIBL effect is lower and the leakage current decreases significantly. Therefore, the FDP-NMOS has excellent subthreshold characteristics. When N B is 10 × 10 17 /cm 3 , the subthreshold swing is 76.8 mV/dec and the on/off ratio is 1.65 × 10 8 , respectively. The on/off ratio of FDP-NMOS increases more than 4 orders compared with the FD-NMOS and the experimental results.

The order of the input information plays a very important role in a distributed information processing system (DIPS). This paper proposes a novel data sorting mechanism named the {\epsilon}-differential agreement (EDA) that can support parallel data sorting. EDA adopts the collaborative consensus mechanism which is different from the traditional consensus mechanisms using the competition mechanism, such as PoS, PoW, etc. In the system that employs the EDA mechanism, all participants work together to compute the order of the input information by using statistical and probability principles on a proportion of participants. Preliminary results show variable fault-tolerant rates and consensus delay for systems that have different configurations when reaching consensus, thus it suggests that it is possible to use EDA in a system and customize these parameters based on different requirements. With the unique mechanism, EDA can be used in DIPS of multi-center decision cluster, not just the rotating center decision cluster., Comment: 6 pages, 4 figures

Multi robot SLAM systems have several advantages e.g. robustness and quicker exploration, which determine they have better performance and importance than single robot SLAM especially in time critical Search and Rescure missions. However, the development of multi robot SLAM is much slower, because of the difficulties such as determination of relative poses of robots and communication, etc. CORB-SLAM as a multi robot SLAM system, of which the clients are based on ORB-SLAM2, inheriting its advantages like low-cost sensor configuration and low computation cost, has not been evaluated in a quantitative method to assess its performance. For multi robot SLAM, the adaptability in long-term scenarios is another important research topic, since multi robots may run in different dates and seasons. Because CORB-SLAM does not employ extra algorithms to suppress illumination changes, its long-term performance is limited. This work quantitatively evaluates CORB-SLAM, and experiments to combine it with illumination variance algorithm to improve its performance under different illumination and season. Master of Science (Computer Control and Automation)

Background Wound infections, especially multidrug-resistant (MDR) bacterial infections, are a major challenge in clinical medicine. Methods In this study, a new type of antibacterial sponge was prepared from a solution containing a chitosan-polyvinyl alcohol (CTS-PVA) emulsion with added polyhexamethylene guanidine hydrochloride (PHMG) in a homogeneous medium using lyophilization technology. The antibacterial ability of and CTS-PVA/PHMG sponge against Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli, Candida albicans, Methicillin-resistant Staphylococcus aureus, multidrug-resistant Pseudomonas aeruginosa, and multidrug-resistant Acinetobacter baumannii in vitro. The structure and physical properties were characterized. The sponge dressing was tested in a Pseudomonas aeruginosa-infected full-thickness mouse skin wound defect model. The effects were evaluated by wound area measurement and histological analysis. Results The CTS-PVA/PHMG sponge showed broad-spectrum antibacterial ability, including for MDR bacterial stains from clinical sources, while maintaining excellent physicochemical properties, including a high swelling degree and good moisture retention capability. Scanning electron microscopy images displayed the surface morphology of the CTS-PVA/PHMG sponge dressing. The detection of the wound healing rate and histological analysis supported that the new dressing can alleviate the inflammation and accelerate the healing speed of infected wounds and in vivo. Conclusions CTS-PVA/PHMG sponge shows broad-spectrum antibacterial activity, which can provide a new pathway for clinical prevention and treatment of superbug-infected wounds.

In order to achieve high efficiency absorption of CO2 by fresh cement paste, a special CO2 absorption device was designed. Then the influences of stirring rate (SR), water-to-cement ratio(W/C), superplasticizer type and its addition order on the CO2 absorption rate (CO2 AR) and CO2 ultimate absorption amount(CO2 UAA)were investigated. Moreover, the fluidity and mechanical properties of cement paste with different CO2 absorption amount were determined. According to the experimental results, with increasing absorption amount of CO2, the cement paste gradually loses fluidity and becomes hardened, indicating CO2 absorption of cement paste has reached to CO2 UAA. Meanwhile, increasing SR can effectively increase CO2 AR, while the CO2 UAA cannot be increased. Increasing W/C results in an increase in both CO2 AR and CO2 UAA. On the other hand, superplasticizer added before CO2 absorption neither can help to increase CO2 UAA nor improve fluidity of the carbonated cement paste. Superplasticizer added after CO2 absorption can improve fluidity of the carbonated cement paste but is barely helpful for increasing the CO2 UAA. The influence of CO2 absorption on early mechanical strength of the carbonated cement mortar is not obvious while the later mechanical strength of the carbonated cement paste can be improved. SEM and EDS analysis indicate that with an increase in CO2 absorption amount (CO2 AA), water in the cement paste is gradually consumed, forming columnar crystals of ettringite and calcium carbonate and increasing the cement particles and the flocculation structure per unit volume of cement paste. As a result, cement paste gradually loses its fluidity.

Wide voltage range converters are significant for photovoltaic power generation system, battery charging, etc. Resonant converter has been preferred topology for these applications, however, some demerits exist: 1) Difficult to realize wide voltage gain range and low circulating current losses; 2) Widely regulated switching frequency to adapt to wide voltage gain range; 3) Poor regulation at light load. Based on above-mentioned shortcomings, dual-resonance-core based step-up/down switched capacitor converter topology is presented in this paper. Utilizing hybrid control strategy, non-overlapping control strategy is used in the Buck-type mode and overlapping control strategy is applied in the Boost-type mode with common switching frequency range, and therefore, reduced frequency range in the presented converter is achieved. Moreover, wide voltage gain range can be obtained due to step-up/down capacity in the proposed converter. Furthermore, conversion ratio of proposed converter can be reduced to zero theoretically by using dual-resonance-core existed in the resonant switched capacitor converter. Finally, comprehensive operational principle analysis and performance analysis are provided, and a 48V to 2890V/100W prototype is built to verify the analysis.

Microplastics (MPs) pollution in the marine has become a global concern. To obtain accurate measurements of MPs in the marine sediments is important for understanding its effects and potential risks to the environment. In this study, we developed a novel method for extracting MPs from sediments by density flotation of sodium dihydrogen phosphate (NaH2PO4) solution, which utilized the heating-assisted to increase the density of separation solution. It features high density, nontoxic, and economic, indicating its promising application in the MPs extraction. A standard procedure based on NaH2PO4 for extracting MPs from sediments was established and validated by the recovery experiments of seven common plastics. Results showed NaH2PO4 solution exhibited higher recovery rates than the commonly-used NaCl solution; three plastics with high density couldn't be recovered by NaCl, while NaH2PO4 was available for all the studied MPs. A large number of real beach samples was collected and validated by this method, and the abundance and characteristic of the collected MPs were further analyzed. Results exhibited the heating-assisted NaH2PO4 could be a promising way for extracting MPs in environmental samples.

The uncertain mechanical parameters of artificial frozen soil in deep alluvium are the key factors for computing the stochastic stress fields and stochastic deformation fields of artificial freezing engineering. In this study, a series of mechanical parameters tests for the artificially frozen clay are carried out, and the statistical data of elastic modulus, strength and Poisson ratio of frozen clay under three different low temperature conditions (−20 °C, −15 °C and −10 °C) are obtained. The experimental semi-variogram values of the mechanical parameters for artificially frozen clay under different temperature conditions are calculated and analyzed. The results show that the test data meets the requirements of the coefficient of variation analysis and the elastic modulus, strength and poisson ratio basically meet the normal distribution. The variability of the mechanical parameters for artificially frozen clay is basically in the degree of moderate to weak variability. The optimal theoretical variation function model of basic mechanical parameters of frozen clay is Gaussian model. The correlation of soil parameters calculated by Gauss model at smaller distances is greater than that of Exponential model and Spherical model. The mechanical parameters of artificially frozen clay are strongly dependent in spatial range. This study can improve our understanding of the variation characteristics of uncertain mechanical parameters for artificial frozen soil in deep alluvium and provide a theoretical basis and reference for the stochastic mechanical analysis of artificial freezing engineering.