Introduction The surface images of steel rails are extremely difficult to detect and recognize due the presence interference such as light changes texture background clutter during acquisition process. Methods To improve accuracy railway defects detection, a deep learning algorithm is proposed rail defects. Aiming at problems inconspicuous edges, small size interference, region extraction, improved Retinex image enhancement, modeling difference, threshold segmentation performed sequentially...

In this article, a novel feature-extraction-based process monitoring method is proposed for multimode processes with common features. Different from the traditional feature extraction methods that consider either scores or weightings between different modes, common-subspace-based takes both and into account developed based on tensor decomposition. addition, specific features each mode are extracted by independent component analysis. Moreover, moving-window Kullback-Leibler-divergence-based...

Customized production, quality variation of raw materials and other factors make industrial processes work in multiple operating modes. In general, complex have strong nonlinearity under each mode. this paper, a Gaussian mixture variational autoencoder (GMVAE) model, which combines with VAE, is proposed to monitor nonlinear Due the distribution limitation latent variable space, GMVAE can not only automatically extract features system, but also these follow distribution. Based on space...

Since the actual chromatogram data of water-flooded layer have characteristics multiple dimension, complexity and noise, it is difficult to accurately identify appraise in oil gas reservoirs. Therefore, this article proposes a recognition modeling approach based on intelligent ensemble classifier, integrated model-free Bayesian AdaBoost algorithm support vector machine algorithm. The effective characteristic information can be obtained using curve fitting method. In order transform sparse...

Single-phase-to-ground(SPG) fault accounts for about 80% of faults in distribution system, which will easily cause system overvoltage and multi-point-grounded short-circuit fault. As a result, location should be accurately quickly determined after SPG occurs. Due to different neutral-point grounding modes the characteristics are various when Therefore, detection methods under acquired data configured equipment may also different. According types monitoring devices applied, diagnosis...

Although the cellular endocytosis or uptake research using confocal laser scanning microscopy (CLSM) and Transwell inserts was frequently reported in experimental cell pharmaceutical research, there is little report on transport process based same techniques. One of main reasons that fluorescence most fluorescent reagents could be clearly definitely seen under CLSM only when they are internalized into concentrated within cells. Here, Madin-Darby Canine Kidney (MDCK) cells Coumarin 6 labeled...

Began with a brief introduction of the basic principle and structure SCADA system, this paper analyses 3 practical problems exist in traditional state estimation, which is based on an isolated time section. It also points out that cannot be eliminated by directly revising existing algorithm hence new method estimation should found from methodology. On basis, approach using process measurement data proposed paper. The fundamental idea includes two steps; first one to construct relatively...

At present, manufacturing in China faces both opportunity and challenge, is a stage of rapid development industrialization, relying heavily on manufacturing, facing international pressure to control carbon-emission. Under this background, based the operation process radical innovation, paper builds up innovation mechanism model low-carbon technology which can be divided into mechanism, safeguard feedback anti-driving further explaining relationship between these mechanisms their respective...

In research field of ocean wave simulation, especially in spectrum and the direction spectrum, there are irreal problems narrow steep peak wide flat valley. this paper, a synthesis model is proposed which utilizes Gerstner waveform combined with LHM function to achieve more realistic simulation sea surface. addition, an improved method based on particle modeling proposed. method, collision detection technology used process between waves ship, algorithm control number excitable particles....

Abstract With Entity recognition is an important basic tool for many natural language processing tasks such as information extraction, question answering systems, and syntactic analysis. divided into general field specific field. There were different methods fields. At present, one of the more common entity technologies a rule-based method, other statistical method. Due to data sets scarcity in aerospace engine field, there lack corresponding research This paper proposes statistical-based...

Abstract. Natural source Super Low Frequency(SLF) electromagnetic prospecting methods have become an increasingly promising way in the resource detection. The capacity estimation of reservoirs is great importance to evaluate their exploitation potency. In this paper, we built a signal-estimate model for SLF signal and processed monitored data with adaptive filter. non-normal distribution test showed that was obviously different from Gaussian probability distribution, Class B instantaneous...

In this paper, a method based on neighborhood information is proposed, which adjusts the parameters by using of chi square with Gauss weighted distance and selfadaptability shared neighbor weighting. And integrating nearest adaptive method, each pixel automatically given scale parameter to reduce need adjust parameters. local introduced in construct similarity matrix directly original image. At same time, existing graph-cut image segmentation necessary better network graph maximum flow...

In process industries like the hot tandem rolling mill (HTRM) process, faults occurred in different places will cause impacts on final product qualities. Thus, a comprehensive maintenance should be considered to guarantee stable and reliable production process. Different from traditional fault detection (FD) methods that can only determine time by means of successful time, this paper proposes multi -step FD method also reveals hazard fault, so levels given engineers. Firstly, HTRM is divided...

Nowadays, many industrial processes generate large amounts of multimode data, which generally have a natural tensor structure, causing some faults invisible with traditional process monitoring (PM) and fault diagnosis (FD) methods. Tensor decomposition (TD) is more practical approach for its effectiveness in solving high dimensionality problems as well indicating the links between different modes. This paper proposes common individual feature extraction method based on TD, identifies...