融合CNN和ViT的声信号轴承故障诊断方法

针对轴承故障诊断任务数据量少、故障信号非平稳等特点,提出一种短时傅里叶变换、卷积神经网络和视觉转换器相结合的轴承故障诊断方法。首先,利用短时傅里叶变换将原始声信号转换为包含时序信息和频率信息的时频图像。其次,将时频图像作为卷积神经网络的输入,用于隐式提取图像的深层特征,其输出作为视觉转换器的输入。视觉转换器用于提取信号的时间序列信息。并在输出层利用Softmax函数实现故障模式的识别。试验结果表明,该方法对于轴承故障诊断准确率较高。为了更好解释和优化提出的轴承故障诊断方法,利用t-分布领域嵌入算法对分类特征进行了可视化展示。

基于改进INFO-Bi-LSTM模型的SO_(2)排放质量浓度预测

针对火电机组SO_(2)排放质量浓度的影响因素众多,难以准确预测的问题,提出一种改进向量加权平均(weighted mean of vectors,INFO)算法与双向长短期记忆(bi-directional long short term memory,Bi-LSTM)神经网络相结合的预测模型(改进INFO-Bi-LSTM模型)。采用Circle混沌映射和反向学习产生高质量初始化种群,引入自适应t分布提升INFO算法跳出局部最优解和全局搜索的能力。选取改进INFO-Bi-LSTM模型和多种预测模型对炉内外联合脱硫过程中4种典型工况下的SO_(2)排放质量浓度进行预测,将预测结果进行验证对比。结果表明:改进INFO算法的寻优能力得到提升,并且改进INFO-Bi-LSTM模型精度更高,更加适用于SO_(2)排放质量浓度的预测,可为变工况下的脱硫控制提供控制理论支撑。

An Integrated Use of Advanced T2 Statistics and Neural Network and Genetic Algorithm in Monitoring Process Disturbance

Integrated use of statistical process control (SPC) and engineering process control (EPC) has better performance than that by solely using SPC or EPC. But integrated scheme has resulted in the problem of “Window of Opportunity” and autocorrelation. In this paper, advanced T2 statistics model and neural networks scheme are combined to solve the above problems: use T2 statistics technique to solve the problem of autocorrelation;adopt neural networks technique to solve the problem of “Window of Opportunity” and identification of disturbance causes. At the same time, regarding the shortcoming of neural network technique that its algorithm has a low speed of convergence and it is usually plunged into local optimum easily. Genetic algorithm was proposed to train samples in this paper. Results of the simulation ex-periments show that this method can detect the process disturbance quickly and accurately as well as identify the dis-turbance type.

Numerical simulation of rock pore-throat structure effects on NMR T_2 distribution

We built a three-dimensional irregular network model which can adequately describe reservoir rock pore-throat structures. We carried out numerical simulations to study the NMR T2 distribution of water-saturated rocks. The results indicate that there is a good correlation between T2 distribution and the pore radius frequency histogram. The total T2 distribution can be partitioned into pore body and pore throat parts. The effect of parameters including throat radius, pore-throat ratio, and coordination number of the micro- pore structure on the T2 distribution can be evaluated individually. The result indicates that： 1 ） with the increase of the pore throat radius, the T2 distribution moves toward longer relaxation times and its peak intensity increases; 2） with the increase of the pore-throat ratio, the T2 distribution moves towards longer T2 with the peak intensity increasing and the overlap between pore body T2 and pore throat T2 decreasing; 3） With the increase of connectivity, the short T2 component increases and peak signal intensity decreases slightly.

区间二型T-S模型网络系统的安全控制设计

研究了具有参数不确定的区间二型模糊网络系统在遭受到拒绝服务攻击时的控制器设计问题。通过选取合适的Lyapunov-Krasovskii泛函,综合应用自由权矩阵、凸组合、Jensen不等式的方法并结合现有处理拒绝服务攻击的方式,设计了弹性事件触发策略下的安全控制器,并通过数例仿真验证设计方法的有效性。

区间二型T-S模型网络控制系统的鲁棒H_∞控制

针对一类具有网络诱导时滞和参数不确定的网络控制系统,研究鲁棒H_∞控制器设计问题。基于区间二型T-S模型,通过构造合适的Lyapunov-Krasovskii泛函,引入自由权矩阵来表示Newton-Leibniz公式中各项之间的关系,加入一些附加项,运用不等式放缩和矩阵分解技巧,设计了使系统渐近稳定并满足H_∞性能指标的状态反馈控制器。最后,通过数值仿真实例验证了设计方法的有效性。

基于多源数据融合的医用影像辅助诊断模型设计

为了实现对患者的准确分类,辅助医生进行疾病诊断识别,文中提出一种基于多源数据融合的影像数据辅助诊断模型。该模型将MRI和PET图像进行融合,并以改进的Transformer网络T2T-ViT为主干分类网络,通过迁移学习ImageNet数据集的参数,实现对阿尔茨海默病的分类。在公开数据集上进行的实验结果表明,所提出模型对于阿尔茨海默症患者的识别准确率可达0.95,优于目前的主流图像分类网络,证明其有效性,能够辅助影像医生进行疾病诊断,具有一定的应用价值。

Feature-Grounded Single-Stage Text-to-Image Generation

Recently,Generative Adversarial Networks(GANs)have become the mainstream text-to-image(T2I)framework.However,a standard normal distribution noise of inputs cannot provide sufficient information to synthesize an image that approaches the ground-truth image distribution.Moreover,the multistage generation strategy results in complex T2I applications.Therefore,this study proposes a novel feature-grounded single-stage T2I model,which considers the“real”distribution learned from training images as one input and introduces a worst-case-optimized similarity measure into the loss function to enhance the model's generation capacity.Experimental results on two benchmark datasets demonstrate the competitive performance of the proposed model in terms of the Frechet inception distance and inception score compared to those of some classical and state-of-the-art models,showing the improved similarities among the generated image,text,and ground truth.

The interaction mechanism of nickel ions with L929 cells based on integrative analysis of proteomics and metabolomics data

The aim of this article was to study the toxicity mechanism of nickel ions(Ni^(2+))on L929 cells by combining proteomics and metabolomics.First,iTRAQ-based proteomics and LC/MS metabolomics analyses were used to determine the protein and metabolite expression profiles in L929 cells after treatment with 100μMNi^(2+)for 12,24 and 48 h.A total of 177,2191 and 2109 proteins and 40,60 and 74 metabolites were found to be differentially expressed.Then,the metabolic pathways in which both differentially expressed proteins and metabolites were involved were identified,and three pathways with proteins and metabolites showing upstream and downstream relationships were affected at all three time points.Furthermore,the protein-metabolite-metabolic pathway network was constructed,and two important metabolic pathways involving 4 metabolites and 17 proteins were identified.Finally,the functions of the important screened metabolic pathways,metabolites and proteins were investigated and experimentally verified.Ni^(2+)mainly affected the expression of upstream proteins in the glutathione metabolic pathway and the arginine and proline metabolic pathway,which further regulated the synthesis of downstream metabolites,reduced the antioxidant capacity of cells,increased the level of superoxide anions and the ratio of GSSG to GSH,led to oxidative stress,affected energy metabolism and induced apoptosis.

基于RNN的中文二分结构句法分析

为了构建一个简单易扩展的中文句法分析器,我们依据朱德熙和陆俭明先生的中文二分结构的层次分析句法理论,手工构建了一个3万句的二分结构的中文句法树库,并使用哈夫曼编码方式来简化表示完全二叉树的层次结构。该文将中文句法分析转换为迭代二分的序列标注问题,并根据该任务的特点,提出了在词的间隔上进行标记的序列标注模型(RNN-Interval,RNN-INT),与常用的循环神经网络模型(RNN,LSTM)和条件随机场模型(CRF)进行对比实验,使用mx2交叉验证序贯t-检验来比较模型。实验结果表明,RNN-INT模型在窗口为1的词特征就可达到最好的性能,并好于其他窗口大小和其他序列标注模型(RNN,LSTM,CRF)。最后,在测试集上,在人工分词下,RNN-INT在短语级别的F1值(块F1)达到71.25%,在句子级别的准确率达到约43%。