基于改进DeepLabv3+的轻量化作物杂草识别方法

为在存储资源与计算能力有限的设备上实现田间作物和杂草的识别,本文提出一种基于改进DeepLabv3+的轻量化语义分割网络。首先,以MobileNet v2作为DeepLabv3+的特征提取骨干网络,提出双分支残差模块替换倒残差模块,并删除后两层卷积以降低模型参数量。其次,在空洞空间金字塔池化(Atrous Spatial Pyramid Pooling,ASPP)模块中引入分组逐点卷积,使用深度扩张卷积替换标准卷积,并将卷积后的特征图进行多尺度特征融合增强对作物和杂草深层特征的提取能力。最后,将原有的非线性激活函数替换为Leaky ReLU激活函数来提升分割精度。实验结果表明:改进后网络的mIOU达到86.75%,参数量仅为0.69M,FPS达到了98,与原始DeepLabv3+以及3个典型轻量化语义分割网络的相比,参数量最小,在对比的轻量化网络中具有最高的分割精度。

改进DeepLab v3+模型下的梯田遥感提取研究

[目的与意义]梯田作为农业生产的关键要素之一,其面积估算对于农业政策制定、土地规划和资源管理至关重要。为解决复杂的地形条件、种植环境导致传统遥感数据和监测方法难以开展梯田自动化提取问题,探索一种利用深度学习技术在高分辨率遥感影像中精准提取梯田面积的方法。[方法]以休耕期梯田高分六号影像构建语义分割数据集,同时提出一种改进的DeepLab v3+模型。该模型使用轻量级网络MobileNet v2作为骨干网络,为了同时兼顾局部细节和全局语境,使用多尺度特征融合(Multi-scale Feature Fusion module,MSFF)模块代替空洞空间金字塔池化(Atrous Spatial Pyramid Pooling,ASPP)模块,利用扩张率依次增大的空洞卷积级联模式改善信息丢失的问题。此外,对浅层特征和深层特征使用坐标注意力机制以加强网络对于目标的学习。[结果与讨论]利用红、绿和近红外波段组合方式在梯田提取的精度和效果上表现最佳。相比于原始DeepLab v3+网络,精确率、召回率、F_(1)评分和交并比指标分别提升4.62%、2.61%、3.81%和2.81%。此外,与UNet和原始DeepLab v3+相比,改进的DeepLab v3+在参数量上和浮点运算数有着更为优越的性能,其参数量仅为UNet的28.6%和原始DeepLab v3+的19.5%,同时浮点运算数仅为UNet和DeepLab v3+的1/5。这不仅提高了计算效率,也使得改进后的模型更适用于资源有限或计算能力较低的环境中。[结论]深度学习在高分辨率遥感影像梯田识别中具有较高的精度,有利于为梯田精细化监测和管理提供参考依据。

改进Deeplabv3+的双注意力融合作物分类方法

近年来,卷积神经网络(convolutional neural networks,CNN)在农作物分类研究中不断取得新进展,但在建模长期依赖关系方面表现出一定的局限性,对农作物全局特征的捕获存在不足。针对以上问题,将Transformer引入Deeplab v3+模型,提出了一种用于无人机影像农作物分类的并行分支结构--DeepTrans(Deeplab v3+with Trans-former)模型。DeepTrans以一种并行的方式将Transformer和CNN结合在一起,利于全局特征与局部特征的有效捕获。通过引入Transformer来增强图像中信息的远距离依赖关系,提高了作物全局信息的提取能力;加入通道注意力机制和空间注意力机制加强Transformer对通道信息的敏感度及ASPP(atrous spatial pyramid pooling)对作物空间信息捕获能力。实验结果表明,DeepTrans模型在MIoU指标上可达0.812,相较于Deeplab v3+模型提高了3.9%,该模型在五类作物的分类中精度均有提升,对于容易错分的甘蔗、玉米和香蕉三种作物,其IoU分别提高了2.9%、4.7%、13%。由此可见,DeepTrans模型在农作物分类图像的内部填充和全局预测方面有着更好的分割效果,有助于更准确地监测农田作物的种植结构及规模。

基于v3洋葱域名的比特币地址威胁程度分析

比特币可以在不透露使用者身份的情况下进行交换,导致其成为不法分子在暗网上进行违法活动的主要方式。为了追踪比特币非法交易,传统方法根据比特币的伪匿名性,在整个区块链上进行启发式地址聚类,没有充分利用比特币地址在暗网上的信息。2021年Tor官方全面启用v3洋葱域名,使得以往的v2洋葱域名数据无法再作为分析的依据。设计并实现基于v3洋葱域名的比特币地址威胁程度的一体化分析框架TLAFDB。信息收集模块使用境外服务器解决地域限制并设置socks5h代理以支持暗网爬虫运行,使用洋葱种子地址在暗网中爬行收集最新的v3洋葱域名数据,信息清洗模块采用可同时覆盖Base58和Bech32编码的正则表达式以提取v3洋葱域名网页中的比特币地址,通过区块链搜索引擎Blockchain.com筛选存在真实交易的比特币地址,并建立其和所在v3洋葱域名的关联关系,信息分析模块采用人工分析和关键词匹配相结合的方法分类v3洋葱域名,赋予其关联的比特币地址类别和流行度并判定威胁程度。实验结果表明,TLAFDB收集了23627个v3洋葱域名网页,提取并分析1141个存在真实交易的比特币地址的类别、流行度和威胁程度,发现在暗网中同一个比特币地址常出现在大量的镜像洋葱域名网页上,超过95%的比特币地址被恶意使用,并且庞氏骗局交易量占高危比特币地址总交易量的99%。

基于改进MobileNet v3的苹果叶片病害识别研究

为解决移动端和嵌入式设备中苹果叶片病害识别准确率不高、效率低下的问题,提出了一种新的基于MobileNet v3网络的分类模型,以实现更加高效和准确的苹果叶片病害识别。首先通过数据增广方法增强数据集,按照9∶1的比例划分训练集和验证集;然后在MobileNet v3网络核心倒残差结构的升维部分引入全维动态卷积,以加强对不同维度注意力权重的学习,从而增强网络的拟合能力;最后在降维部分引入修改后的ConvNext Block模块,减少信息损失并增加全局感受野。采用PyTorch作为分类网络的深度学习框架,使用交叉熵损失函数作为分类任务的损失函数,Adam作为优化器,通过多组对比试验可知,MobileNet v1、MobileNet v2、ResNet34、MobileNet v3以及改进后的MobileNet v3 ODConvNext网络的准确率分别为94.5%、95.7%、97.2%、96.9%及97.5%。可见,MobileNet v3 ODConvNet网络拥有最高的Top-1准确率,相较于MobileNet v3网络和结构更为复杂的ResNet34网络分别提升了0.6、0.3百分点;在运算频率方面,相对于MobileNet v3网络仅增加了1.00×10^(6)次/s,并且仅为ResNet34网络参数量的11.84%。因此,该试验结果证明了改进后的MobileNet v3 ODConvNext模型具有更加轻量级和更高准确率的优点,满足在移动端真实场景下进行苹果叶片病害识别的要求,有助于苹果叶片病害的防治工作。

The intracellular mechanism of alpha-fetoprotein promoting the proliferation of NIH 3T3 cells

AIM: The existence and properties of alpha-fetoprotein (AFP) receptor on the surface of NIH 3T3 cells and the effects of AFP on cellular signal transduction pathway were investigated. METHODS: The effect of AFP on the proliferation of NIH 3T3 cells was measured by incorporation of 3H-TdR. Receptor-binding assay of 125I-AFP was performed to detect the properties of AFP receptor in NIH 3T3 cells. The influences of AFP on the [cAMP]i and the activities of protein kinase A (PKA) were determined. Western blot was used to detect the change of K-ras P21 protein expression. RESULTS: The proliferation of NIH 3T3 cells treated with 0-80 mg/L of AFP was significantly enhanced. The Scatchard analysis indicated that there were two classes of binding sites with KD of 2.722 x 10(-9)M (Bmax=12810 sites per cell) and 8.931 x 10(-8)M (Bmax=119700 sites per cell) respectively. In the presence of AFP (20 mg/L), the content of cAMP and activities of PKA were significantly elevated . The level of K-ras P21 protein was upregulated by AFP at the concentration of 20 mg/L. The monoclonal antibody against AFP could reverse the effects of AFP on the cAMP content, PKA activity and the expression of K-ras p21 gene. CONCLUSION: The effect of AFP on the cell proliferation was achieved by binding its receptor to trigger the signal transduction pathway of cAMP-PKA and alter the expression of K- ras p21 gene.

基于DeepLab v3+的综放工作面含矸率预测研究

针对综放工作面真实煤矸堆叠状态下的体积含矸率很难获取的问题,提出一种基于DeepLab v3+的综放工作面含矸率预测方法。构建了煤矸堆积体图像数据集,采用半自动的数据标注方法和限制对比度自适应直方图均衡化法对煤矸图像进行预处理。运用DeepLab v3+模型进行煤矸图像语义分割,进而计算煤矸图像的投影面积含矸率。利用PFC3D数值模拟软件,基于重建的三维煤矸块体建立数值模型,模拟顶煤放落和刮板输送机运煤过程,通过fish语言读取每个矸石或煤的体积,计算得到煤矸堆积体体积含矸率。通过分析不同顶煤厚度条件下刮板输送机上煤矸堆积体的投影面积含矸率与体积含矸率的量化关系,建立了煤流的体积含矸率预测模型。实验结果表明:DeepLab v3+模型的准确率、平均像素准确率和平均交并比分别为97.68%,97.72%,95.33%,均高于经典语义分割模型FCN8s和PSPNet,实现了煤矸堆积体投影面积含矸率的精准快速识别;体积含矸率预测模型的决定系数R^(2)为0.9828,预测效果较好。

基于改进MobileNet v3的苹果叶片病害识别方法及移动端应用

准确识别苹果叶片病害种类以进行及时防治对于苹果增量增产具有重要的意义,为实现在移动设备实时对苹果叶片进行病害识别,提高苹果的产量,减少种植者的损失。首先收集了黑星病、斑点落叶病、锈病、白粉病、混合病、褐斑病等6种苹果叶部病害和健康叶片的图像,并使用Retinex算法对图像进行数据增强,以提高数据集质量,然后将数据集按照8∶1∶1的比例划分为训练集、验证集和测试集;其次对MobileNet v3网络模型进行改进优化调整,在精简网络结构的同时减少冗余参数,并在非线性激活层后加入批归一化层,以提高网络的特征提取能力;同时,为了提升在低精度移动设备上的准确性和模型运行效率,将全连接层中的激活函数替换为ReLU6函数;最后,在模型训练时使用动量随机梯度下降优化器来进行模型权重系数的寻优,以减少训练时间和达到更高的分类准确率。试验结果表明,改进后的MobileNet v3-A3网络对苹果叶片病害图像的识别准确率为96.48%,模型权重为2.98 MB,识别速率为8.82 ms/幅图片,与其他同量级卷积神经网络相比识别精度更高、模型更小、识别速度更快。本研究使用Android Studio将权重模型封装到安卓软件中,实现了移动设备对苹果叶片病害的准确快速识别。

Additively manufactured Ti–Ta–Cu alloys for the next-generation load-bearing implants

Bacterial colonization of orthopedic implants is one of the leading causes of failure and clinical complexities for load-bearing metallic implants. Topical or systemic administration of antibiotics may not offer the most efficient defense against colonization, especially in the case of secondary infection, leading to surgical removal of implants and in some cases even limbs. In this study, laser powder bed fusion was implemented to fabricate Ti3Al2V alloy by a 1:1 weight mixture of CpTi and Ti6Al4V powders. Ti-Tantalum(Ta)–Copper(Cu) alloys were further analyzed by the addition of Ta and Cu into the Ti3Al2V custom alloy. The biological,mechanical, and tribo-biocorrosion properties of Ti3Al2V alloy were evaluated. A 10 wt.% Ta(10Ta) and 3 wt.% Cu(3Cu) were added to the Ti3Al2V alloy to enhance biocompatibility and impart inherent bacterial resistance. Additively manufactured implants were investigated for resistance against Pseudomonas aeruginosa and Staphylococcus aureus strains of bacteria for up to 48 h. A 3 wt.% Cu addition to Ti3Al2V displayed improved antibacterial efficacy, i.e.78%–86% with respect to CpTi. Mechanical properties for Ti3Al2V–10Ta–3Cu alloy were evaluated, demonstrating excellent fatigue resistance, exceptional shear strength, and improved tribological and tribo-biocorrosion characteristics when compared to Ti6Al4V. In vivo studies using a rat distal femur model revealed improved early-stage osseointegration for alloys with10 wt.% Ta addition compared to CpTi and Ti6Al4V. The 3 wt.% Cu-added compositions displayed biocompatibility and no adverse infammatory response in vivo. Our results establish the Ti3Al2V–10Ta–3Cu alloy’s synergistic effect on improving both in vivo biocompatibility and microbial resistance for the next generation of load-bearing metallic implants.

基于改进YOLO v3的轴承端面缺陷检测算法

为提高轴承端面缺陷检测的速度以及检测精度,提出一种基于改进YOLO v3的轴承端面缺陷检测算法。首先,对图像数据集进行数据增强处理以防止产生过拟合现象;其次,通过改进K-means聚类算法重新聚类出目标检测的Anchor Boxes,并引入SKNet注意力机制模块对原网络结构以及输出层结构进行改进;最后对改进的YOLO v3算法进行实验验证,并与原YOLO v3算法进行对比分析。结果表明,改进后的YOLO v3算法相比原YOLO v3算法对轴承端面缺陷检测的mAP值提升了7.03%,检测速度提升了34.7帧/s,验证了改进算法的有效性。

基于改进MobileNet v3-Small模型的草莓病害识别方法

为了对草莓病害进行及时的诊断与治疗而提升草莓产量,将深度学习与农业生产结合以快速高效地进行病害检测。传统神经网络进行病害识别时间较长,参数量较大,难以迁移到移动端设备上,基于此提出一种改进MobileNet v3-Small模型的识别方法。首先收集了7类常见草莓病害图像样本(如角斑病、叶斑病等),通过旋转、镜像等多种数据增强方式对图像进行处理以增加图片数量,提高模型泛化能力。接着以MobileNet v3-Small模型为基础,基于原始Inception_A提出部分卷积权值共享的多尺度卷积结构,以更高效地提取草莓病害不同尺度特征。随后,在网络深层引入了ULSAM轻量级子注意力机制,形成草莓病害更高层次的抽象表示。同时,将深度可分离卷积中的第2个PW卷积替换为CondConv卷积形成PDC结构,克服了PW卷积只拥有局部感受野的缺陷,同时也降低了模型参数量。试验结果表明,改进后的MobileNet v3-Small模型准确率达到98.62%,较原模型94.91%的准确率提高了3.71百分点,并且参数量减少了0.04 M,远优于同级轻量化模型,且以远低于ResNet18的参数量取得更好的特征提取效果。综上所述,本研究所提出的改进后的MobileNet v3_Small模型能更好地在真实场景下进行草莓病害识别,为草莓生产贡献了一份力量,助力智慧农业发展。

Co8W6Mo5Cr4V3粉末高速钢的热变形行为

采用熔炼、电渣重熔和气雾化法制备Co8W6Mo5Cr4V3粉末,再通过热等静压制备Co8W6Mo5Cr4V3粉末高速钢。在不同变形条件下进行热压缩实验,研究粉末高速钢的热变形行为,根据获得的应力-应变曲线,建立热变形本构方程,绘制热加工图。结果表明:Co8W6Mo5Cr4V3粉末高速钢的流变应力随变形温度升高或应变速率减小而减小;通过本构方程预测的流变应力和实验得到的流变应力间的相关系数为0.995,可用此方程对粉末高速钢在不同变形条件下的应力进行预测;根据热加工图,粉末高速钢最优的变形温度为1 100~1 150℃,应变速率为0.1~1.0 s^(-1)。

基于DeepLab V3+估测小龙虾虾头虾钳占比及分级

目的:完善小龙虾分级工作。方法:搭建小龙虾图像拍摄平台,获取小龙虾原始图像,创建分割虾头、虾钳和虾尾3个部位的语义分割数据集。分析小龙虾虾头、虾钳、虾尾3个部位实际质量与数据集中对应部位像素大小之间的相关性,总结得到根据整虾中虾头虾钳占比进行分级的小龙虾分级新标准。使用数据集训练DeepLab V3+神经网络,并用测试集检验模型语义分割效果以及小龙虾分级的准确率,语义分割评价指标为平均交并比(MIoU)、平均像素准确率(MPA)和像素准确率(PA)。结果:小龙虾语义分割测试集的MIoU为94.35%,MPA为96.56%,PA为99.44%,测试集小龙虾分级准确率为85.56%。结论:DeepLab V3+模型可以准确分割小龙虾图像并估测虾头虾钳占比,模型能够完成小龙虾分级任务。

Preparation of Na_(3)V_(2)(PO_(4))_(3) Cathode Materials by Hydrothermal Assisted Sol-Gel Method for Sodium -Ion Batteries

Na_(3)V_(2)(PO_(4))_(3)(NVP)cathode material of the sodium ion battery(1 C=117 mAh g-1)has a NASICON-type structure,which not only facilitates the rapid migration of sodium ions,but also has a small volume deformation during sodium ion de-intercalation and the main frame mechanism remains unchanged,and thus is seen as an energy storage material for a wide range of applications,but has a limited electronic conductivity due to its structure.In this paper,NVP cathode materials with finer primary particles are successfully prepared using a simple hydrothermal treatment-assisted sol-gel method.The increased pore size of the NVP materials prepared under the hydrothermal process allows for more active sites and more effective resistance to the volume deformation of sodium ions during insertion/extraction processes,effectively facilitating the diffusion of ions and electrons.The Na_(3)V_(2)(PO_(4))_(3) material obtained by the optimized process exhibited good crystallinity in XRD characterization,as well as superior electrochemical properties in a series of electrochemical tests.A specific capacitance of 106.3 mAh g^(-1) at 0.2 C is demonstrated,compared to 96.5 mAh g^(-1) for Na_(3)V_(2)(PO_(4))_(3) without hydrothermal treatment,and cycling performance is also improved with 93%capacity retention.The calculated sodium ion diffusion coefficient(DNa=5.68×10^(-14))obtained after EIS curve fitting of the improved sample illustrates that the pore structure is beneficial to the performance of the Na_(3)V_(2)(PO_(4))_(3)cathode material.

基于改进DeepLab v3+的耕地自动化提取方法

耕地是地表覆盖的主要类型之一,也是粮食生产的直接载体。本文针对耕地提提取不准确和效率低的问题,提出一种改进的DeepLab v3+提取方法。首先,使用轻量级网络替代DeepLab v3+模型的特征提取网络Xception,以减少模型参数量,提高训练速度。其次,引用SE注意力加入模型,提高网络对耕地的提取精度。并在GID数据集上进行试验,结果显示,本文方法具有优秀的分割性能,能够有效解决经典模型分割结果中存在的不同程度的误分、漏分等问题,在耕地提取中具有高效性,为后续耕地提取任务提供支持。

Real-Time Monitoring Method for Cow Rumination Behavior Based on Edge Computing and Improved MobileNet v3

[Objective]Real-time monitoring of cow ruminant behavior is of paramount importance for promptly obtaining relevant information about cow health and predicting cow diseases.Currently,various strategies have been proposed for monitoring cow ruminant behavior,including video surveillance,sound recognition,and sensor monitoring methods.How‐ever,the application of edge device gives rise to the issue of inadequate real-time performance.To reduce the volume of data transmission and cloud computing workload while achieving real-time monitoring of dairy cow rumination behavior,a real-time monitoring method was proposed for cow ruminant behavior based on edge computing.[Methods]Autono‐mously designed edge devices were utilized to collect and process six-axis acceleration signals from cows in real-time.Based on these six-axis data,two distinct strategies,federated edge intelligence and split edge intelligence,were investigat‐ed for the real-time recognition of cow ruminant behavior.Focused on the real-time recognition method for cow ruminant behavior leveraging federated edge intelligence,the CA-MobileNet v3 network was proposed by enhancing the MobileNet v3 network with a collaborative attention mechanism.Additionally,a federated edge intelligence model was designed uti‐lizing the CA-MobileNet v3 network and the FedAvg federated aggregation algorithm.In the study on split edge intelli‐gence,a split edge intelligence model named MobileNet-LSTM was designed by integrating the MobileNet v3 network with a fusion collaborative attention mechanism and the Bi-LSTM network.[Results and Discussions]Through compara‐tive experiments with MobileNet v3 and MobileNet-LSTM,the federated edge intelligence model based on CA-Mo‐bileNet v3 achieved an average Precision rate,Recall rate,F1-Score,Specificity,and Accuracy of 97.1%,97.9%,97.5%,98.3%,and 98.2%,respectively,yielding the best recognition performance.[Conclusions]It is provided a real-time and effective method for monitoring cow ruminant behavior,and the proposed federated edge intelligence model can be ap‐plied in practical settings.

Milk fat globule epidermal growth factor 8 alleviates liver injury in severe acute pancreatitis by restoring autophagy flux and inhibiting ferroptosis in hepatocytes

BACKGROUND Liver injury is common in severe acute pancreatitis(SAP).Excessive autophagy often leads to an imbalance of homeostasis in hepatocytes,which induces lipid peroxidation and mitochondrial iron deposition and ultimately leads to ferroptosis.Our previous study found that milk fat globule epidermal growth factor 8(MFG-E8)alleviates acinar cell damage during SAP via binding toαvβ3/5 integrins.MFG-E8 also seems to mitigate pancreatic fibrosis via inhibiting chaperone-mediated autophagy.AIM To speculate whether MFG-E8 could also alleviate SAP induced liver injury by restoring the abnormal autophagy flux.METHODS SAP was induced in mice by 2 hly intraperitoneal injections of 4.0 g/kg L-arginine or 7 hly injections of 50μg/kg cerulein plus lipopolysaccharide.mfge8-knockout mice were used to study the effect of MFG-E8 deficiency on SAPinduced liver injury.Cilengitide,a specificαvβ3/5 integrin inhibitor,was used to investigate the possible mechanism of MFG-E8.RESULTS The results showed that MFG-E8 deficiency aggravated SAP-induced liver injury in mice,enhanced autophagy flux in hepatocyte,and worsened the degree of ferroptosis.Exogenous MFG-E8 reduced SAP-induced liver injury in a dose-dependent manner.Mechanistically,MFG-E8 mitigated excessive autophagy and inhibited ferroptosis in liver cells.Cilengitide abolished MFG-E8’s beneficial effects in SAP-induced liver injury.CONCLUSION MFG-E8 acts as an endogenous protective mediator in SAP-induced liver injury.MFG-E8 alleviates the excessive autophagy and inhibits ferroptosis in hepatocytes by binding to integrinαVβ3/5.

新课标导向下的五年级下学期3v3篮球大单元教学设计方案

本文为五年级下学期3v3篮球大单元学习方案,以“篮球基本技术”为大概念引领整个单元教学。教学设计围绕核心素养培育、“教会、勤练、常赛”等理念,教学过程中以“移动技术、传球、运球、投篮、突破、篮球基础配合”等基本技术为大任务驱动整个单元教学,单元学习过程中为学生创设各类比赛的真实情境,提高学生运用篮球基本技术的能力。最终,通过多种形式的“学、练、赛”提高学生的篮球水平,培育学生的核心素养。

Evolutionary Patterns of the Proviral gp90 V3 to V5 Regions of Equine Infectious Anemia Virus Associated with Immune Selection in Progressors and Nonprogressors

The aim of this study was to determine the genomic evolutionary pattern of virulent equine infectious anemia virus （EIAV） during persistent infection. The evolutionary dynamics of proviral genomes were examined by challenging an EIAV seronegative equine （pony 1） and three EIAV vaccinated equines （ponies 4, 7, and 8） with the Chinese virulent strain EIAV- L. Ponies 1 and 7 succumbed to disease and were called progressors, while ponies 4 and 8 lacked clinical symptoms and were considered nonprogressors. Sequences spanning the V3, V4, and V5 hyper-variable regions of the EIAV-L envelope gp90 gene were sequenced from each pony as evolutionary markers of the provirus. The proviral genome of the EIAV-L inoculum evolved during persistent infection and displayed different patterns between EIA progressors and nonprogressors. Inoculum-like variants were isolated from nonprogressors during persistent infection, but only from progressors during acute infection. Variant mutations from nonprogressors were dispersed throughout the sequenced region, while those from progressors were predominantly localized to V3. Humoral immunity and virus variant population selection analyses indicated that immune selection was positive in chronically infected progressors and weak in nonprogressors. In-frame stop codons were frequently localized to a defect ＂hot spot＂. The high number of defective variants in nonprogressors may promote disease survival.

Wear Properties of Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr Alloys under Different Atmospheres

The study examines the friction and wear properties of Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr alloys under varying gas conditions.The alloy was tested in the presence of oxygen and carbon dioxide using a controlled atmosphere wear tester.The study revealed that the wear environ‐mental embrittlement resulted from the diffusion of reactive atomic hydrogen into the interior of the Ni_(3)Al-Ni_(3)V alloy.The addition of Zr elements decreased the proportion of Al elements on the surface of the Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr alloy and reduced the proportion of H atoms produced by the chemical reaction between atmospheric water vapour and Al elements.This inhibited the environmental embrittlement and improved the performance of the Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr alloy.The wear performance of Ni_(5)Zr alloy is superior to that of Ni_(3)Al-Ni_(3)V.When exposed to air in an air environment,the surface of Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr alloy forms a protective Al_(2)O_(3) oxide film on the workpiece,result‐ing in a reduction of the friction coefficient and wear rate of the alloy.The wear mechanism of the alloy is mainly oxidation wear and abra‐sive wear.In an oxygen environment,the surface of the alloy generates a significant amount of Al_(2)O_(3) oxide film.The flaking of the oxide film leads to an increase in the friction coefficient and wear rate of the alloy.In a carbon dioxide environment,the surface of the alloy un‐dergoes severe deformation,and plough lines become apparent.This is accompanied by flaking Si_(3)N_(4) abrasive chips adhering to the sur‐face of the alloy,which intensifies the wear of the alloy.The primary wear mechanism is abrasive wear.Therefore,the friction coefficient and wear rate of the Ni_(3)Al-Ni_(3)V-Zr-Ni_(5)Zr alloy in the atmosphere are optimal.