Focusing on the problems of complex structure and low feature extraction efficiency that exist in some traditional neural network algorithms, an improved convolutional (CNN), combined with a attention mechanism, is proposed to improve reduce complexity. First, block module (CBAM) was introduced, which can strengthen useful features from data, thereby improving efficiency. Second, residual optimized, original multilayer convolution simplified single-layer convolution, reduces complexity model...

Deformation is a comprehensive reflection of the structural state concrete dam, and research on prediction models for dam deformation provides basis safety monitoring early warning strategies. This paper focuses practical problems such as multicollinearity among factors; subjectivity factor selection; robustness, externality, generalization, integrity deficiencies; unsoundness evaluation systems models. Based rough set (RS) theory long short-term memory (LSTM) network, single-point...

Extracting implicit anomaly information through deformation monitoring data mining is highly significant to determining dam safety status. As an intelligent singular value diagnostic method for concrete monitoring, shallow neural network models result in local optima and overfitting, require manual feature extraction. To obtain diagnosis model that can be used a convolutional (CNN) has advantages of deep learning (DL), such as automatic extraction, good fitting, strong generalizability, was...

Aiming at the problem that rolling bearing vibration signals are susceptible to noise interference in working environment, an anti-noise network (MCAMDN) depending on a selectable branching multiscale convolution and attention mechanism is proposed. Compared single-branch convolution, multibranch module (MConv) used MCAMDN can effectively choose use branches different scales due change of feature map size, which ensures learn more features, appropriately reduce parameters model amount...

Abstract Addressing the challenge of inconsistent data feature distribution and difficulty fault diagnosis in rolling bearings operating under variable conditions, a novel approach is proposed for diagnosis. Dynamic convolution dual-channel fusion are utilized our method. In shallow network layer, we employ convolutional structure, combining large group with small to enhance extraction high-low frequency information from images. The improved GhostNetV2 bottleneck layer was used deeper obtain...

Abstract Time-frequency analysis (TFA) is a widely used reservoir prediction technology. During the identification of sand bodies using TFA, abnormally high values in high-frequency portion spectrum usually indicate presence thin bodies. However, owing to complexity depositional environment, spatial distribution often variable, and multiple phases bodies—superimposed on each other—can occur with uneven thicknesses. A frequency information extraction technique based time-frequency ridges...

In this paper we propose an approach to approximate a truncated singular value decomposition of large structured matrix. By first decomposing the matrix into sum Kronecker products, our can be used number values and vectors more efficiently than other well-known schemes, such as iterative algorithms based on Golub--Kahan--Lanczos bidiagonalization. We provide theoretical results numerical experiments demonstrate accuracy approximation show how solve scale ill-posed inverse problems, either...

Potassium Titanium Oxide Phosphate (KTP) is a new nonlinear frequency-conversion crystal. It has chemical stability, high coefficient, damage threshold, easily-polished surface, and broad transparency range. be used in solid green laser with medium low power widely. The requirement for surface roughness less than 1nm.In this paper, the removal rate are discussed different velocity, pressure size of abrasive powder. In order to satisfy requirement, polishing techniques ultra-precision plane...

An UWB Vivaldi antenna with broad beamwidth is investigated. The analysis explains the design of parameters that decide characteristic and puts forward idea combining two congruent two-element arrays, which requires integration both 3dB power divider directional coupler in whole structure. has been designed optimized to cover S/C band (2-6GHz) accomplish right-hand circular polarization left-hand synchronously.

Model update is an important module in target trackers. It plays role adaptive tracking. Many researches have proven that different model strategies should be adopted, when tracking scenes, especially occlusion and deformation. Though many been proposed recent years, few of them make high improvement good combination on In this paper, we first proved there a close relationship between the scenes response maps. Then, strategy for calculating rate based map. experiments done to compare with...

Q factor is a significant parameter to measure seismic absorption attenuation. Accurate and robust estimation of an important premise enhance resolution, image interpret subsurface, predict characterize reservoirs. Currently, the centroid frequency shifting amplitude spectrum (CFS) method widely used approach. However, assumption Gaussian, boxcar, or triangular shapes for CFS will cause systematic error estimation. Moreover at high low frequencies are easily susceptible random noise, which...

In this paper, we propose a new scheme for modelling the diverse behavior of neurons. We introduce conditional activation, in which neurons activation function is dynamically modified by control signal. apply method to recreate special existing human auditory and visual system. A heterogeneous multilayered perceptron (MLP) incorporating developed models demonstrates simultaneous improvement learning speed performance across various number hidden units layers, compared homogeneous network...

Intelligent recognition camera for electric transmission line status plays an important role in power our country, and the ice environment will bring severe challenges to its normal use. Based on analysis of harm caused by icing intelligent camera, this paper proposes that it should be very necessary carry out test laboratory reform weak points products. According proposed verification conditions, device is designed developed, tested laboratory. Finally, proved meets requirements conditions...