基于CNN-BiLSTM-Transformer的舰船中压直流全电推进系统故障诊断设计

针对舰船中压直流全电推进系统结构复杂度高、单元耦合性强、运行环境多变等特点造成的故障诊断准确性低和实时性差等问题,开展了基于CNN-BiLSTM-Transformer的故障诊断设计。首先,基于卷积神经网络CNN构建单点特征级联网络,开展单一时刻下故障信号空间特征的深入提取,以提升故障特征提取的有效性;其次,以双向长短期记忆网络BiLSTM为核心设计多点特征依赖网络,利用门控机制和双向时序学习机制,实现故障信号在多个时刻之间特征依赖关系的有效学习,以提升故障诊断的准确性;然后,以Transformer为核心建立序列特征并行处理网络,通过自注意力机制实现对故障特征上下文关系的精确刻画,进而利用多头注意力机制实现特征序列的并行处理,以提升故障诊断的实时性;最后,设计舰船中压直流全电推进系统故障诊断实验方案,并开展不同故障模式下的诊断性能评估。该文方法在多种故障模式下诊断准确率和实时性均优于现有的主流故障诊断方法,有助于为舰船中压直流全电推进系统的安全运行提供更有力的技术保障。

基于双域Transformer耦合特征学习的CT截断数据重建模型

目的为解决CT扫描视野(FOV)不足导致的截断伪影和图像结构失真问题,本文提出了一种基于投影和图像双域Transformer耦合特征学习的CT截断数据重建模型(DDTrans)。方法基于Transformer网络分别构建投影域和图像域恢复模型,利用Transformer注意力模块的远距离依赖建模能力捕捉全局结构特征来恢复投影数据信息,增强重建图像。在投影域和图像域网络之间构建可微Radon反投影算子层,使得DDTrans能够进行端到端训练。此外,引入投影一致性损失来约束图像前投影结果,进一步提升图像重建的准确性。结果Mayo仿真数据实验结果表明,在部分截断和内扫描两种截断情况下,本文方法DDTrans在去除FOV边缘的截断伪影和恢复FOV外部信息等方面效果均优于对比算法。结论DDTrans模型可以有效去除CT截断伪影,确保FOV内数据的精确重建,同时实现FOV外部数据的近似重建。

Transforming growth factor-beta 1 enhances discharge activity of cortical neurons

Transforming growth factor-beta 1(TGF-β1)has been extensively studied for its pleiotropic effects on central nervous system diseases.The neuroprotective or neurotoxic effects of TGF-β1 in specific brain areas may depend on the pathological process and cell types involved.Voltage-gated sodium channels(VGSCs)are essential ion channels for the generation of action potentials in neurons,and are involved in various neuroexcitation-related diseases.However,the effects of TGF-β1 on the functional properties of VGSCs and firing properties in cortical neurons remain unclear.In this study,we investigated the effects of TGF-β1 on VGSC function and firing properties in primary cortical neurons from mice.We found that TGF-β1 increased VGSC current density in a dose-and time-dependent manner,which was attributable to the upregulation of Nav1.3 expression.Increased VGSC current density and Nav1.3 expression were significantly abolished by preincubation with inhibitors of mitogen-activated protein kinase kinase(PD98059),p38 mitogen-activated protein kinase(SB203580),and Jun NH2-terminal kinase 1/2 inhibitor(SP600125).Interestingly,TGF-β1 significantly increased the firing threshold of action potentials but did not change their firing rate in cortical neurons.These findings suggest that TGF-β1 can increase Nav1.3 expression through activation of the ERK1/2-JNK-MAPK pathway,which leads to a decrease in the firing threshold of action potentials in cortical neurons under pathological conditions.Thus,this contributes to the occurrence and progression of neuroexcitatory-related diseases of the central nervous system.

基于Transformer的肺肿瘤三维CT图像分割

基于信息学技术自动分割病人的肺部CT图像,有助于医生对于肺癌患者的早期诊断,提取和整合图像区域间的空间关联,对于提升肺肿瘤分割性能是十分重要的。本文提出了一个新的基于Transformer的分割模型,用于肺肿瘤三维CT图像分割、学习和整合此类关联。本文分别设计了带有混合多头图像区域节点注意力的Transformer模块和类别注意力模块,学习并融合了肺部CT图像的空间层面和通道层面的信息。将新的基于Transformer的分割模型同其他较为先进的模型进行了对比实验,实验结果表明新的模型在骰子系数、交并比和豪斯多夫距离等方面优于其他模型。

MCIF-Transformer Mask RCNN:Multi-Branch Cross-Scale Interactive Feature Fusion Transformer Model for PET/CT Lung Tumor Instance Segmentation

The precise detection and segmentation of tumor lesions are very important for lung cancer computer-aided diagnosis.However,in PET/CT(Positron Emission Tomography/Computed Tomography)lung images,the lesion shapes are complex,the edges are blurred,and the sample numbers are unbalanced.To solve these problems,this paper proposes a Multi-branch Cross-scale Interactive Feature fusion Transformer model(MCIF-Transformer Mask RCNN)for PET/CT lung tumor instance segmentation,The main innovative works of this paper are as follows:Firstly,the ResNet-Transformer backbone network is used to extract global feature and local feature in lung images.The pixel dependence relationship is established in local and non-local fields to improve the model perception ability.Secondly,the Cross-scale Interactive Feature Enhancement auxiliary network is designed to provide the shallow features to the deep features,and the cross-scale interactive feature enhancement module(CIFEM)is used to enhance the attention ability of the fine-grained features.Thirdly,the Cross-scale Interactive Feature fusion FPN network(CIF-FPN)is constructed to realize bidirectional interactive fusion between deep features and shallow features,and the low-level features are enhanced in deep semantic features.Finally,4 ablation experiments,3 comparison experiments of detection,3 comparison experiments of segmentation and 6 comparison experiments with two-stage and single-stage instance segmentation networks are done on PET/CT lung medical image datasets.The results showed that APdet,APseg,ARdet and ARseg indexes are improved by 5.5%,5.15%,3.11%and 6.79%compared with Mask RCNN(resnet50).Based on the above research,the precise detection and segmentation of the lesion region are realized in this paper.This method has positive significance for the detection of lung tumors.

Segmentation of Head and Neck Tumors Using Dual PET/CT Imaging:Comparative Analysis of 2D,2.5D,and 3D Approaches Using UNet Transformer

The segmentation of head and neck(H&N)tumors in dual Positron Emission Tomography/Computed Tomogra-phy(PET/CT)imaging is a critical task in medical imaging,providing essential information for diagnosis,treatment planning,and outcome prediction.Motivated by the need for more accurate and robust segmentation methods,this study addresses key research gaps in the application of deep learning techniques to multimodal medical images.Specifically,it investigates the limitations of existing 2D and 3D models in capturing complex tumor structures and proposes an innovative 2.5D UNet Transformer model as a solution.The primary research questions guiding this study are:(1)How can the integration of convolutional neural networks(CNNs)and transformer networks enhance segmentation accuracy in dual PET/CT imaging?(2)What are the comparative advantages of 2D,2.5D,and 3D model configurations in this context?To answer these questions,we aimed to develop and evaluate advanced deep-learning models that leverage the strengths of both CNNs and transformers.Our proposed methodology involved a comprehensive preprocessing pipeline,including normalization,contrast enhancement,and resampling,followed by segmentation using 2D,2.5D,and 3D UNet Transformer models.The models were trained and tested on three diverse datasets:HeckTor2022,AutoPET2023,and SegRap2023.Performance was assessed using metrics such as Dice Similarity Coefficient,Jaccard Index,Average Surface Distance(ASD),and Relative Absolute Volume Difference(RAVD).The findings demonstrate that the 2.5D UNet Transformer model consistently outperformed the 2D and 3D models across most metrics,achieving the highest Dice and Jaccard values,indicating superior segmentation accuracy.For instance,on the HeckTor2022 dataset,the 2.5D model achieved a Dice score of 81.777 and a Jaccard index of 0.705,surpassing other model configurations.The 3D model showed strong boundary delineation performance but exhibited variability across datasets,while the 2D model,although effective,generally underperformed compared to its 2.5D and 3D counterparts.Compared to related literature,our study confirms the advantages of incorporating additional spatial context,as seen in the improved performance of the 2.5D model.This research fills a significant gap by providing a detailed comparative analysis of different model dimensions and their impact on H&N segmentation accuracy in dual PET/CT imaging.

Effect of Modulation Error on All Optical Fiber Current Transformers

For actively modulated In-line Sagnac interferential all optic fiber current transformers (AOFCTs), the accuracies are directly affected by the amplitude of the modulation signal. In order to deeply undertand the function of the modulator, a theoretical model of modulation effect to AOFCTs is built up in this paper. The effect of the amplitude of the modulation signal to the output intensity of AOFCTs is theoretically formulated and numerical calculated. The results show that the modulation voltage variation could affect the output accuracies significantly. This might be some references on the investigation for practical applications of AOFCTs.

An active-passive beam current transformer

In this article,a new type current transformer was developed using an active-passive circuit to improve low frequency response of the system without impairing high frequency response.The active-passive current transformer with In-flange was fabricated.Theoretical analysis,numerical simulations and experimental results are given.

Analysis on the Transfer Characteristics of Rogowski-coil Current Transformer and Its Influence on Protective Relaying

This paper systematically analyzes the transfer characteristics of the Rogowski-coil Current Transformer and its effect on protective relaying through theoretical analysis, experiments and simulations. The frequency characteristics and transient characteristics of Rogowski transducer and Rogowski-coil Current Transformer are deeply analyzed based on the physical structure of the transformer.?It is revealed that broad bandwidth of the transformer can improve the performance of protective relaying, and the bandwidth is determined mainly by the parameters of the Rogowski transducer and signal processing circuits. It is also discovered that the measurement errors of transient current mainly depend on the abilities for the current transformer to reproduce an accurate replica of the decaying dc components, which is mainly decided by the decay time constant of the aperiodic component of transient current and the parameters of the integral unit. Finally, some measures are proposed for the performance improvement of Rogowski-coil Current Transformer to meet the requirements of protective relaying system in terms of structural design and testing standards.

Optical fiber-based power and data delivery system for high voltage electronic current transformer

This paper describes a prototype power delivery system developed for high voltage electronic current transformer （ECT） that uses laser light to transfer power to and communicates with the primary converter. The design is based on optical-to-electrical power converters, solid-state diode lasers and optical fibers. Command signals are transmitted via the same up-fiber used to send power from secondary power supply to primary converter. The upward data transmission is completed during the brief interruption of power delivery without affecting steady power-supply. A simple comparator added to the primary converter can take the command data. Experimental results show that the fibers can provide reliable up-link for data transmission at 200 kb/s from the secondary to the primary converter. Based on the delivery system, the secondary converter can control three auxiliary channels to provide additional information. These monitoring channels are used in a time-multiplexing mode to provide information about the operation temperature, voltage and current at the remote unit for monitoring the ECT. This preventive maintenance or built-in test can increase reliability by giving early warning for necessary maintenance request.

Transformer Internal and Inrush Current Fault Detection Using Machine Learning

Preventive maintenance in the transformer is performed through a dif-ferential relay protection system,and it protects the transformer from internal and external faults.However,the Current Transformer(CT)in the differential protec-tion system mal-operates during inrush currents.CT saturates due to magnetizing inrush currents and causes false tripping of the differential relays.Moreover,iden-tification of tripping in protection relay either due to inrush current or internal faults needs to be diagnosed.For the above problem,continuous monitoring of transformer breather and CT terminals with thermal camera helps detect the trip-ping in relay due to inrush or internal fault.The transformer’s internal fault leads to high breathing process in the transformer breather,never for inrush currents.During inrush currents,CT temperature is increased.Continuous monitoring of breather and CT of the transformer through thermal imaging and radiometric pix-els detect the causes of CT saturation and differentiates maloperation.Hybrid wavelet threshold image analytics(HWT-IA)based radiometric pixels analysis of the transformer breather and CT after de-noising provides an accurate result of about 95%for identification of the false tripping of differential protection system of transformer.

Study on the Effect of Multi-Type Current Transformers Hybrid Operation on Differential Protection of the Bus Based on Dynamic Simulation

This paper analyzes characteristics of multi type current transformers hybrid operation for each branch of the bus and their effects on differential protection of the bus. By theoretically analyzing transmission characteristics of multi type current transformers and their influence factors, we study the dynamic model testing method of multi type current transformers for the bus, and design 3 kinds of testing schemes by making the equivalent model based on the field of P-level current transformer, TPY-level current transformer and electronic current transformer, and build the hybrid operation testing platform of multi type current transformers. Finally, we compare and analyze the transmission characteristics difference of multi type current transformers on the same branch and the characteristics difference of hybrid operation in two successive external faults, analyze the cause behind the differences, and put forward the corresponding improvement measures.

Solutions to Consider in Current Transformer Selection for APR1400 Nuclear Power Plant Medium Voltage Switchgears

This paper discusses a preferable solution to mitigate the CT （current transformer） saturation problem, and at same time, reduce the accuracy errors when considering the selection of CTs for installation on the medium voltage switchgear of a nuclear power plant. This consideration is important for both measurement and protection functions of the digital protective relays. This is a study to ascertain the best options for a suitable solution to prevent CT saturation in relations to its protective capabilities during short circuit fault without compromising the CT accuracy class during normal operation of the system, while ensuring its conformity to the design requirement is within limit. The advantages of current transformers have proven not only to be reliable and safe, but also are of easy handling, reduction of the cost and components on the MV （medium voltage） switchgear. The purpose of this research is to identify best approach to resolve the existing problems in the current protection system. With the view of LPCT （low power current transformer） which has been newly constructed by few manufacturers to provide good protection and a wide range of measuring function without errors, some other solutions will be considered in this research.

APPLICATION OF NEURAL NETWORK FOR THE TRANSFORMER PROTECTIVE RELAYS

A neural network method used to identify the different operating states of transformers has been proposed and established.It is superior to the traditional transformer protective principles and can correctly identify,within half cycle from the fault inception,the internal faults,magnetizing inrush current state,external faults and switching on the internal faults of a no load transformer.In addition,this method has broad availability and high fault tolerant ability.A lot of simulations have demonstrated its superiority.

SIMULATING METHOD OF MAGNETIZING INRUSH CURRENT OF POWER TRANSFORMERS USING CONCEPT OF INSTANTANEOUS POWER

In this paper,a new simulating method is presented,using only the normal magnetizing curve (B-H) of the transformer core material,its geometric dimensions,the no-load power loss data and the concept of instantaneous power. At the end of this paper the simulating calculation using EMTP has been also performed for the same transformer. The comparison shows that the two sets of results are very close to each other,and proves the correctness of the new method. The new method presented in this paper is helpful to verify the correctness of the power transformer design,analyze the behavior of the transformer protection under switching and study the new transformer protection principles.

基于增强CT图像和Swin Transformer网络的食管癌T分期智能诊断模型的构建与评估

目的基于增强CT图像和Swin Transformer网络,拟构建食管癌T分期智能诊断模型。方法收集2018年1月至2022年4月在陆军军医大学第一附属医院和山西省肿瘤医院胸外科经病理证实为食管癌的150例患者的45000张术前增强CT图像。经过UperNet Swin网络自动分割和肿瘤体积的计算,使用ResNet50、Swin Transformer和VIT 3个网络进行食管癌T分期智能诊断模型的构建。使用精准率、召回率、F1-score、特异度以及阴性预测值(negative predictive value,NPV)等指标在150例内部数据集上评价模型性能,描绘混淆矩阵和ROC曲线。结果在3个食管癌T分期诊断的模型中,Swin Transformer模型结合肿瘤体积、病理信息等特征的分期诊断效果最好,T1~T4期的精准率分别为1.00、0.67、0.83、1.00,AUC为0.861,优于ResNet50和VIT分期诊断模型,它们的精准率分别为0.13、0.27、0.59、0.81和0.03、0.14、0.56、0.75,AUC分别是0.611和0.542。结论与ResNet50和VIT网络比较,Swin Transformer网络能够更精准进行食管癌智能T分期诊断。

基于Transformer的脊柱CT图像分割

脊柱侧弯是青少年群体常见的一种脊柱疾病,老年群体中因脊柱骨质疏松引起的脊柱骨折也尤为普遍。CT成像技术作为脊柱外科的主要检查手段之一,广泛用于临床以及研究目的的筛查,诊断和图像引导治疗。研究以脊柱CT图像为研究对象,将目前在NLP领域表现优异的Transformer模型与经典的U-Net图像分割网络相结合,运用到CT图像的分割处理工作当中;同时在模型训练过程中基于脊柱自身的结构特点,采用由粗到精的训练方法,首先对脊柱的各个椎骨进行定位模型训练,然后在定位结果的基础上再训练分割模型。最终模型的分割结果与真实值之间的Dice相似系数达到了94.37%以上,实验结果表明了该方法的有效性以及临床应用的可行性。

Study of wavelet transform type high-current transformer

The wavelet transformation is applied to the high current transformer.The high current transformer elaborated in the paper is mainly applied to the measurement of AC/DC high current.The principle of the transformer is the Hall direct measurement principle.The transformer has the following three characteristics:firstly, the effect of the remnant field of the iron core on the measurement is decreased;secondly,because the temperature compensation is adopted,the transformer has good temperature charactreristic;thirdly,be cause the wavelet transfomation technology is adopted,the transformer has the capacity of good antijanming.

基于Transformer和CNN的低剂量CT图像去噪网络

低剂量计算机断层扫描(Low-dose Computed Tomography, LDCT)在临床中有着广泛的应用,可以有效减轻对病人的辐射剂量。但是成像后的低剂量CT图像中含有明显的噪声和条形伪影,影响医师的诊断。提出了一种基于Transformer和CNN的去噪网络,该网络是一种改进的编解码网络架构,其编码端的每一层由卷积模块与Transformer模块融合而成,用来提取每一层的局部特征和全局特征,同时引入融合模块用来有效地融合提取的局部特征和全局特征。并把融合后的特征通过跳跃连接融入解码端对应的层,解码端的每一层通过卷积模块提取有效特征进而重建去噪后的图像。在真实数据集Mayo上的实验结果说明所提出的网络不仅可以有效去除噪声,还能够保持图像的边缘。

Swin-Transformer故障信息挖掘的海底观测网故障定位方法

海底观测网长期受海洋环境与人为因素影响,易使光电复合缆绝缘破损与海水接触形成电学故障点。如何准确地定位电学故障点,对提高海底观测网输电与信息传输的可靠性至关重要。首先根据海底观测网输电结构建立海底观测网输电模型,推导与模拟电学故障点传播至观测点的暂态电流,然后由连续小波变换提取暂态电流与故障点对应的内在关联特征量,最后通过Swin-Transformer神经网络挖掘内在关联特征量与故障距离的匹配关系来定位电学故障点。研究结果表明,在内在关联特征量样本测试集条件下,光电复合缆≤160 km的电学故障点定位误差小于400 m,可为长距离光电复合缆的海底观测网电学故障点定位提供参考。