Search

Predicting stock prices has long been the holy grail for providing guidance to investors. Extracting effective information from Limit Order Books (LOBs) is a key point in high-frequency trading based on stock-movement forecasting. LOBs offer many details, but at the same time, they are very noisy. This paper proposes a differential transformer neural network model, dubbed DTNN, to predict stock movement according to LOB data. The model utilizes a temporal attention-augmented bilinear layer (TABL) and a temporal convolutional network (TCN) to denoise the data. In addition, a prediction transformer module captures the dependency between time series. A differential layer is proposed and incorporated into the model to extract information from the messy and chaotic high-frequency LOB time series. This layer can identify the fine distinction between adjacent slices in the series. We evaluate the proposed model on several datasets. On the open LOB benchmark FI-2010, our model outperforms other comparative state-of-the-art methods in accuracy and F1 score. In the experiments using actual stock data, our model also shows great stock-movement forecasting capability and generalization performance. [ABSTRACT FROM AUTHOR]

The aero-engine is the heart of an aircraft; its performance deteriorates rapidly due to the high temperature and high-pressure environment during flights. It is necessary to predict the remaining useful life (RUL) to improve the reliability of aero-engines and provide security for reliable flights. In previous flights, the sensors collected a lot of performance parameter data and formed a database regarding the aero-engine degradation process. These performance parameters cannot reflect the degradation process directly. In this paper, fuzzy clustering is applied to divide the degradation stages of the aero-engine, construct the health indicator, and describe the degradation process. Time-series decomposition modeling is applied to predict the degradation process of the health indicator. Based on the idea of similarity comparison, the RUL is predicted by comparing the similarity of time series through example learning. The method is verified and analyzed on the dataset published by National Aeronautics and Space Administration (NASA), and the mean square error (MSE) is 528. The result is better than the comparative method. [ABSTRACT FROM AUTHOR]

Refracturing treatment is an economical way to improve the productivity of poorly or damaged fractured horizontal wells in tight reservoirs. Fracture reorientation and fracture face damage may occur during refracturing treatment. At present, there is still no report on the rate decline solution for refractured horizontal wells in tight reservoirs. In this work, by taking a semi-analytical method, traditional rate decline and Blasingame-type rate decline solutions were derived for a refractured horizontal well intercepted by multiple reorientation fractures with fracture face damage in an anisotropic tight reservoir. The accuracy and reliability of the traditional rate decline solution were verified and validated by comparing it with a classic case in the literature and a numerical simulation case. The effects of fracture reorientation and fracture face damage on the rate decline were investigated in depth. These investigations demonstrate that fracture face damage is not conducive to increasing well productivity during the early flow period and there is an optimal matching relationship between the principal fracture section angle and permeability anisotropy, particularly for the reservoirs with strong permeability anisotropy. The fracture length ratio and fracture spacing have a weak effect on the production rate and cumulative production while the fracture number shows a strong influence on the rate decline. Furthermore, multifactor sensitivity analysis indicates that fracture conductivity has a more sensitive effect on well productivity than fracture face damage, implying the importance of improving fracture conductivity. Finally, a series of Blasingame-type rate decline curves were presented, and type curve fitting and parameter estimations for a field case were conducted too. This work deepens our understanding of the production performance of refractured horizontal wells, which helps to identify reorientation fracture properties and evaluate post-fracturing performance. [ABSTRACT FROM AUTHOR]

A fractured horizontal well is an effective technology to obtain hydrocarbons from tight reservoirs. In this study, a new semi-analytical model for a horizontal well intercepted by multiple finite-conductivity reorientation fractures was developed in an anisotropic rectangular tight reservoir. Firstly, to establish the flow equation of the reorientation fracture, all reorientation fractures were discretized by combining the nodal analysis technique and the fracture-wing method. Secondly, through coupling the reservoir solution and reorientation fracture solution, a semi-analytical solution for multiple reorientation fractures along a horizontal well was derived in the Laplace domain, and its accuracy was also verified. Thirdly, typical flow regimes were identified on the transient-pressure curves. Finally, dimensionless pressure and pressure derivative curves were obtained to analyze the effect of key parameters on the flow behavior, including fracture angle, permeability anisotropy, fracture conductivity, fracture spacing, fracture number, and fracture configuration. Results show that, for an anisotropic rectangular tight reservoir, horizontal wells should be deployed parallel to the direction of principal permeability and fracture reorientation should be controlled to extend along the direction of minimum permeability. Meanwhile, the optimal fracture number should be considered for economic production and the fracture spacing should be optimized to reduce the flow interferences between reorientation fractures. [ABSTRACT FROM AUTHOR]

Many oil wells in closed reservoirs continue to produce in the pseudo-steady-state flow regime for a long time. The principal objective of this work is to investigate the characteristics of two key pseudo-steady-state parameters—pseudo-steady-state constant (bDpss) and pseudo-skin factor (S)—for a well penetrated by a fracture with an azimuth angle (θ) in an anisotropic reservoir. Firstly, a general analytical pressure solution for a finite-conductivity fracture with or without an azimuth angle in an anisotropic rectangular reservoir was developed by using the point-source function and spatial integral method, and two typical cases were employed to verify this solution. Secondly, with the asymptotic analysis method, the expressions of pseudo-steady-state constant and pseudo-skin factor were obtained on the basis of their definitions, and the effects of permeability anisotropy, fracture azimuth angle, fracture conductivity and reservoir shape on them were discussed in detail. Results show that all the bDpss-θ and S-θ curves are symmetric around the vertical line, θ = 90° and form a hump or groove shape. The optimized fracture direction in an anisotropic reservoir is perpendicular to the principal permeability axis. Furthermore, a new formula to calculate pseudo-skin factor was successfully proposed based on these two parameters' relationship. Finally, as an application of pseudo-steady-state constant, a set of Blasingame format rate decline curves for the proposed model were established. [ABSTRACT FROM AUTHOR]

Fracture aperture of a fractured reservoir can be affected by both matrix elasticity and fracture compressibility when the reservoir pressure decreases, namely stress sensitivity. An elasticity parameter coupling Young's modulus and Poisson's ratio was introduced to reflect this geomechanical behavior, and a new model incorporating geomechanics was developed to analyze the flow behavior of a horizontal well in a naturally fractured reservoir. Pressure solutions for two cases—uniform-flux and infinite-conductivity—were derived, respectively. For the uniform-flux case, the effect of dimensionless elasticity parameter on the pressure-drop profile becomes stronger with continuing production, and the profile may be like a bow. Nine flow regimes can be observed on the transient response of the infinite-conductivity case. Stress sensitivity mainly affects the late-flow period and a larger dimensionless elasticity parameter causes a greater pressure drop. Due to stress sensitivity, the pressure derivative curve exhibits an upward tendency in the pseudo-radial flow regime, and the slope is greater than "1" in the pseudo-steady flow regime. For KT-I formation in the North Truva field, its elasticity parameter decreases with the increase of Young's modulus or Poisson's ratio and ranges from 8 × 10−8 Pa−1 to 1.1 × 10−7 Pa−1. Meanwhile, the transient response of H519 has a slight negative correlation with Young's modulus and Poisson's ratio in the pseudo-steady flow regime. [ABSTRACT FROM AUTHOR]