Search

Partial discharge (PD) pattern recognition approaches are designed to identify the types or severities of the insulation defects within the high voltage equipment, which is vital for evaluating potential harmfulness and making follow-up maintenance plan. In recent years, many advanced machine learning (ML) algorithms have been introduced to this field and achieved remarkable results. As the second one of the two-part papers, we aim to give a comprehensive review regarding the pattern recognition approaches for ultra-high frequency (UHF) PD data in this paper. These methods are grouped into three categories, which are the traditional ML-based PD type recognition, the deep learning-based (DL) PD type recognition, and PD severity assessment. Specifically, for the first topic, feature extraction methods, dimensionality reduction methods and classification methods are reviewed separately. For the second topic, many state-of-the-art DL methods are discussed, including the deep belief network (DBN), deep autoencoder network (DAN), convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), graph convolutional network (GCN), deep ensemble learning (DEL), etc. For the third topic, the relevant algorithms are also divided into the conventional ML-based ones and the DL-based ones, which are studied in detail respectively. Finally, a brief discussion about the application effects of the above technologies is given, and some future directions are suggested. This paper covers almost every aspect of the PD pattern recognition and highlights the latest progress, which can provide valuable references for scholars in this field.

Existing goaves (e.g., shafts and roadways) in mines represent important hidden dangers during the production of underlying coal seams. In this view, the accurate identification, analysis, and delimitation of the scope of goaves have become important in the 3D seismic exploration of mines. In particular, an accurate identification of the boundary swing position of goaves for 3D seismic data volumes within a certain depth interval is key and difficult at the same time. Here, a wide-band and wide-azimuth observation system was used to obtain high-resolution 3D seismic data. The complex structure of a mine was analyzed, and a seismic double processing system was applied to verify the fine processing effect of a goaf and improve the resolution of the 3D seismic data. Based on the seismic attribute identification characteristics of the goaf structure, we decided to adopt multi-attribute comprehensive identification and data fusion technologies to accurately determine the position of the goaf and of its boundary. Combining this information with the mine roadway engineering layout, we verified the accurateness and correctness of the goaf boundary location. Our study provides a good example of the accurate identification of the 3D seismic data of a roadway goaf.

To study the influence of coal mining on the water richness overlying strata in the mining area using time-lapse electrical monitoring technology, four dataset acquisitions were completed with the same acquisition method, equipment, parameters, and processing flow. According to the characteristics of the data, major problems such as topographic correction, high-precision denoising, spatial and temporal normalization, and resistivity data inversion have been solved. Precise tomographic imaging was achieved through high-precision data processing and difference inversion. The results show that the electrical stratification characteristics of the overlying soil and rock layers are clear, the resistivity from the surface down gradually increases, and the electrical layers are not uniform locally. During mining, the overlying strata are affected by mining, the electrical resistivity of the underlying aquifers increased to varying degrees, and the fluctuation of electrical resistivity increased while the aquifer’s water content decreased. After mining, the overlying aquifer has the phenomenon of ‘reduced resistivity and water recovery’. After a period of time, the overlying soil disturbance and overlying rock failure zone will gradually tend to be stable. Meanwhile, the aquifer structure and water content will also gradually recover. Our results could provide guidance for water resources protection in this region.

With increasing data availability and interpretation method development, the inter-well tracer test plays more and more important role in oilfield production and development. This paper introduces the trace substance tracer testing in CQ oilfield, and interprets the testing data by using the residence time distribution method. Through the testing results, the connection relationship between oil production and water injection wells, the preferential direction of water flooding, the speed of water line propulsion, the spread of injected water and the presence of high permeability bands between wells in the testing well group are clarified. Meanwhile, the heterogeneity of the reservoir between wells is more clearly understood, which provides a basis for the optimization of subsequent development plans. Through verification with production data, the testing results show that the application of trace substance tracer in the study area is effective, indicating that trace substance tracer can also be well applied in low permeability reservoirs.

Abstract The closure fracture phenomenon increases the complexity of well testing and reduces the accuracy of productivity forecasts when tight oil reservoirs are exploited. However, most existing productivity models tend to ignore this. Therefore, a productivity prediction model for multi-fractured horizontal wells in tight oil reservoirs considering fracture closure has been developed by considering the stress sensitivity in the formation and combining the physical parameters of actual production. The model is solved by the Laplace transform, perturbation transform, Pedrosa transform, and Stehfest numerical inversion. Drawing productivity impact curves and discussing productivity influence factors based on the model results from this study show that the model is reasonable. In the actual production process, the hydraulic fracture parameter values are not as high as possible, and they have a reasonable range of values. The fracture closure pressure has a significant impact on the production of tight oil reservoirs. The higher the fracture closure pressure is, the greater the fracture conductivity decreases sharply, and the larger the proppant elastic modulus is, the stronger the fracture conductivity. The influence of fracture conductivity on the production in tight oil reservoirs has an obvious point, and when the value is less than the point, the production effect is good. Improved production can be achieved by balancing the relationship between fracture parameters. The findings of this study can help to better understand the influence of fracture parameters on productivity and contribute to increasing well production and improved development of tight oil reservoirs.

In order to reveal the basic enzymatic properties of the chitinase produced by an endophytic bacterium Bacillus thuringiensis Bt028 isolated from sour orange fruit, the chitinase in the strain fermentation broth were purified by centrifugation, ammonium sulfate precipitation, dextran gel G-100 chromatography and SDS-polyacrylamide gel electrophoresis and the optimum temperature, optimum pH and catalytic kinetic parameters of the chitinase were also investigated. The results showed that electrophoretic pure chitinase was obtained by centrifugation, ammonium sulfate precipitation and dextran G-100 gel chromatography from fermentation broth of the strain Bt028, with the specific activity of 681.78 U/mg, the purification factor of 3.21 and enzymatic activity recovery of 15.52%, and the molecular mass of the chitinase was determined to be 65 kDa by SDS-polyacrylamide gel electrophoresis. The enzymatic characteristics study results showed that the optimum reaction temperature of the chitinase was 60 ℃ with good stability when temperature lower than 60 ℃; and the optimum reaction pH was 6.5 with good stability at pH5.5~7.5. Mg2+, Ca2+, Hg2+ and Co2+ had inhibitory effect on the enzyme activity, while Cu2+ and Fe3+ had a certain promotion effect. Low concentrations of methanol, ethanol, n-propanol and dimethyl sulfoxide increased the enzymatic activity, however, the chitinase would be inhibited when the concentration of these organic solvents were increased to a certain level. The chitinase could be activated by acetone but inhibited by formaldehyde. Under the optimal catalytic conditions, the value of Km, Vmax and Kcat of the chitinase-catalyzed reaction were 29.533 mg/mL, 108.696 μmoL/(L·min) and 0.527/min, respectively. Research results provide technical parameters for the practical application of the chitinase.

Non-small-cell lung cancer (NSCLC) is the main cause of cancer-related deaths worldwide, with a low five-year survival rate, posing a serious threat to human health. In recent years, the delivery of antitumor drugs using a nanostructured lipid carrier (NLC) has become a subject of research. This study aimed to develop an apigenin (AP)-loaded nanostructured lipid carrier (AP-NLC) by melt sonication using glyceryl monostearate (GMS), glyceryl triacetate, and poloxamer 188. The optimal prescription of AP-NLC was screened by central composite design response surface methodology (CCD-RSM) based on a single-factor experiment using encapsulation efficiency (EE%) and drug loading (DL%) as response values and then evaluated for its antitumor effects on NCI-H1299 cells. A series of characterization analyses of AP-NLC prepared according to the optimal prescription were carried out using transmission electron microscopy (TEM), differential scanning calorimetry (DSC), X-ray diffraction (XRD), and Fourier transform infrared spectroscopy (FT-IR). Subsequent screening of the lyophilization protectants revealed that mannitol could better maintain the lyophilization effect. The in vitro hemolysis assay of this formulation indicated that it may be safe for intravenous injection. Moreover, AP-NLC presented a greater ability to inhibit the proliferation, migration, and invasion of NCI-H1299 cells compared to AP. Our results suggest that AP-NLC is a safe and effective nano-delivery vehicle that may have beneficial potential in the treatment of NSCLC.

Traffic flow forecasting is a basic function of intelligent transportation systems, and the accuracy of prediction is of great significance for traffic management and urban planning. The main difficulty of traffic flow predictions is that there is complex underlying spatiotemporal dependence in traffic flow; thus, the existing spatiotemporal graph neural network (STGNN) models need to model both temporal dependence and spatial dependence. Graph neural networks (GNNs) are adopted to capture the spatial dependence in traffic flow, which can model the symmetric or asymmetric spatial relations between nodes in the traffic network. The transmission process of traffic features in GNNs is guided by the node-to-node relationship (e.g., adjacency or spatial distance) between nodes, ignoring the spatial dependence caused by local topological constraints in the road network. To further consider the influence of local topology on the spatial dependence of road networks, in this paper, we introduce Ollivier–Ricci curvature information between connected edges in the road network, which is based on optimal transport theory and makes comprehensive use of the neighborhood-to-neighborhood relationship to guide the transmission process of traffic features between nodes in STGNNs. Experiments on real-world traffic datasets show that the models with Ollivier–Ricci curvature information outperforms those based on only node-to-node relationships between nodes by ten percent on average in the RMSE metric. This study indicates that by utilizing complex topological features in road networks, spatial dependence can be captured more sufficiently, further improving the predictive ability of traffic forecasting models.

Partial discharge (PD) detection and diagnosis based on the ultra-high frequency (UHF) signals is one of the most widely adopted methods to evaluate the internal insulation status of high voltage equipment. Benefit from the rapid development of computing hardware and data processing algorithms, the intelligent PD fault diagnosis method based on the UHF data has made considerable progress in the past two decades. This two-part paper aims to give a comprehensive review about the application of signal processing and machine learning technologies in UHF PD detection and diagnosis. These technologies are divided into three categories according to their respective purpose, which are the preprocessing technology, source localization technology and pattern recognition technology. As the first one of the two-part review, we focus on the preprocessing and localization approaches in this paper. Specifically, for the preprocessing topic, the methods for signal denoising, multi-source separation, and pulse segmentation are included. While for the localization topic, the time difference of arrival (TDOA) method, direction of arrival (DOA) method, received signal strength indicator (RSSI) method, and other latest methods are reviewed. For each topic, the basic ideas, recent research progresses, advantages and limitations are discussed in detail. Before the conclusion, we also make a discussion about the application effects of the above technologies and prospect some future directions accordingly. In the second paper, the pattern recognition problems based on the UHF PD data will be concentrated, especially the application of deep learning algorithms.

ObjectiveEarly screening contributes to the early detection of children with autism spectrum disorder (ASD). We conducted a longitudinal ASD screening study in a large community setting. The study was designed to investigate the diagnostic rate of ASD screening and determine the effectiveness of ASD screening model in a community-based sample.MethodsWe enrolled children who attended 18- and 24-month well-child care visits in Shanghai Xuhui District. Modified Checklist for Autism in Toddlers, Revised with Follow-up (M-CHAT-R/F) and Binomial Observation Test (BOT) were selected as screening instruments. Screen-positive children were referred to a tertiary diagnostic center for comprehensive ASD diagnostic evaluation. Screen-negative children received well-child checkups and follow-up every 3–6 months until age three and were referred if they were suspected of having ASD.ResultsA total of 11,190 toddlers were screened, and 36 screen-positive toddlers were diagnosed with ASD. The mean age at diagnosis for these children was 23.1 ± 4.55 months, diagnosed 20 months earlier than ASD children not screened. The diagnostic rate of ASD was 0.32% (95% CI: 0.23–0.45%) in this community-based sample. In addition, 12 screen-negative children were diagnosed with ASD during subsequent well-child visit and follow-up. The average diagnostic rate of ASD rose to 0.43% (95% CI: 0.32–0.57%) when toddlers were followed up to 3 years old. The positive predictive values (PPVs) of M-CHAT-R/F, M-CHAT-R high risk, and BOT for ASD were 0.31, 0.43, and 0.38 respectively.ConclusionOur findings provide reliable data for estimating the rate of ASD detection and identifying the validity of community-based screening model. M-CHAT-R/F combined with BOT can be an effective tool for early detection of ASD. This community-based screening model is worth replicating.

Remote sensing image (RSI) scene classification is the foundation and important technology of ground object detection, land use management and geographic analysis. During recent years, convolutional neural networks (CNNs) have achieved significant success and are widely applied in RSI scene classification. However, crafted images that serve as adversarial examples can potentially fool CNNs with high confidence and are hard for human eyes to interpret. For the increasing security and robust requirements of RSI scene classification, the adversarial example problem poses a serious problem for the classification results derived from systems using CNN models, which has not been fully recognized by previous research. In this study, to explore the properties of adversarial examples of RSI scene classification, we create different scenarios by testing two major attack algorithms (i.e., the fast gradient sign method (FGSM) and basic iterative method (BIM)) trained on different RSI benchmark datasets to fool CNNs (i.e., InceptionV1, ResNet and a simple CNN). In the experiment, our results show that CNNs of RSI scene classification are also vulnerable to adversarial examples, and some of them have a fooling rate of over 80%. These adversarial examples are affected by the architecture of CNNs and the type of RSI dataset. InceptionV1 has a fooling rate of less than 5%, which is lower than the others. Adversarial examples generated on the UCM dataset are easier than other datasets. Importantly, we also find that the classes of adversarial examples have an attack selectivity property. Misclassifications of adversarial examples of RSIs are related to the similarity of the original classes in the CNN feature space. Attack selectivity reveals potential classes of adversarial examples and provides insights into the design of defensive algorithms in future research.

Numerical simulation is very important to solve geotechnical problems. However, it is difficult to obtain required comprehensive and accurate information such as parameters, boundary conditions, and etc. In this paper, a grey distributed parameter model, which integrates the finite element method (FEM) with the grey system theory, was proposed to address the issue. The analysis of grey properties on rock and soil system was performed. The equilibrium equations, geometric equations, physics equations and related differential equations were obtained, each of the equations contains grey parameters and variables. And the discretization and solution methods of the FEM with the grey variables were discussed. An example of deep-buried circular mining tunnel was applied to test the proposed model. The calculation results were compared with those of the exact solution (analytical solution) and the classical FEM, respectively, through which the rationality of the proposed model was demonstrated. For the first time, grey variables and grey parameters are defined in geotechnical numerical simulation. The expressions of basic equations with grey variables are given respectively. A grey distributed parameter model which integrates the FEM with the grey system theory is proposed to solve geotechnical problems, and the optimal solution to the proposed model is determined through calculation and comparison of an application example. The proposed numerical model with grey variables not only has the advantage of grey system theory, but also greatly improves the adaptability and application effect of the model, which contributes to the prediction and evaluation problems in geological engineering, geotechnical engineering, water conservancy engineering and civil engineering with complex structures.

Ecological red line policy is an environmental policy that the national government aims to protect important ecosystems by dividing them into different functional red lines. Ecological red lines is the guarantee lines of ecological services and the bottom lines of ecological security. Quantifying the value of ecological red lines is the inherent requirement of the development of ecological red lines policy. This study investigated whether residents in Nanjing were willing to pay for ecological red lines protection by using the theoretical framework of the double-bounded dichotomous choice method. The survey collected a sample of 1,058 respondents, of whom 835 were willing to pay to protect the ecological red lines, and most of those unwilling to pay believed that protection of ecological red lines should be paid by governments or polluters. The calculated mean Willingness to Pay was 300.07 CNY·year−1, among which the respondents showed the highest willingness to pay for nature reserve areas and water resources reserve areas and the lowest willingness to pay for cultural landscape areas. The final estimated non-use value of ecological red lines in Nanjing was 2550 million CNY. By analyzing the results of the logistic regression model, education level and family annual income have a significant positive effect on Willingness to Pay, while age, the influence of ecological red lines on themselves, and the initial bid value are negatively significant. The evaluation results of this study have certain value mainly in the framework of sustainable development of ecological red lines, to improve residents’ awareness and participation in the protection of ecological red lines, and to provide scientific support for the management plan and policies to improve the environmental quality of ecological red lines.

The quantitative detection of faults using the channel wave seismic method has been a major but challenging area of interest. In this study, we adopted an effective technical process to evaluate fault attribution. First, we use integrated transmission and reflection channel wave information to improve the accuracy of extraction velocity. Then, the location of the fault is determined by the elliptical tangent offset method, and feature extraction and fault location extension determination are achieved through logistic regression and a neural network. This is combined with the prior geological information, the fractional dimension D to the quantitative analysis of the fault throw. Data regarding the 4203 working face of a mine in Shanxi, China, are considered as an example. Two groups of faults were predicted, with the location error in the f30 fault position as 6.7 m. In addition, the f29 fault throw first increased, and then gradually decreased from the return airway to the haulage gateway. These predicted results have been drill-verified and were used to modify the original design. The proposed method has good stability and promising application prospects for fault evaluation.

Purpose: In order to compensate for the early intrauterine growth restriction, small-for-gestational age (SGA) infants have “catch-up growth” after birth. Increased caloric intake has been suggested for SGA infants conventionally. It is important to determine if the early growth rate of body mass index (BMI) is associated with risk of persistent obesity later in life. In this longitudinal cohort study, we assessed the BMI of a large cohort of children who were SGA at birth to determine their risk of persistent obesity at school age (6–7 years) due to excessive weight gain in the first 3 years of life.Methods: We collected the height and weight data of 23,871 SGA babies. A polynomial function was used to fit the BMI-for-age z-score (BAZ) values of 0–6 years old SGA children and interpolate their growth trajectory. In addition, we screened out 6,959 children from 23,871 children to further evaluate the dynamic changes of early childhood BMI. We divided the school-age children into groups as non-obese (BAZ < 2) and obese (BAZ > 2), and determined the association between changes in BMI and school-age obesity.Results: From the perspective of BMI distribution, the interpolated growth trajectory indicated that SGA children reaching overweight status or developing obesity by 3 years of age, continued to have obesity until school age (R2, 0.65; R2, 0.21). The retrospective analysis showed that children who were overweight and had obesity during school age had a high BMI from early age. By analyzing the changes in early BMI, we found that the fastest growth of SGA children occurred in the early infancy before 6 months and they continued to grow rapidly for a period of time. Interestingly, former SGA children who maintained a near overweight (1 < BAZ < 2) status before the age of 2 maintained an appropriate growth rate and usually did not develop obesity.Conclusions: A rapid increase in BMI during early infancy in former SGA newborns leads to a persistent risk of obesity. The energy intake of SGA infants should appropriately meet the infants' growth needs and early BMI changes should be closely monitored for an optimal integrated management.

The computed tomography (CT) reconstruction algorithm is one of the crucial components of the CT system. To date, total variation (TV) has been widely used in CT reconstruction algorithms. Although TV utilizes the a priori information of the longitudinal and lateral gradient sparsity of an image, it introduces some staircase artifacts. To overcome the current limitations of TV and improve imaging quality, we propose a multidirectional anisotropic total variation (MATV) that uses multidirectional gradient information. The surrounding rock of coal mining faces uses principles of tomography similar to those of medical X-rays. The velocity distribution for the surrounding rock can be obtained by the first-arrival traveltime tomography of the transmitted waves in the coal mining face. Combined with the geological data, we can interpret the geological hazards in the coal mining face. To perform traveltime tomography, we first established the objective function of the first-arrival traveltime tomography of the transmitted waves based on the MATV regularization and then used the split Bregman method to solve the objective function. The simulated data and real data show that the MATV regularization method proposed in this paper can better maintain the boundaries of geological anomalies and reduce the artifacts compared with the isotropic total variation regularization method and the anisotropic total variation regularization method. Furthermore, this approach describes the distribution of geological anomalies more accurately and effectively and improves imaging accuracy.

In China, south of the Yangtze River, there are a large number of carbon sources, while the Sichuan Basin is the largest sedimentary basin; it makes sense to select the targets for CO2 geological storage (CGUS) early demonstration. For CO2 enhanced oil and gas, coal bed methane recovery (CO2-EOR, EGR, and ECBM), or storage in these depleted fields, the existing oil, gas fields, or coal seams could be the target areas in the mesoscale. This paper proposed a methodology of GIS superimposed multisource information assessment of geological suitability for CO2 enhanced water recovery (CO2-EWR) or only storage in deep saline aquifers. The potential per unit area of deep saline aquifers CO2 storage in Central Sichuan is generally greater than 50 × 104 t/km2 at P50 probability level, with Xujiahe group being the main reservoir. CO2 storage potential of depleted gas fields is 53.73 × 108 t, while it is 33.85 × 108 t by using CO2-EGR technology. This paper recommended that early implementation of CGUS could be carried out in the deep saline aquifers and depleted gas fields in the Sichuan Basin, especially that of the latter because of excellent traps, rich geological data, and well-run infrastructures.

With the application of artificial intelligence in modern industry, the demand for intelligent fault diagnosis systems is increasing, which has attracted much attention of researchers. There is a popular methodology to use the knowledge or rules defined by experts for fault diagnosis. However, the expert defined information that needs to be manually annotated for faults by human with expertise knowledge costs too expensive in real industrial scenarios and is not always practicable to unknown fault category. In order to realize the zero-shot intelligent fault diagnosis and extend to a more general case, we propose an intelligent fault diagnosis system in this paper, using the statistical attributes automatically extracted from the vibration signal-self. To obtain the statistical attributes, the signals are decomposed by Ensemble Empirical Mode Decomposition (EEMD) to get their detailed time and frequency characteristics. And the statistical description of the time and frequency domains is taken as the attributes to relate the unseen categories with the seen ones. With the attributes, Gaussian models of the seen categories are established. And those of the unseen category are estimated by the attribute strength relation between the seen and unseen categories. The unseen category is finally predicted by Maximum A Posterior probability (MAP). The experiments on fault dataset show that the proposed system can obtain well recognition performance compared with the classic methods. Among them, the accuracy of zero-shot fault diagnosis can approach to 86%. And the performance can be further improved through attribute selection.