Search

Based on the generalized Tellegen theorem, this paper establishes a mathematical model for the sensitivity calculation of power networks and derives the sensitivity formula of power networks. Next, the influence of node reactive power on network loss is studied, combined with the sensitivity of node voltage to network parameters in the generalized Tellegen theorem, a reactive power optimization model is established, a reactive power step function is constructed, and the optimal step is obtained by loop iteration Long, optimize the reactive power of the node to minimize the network loss. The IEEE-11 node and 14node systems are analyzed using the reactive power optimization algorithm proposed in this paper, and the results of network loss before and after optimization are compared. The method proposed in this paper is further verified to be correct.

Tibet is located in southwestern China, with many mountains and complex geological conditions, which are prone to natural geological disasters such as landslides and debris flows. How to extract landslide information efficiently and accurately in real time is of great significance to the rapid response after disasters. The application of deep learning semantic segmentation algorithms to information extraction of geological disasters is a new exploration in recent years. An improved PSPNet network model was proposed for semantic segmentation of landslide in Nyingchi, Tibet Province. Considering the segmentation performance and computing speed, the main feature extraction network in front of PSPNet adopts MobileNetV2 structure, which reduced the number of network parameters, reduced the computational complexity of the network, and effectively improved the convergence speed of the network. Based on the landslide images collected in Nyingchi prefecture, Tibet province, the improved model was compared with the classical SegNet, Unet and the original PSPNet algorithm. The improved algorithm effectively reduces the misclassification and divides objects more accurately.

Through the analysis of the silicon solar cell mathematical model, this paper presents a method to determine the model parameters by using the open circuit voltage and short circuit current of silicon solar cell and the voltage and current at the maximum output power point, a method to modify the model parameters of silicon solar cells under different light intensity is proposed; The silicon solar cell was simulated and experimentally measured, the data analysis results show that the model determined by this method can reflect the characteristics of silicon solar cells at low light intensity.

The paper presented an optimization design of filament-wound composites pressure hull subjected to hydrostatic pressure for underwater vehicle application. An Optimization Platform was set up by interworking a genetic algorithm and numerical simulation. A sensitivity analysis was performed to study the effects of design variables on the buckling pressure, material failure pressure and design pressure. Results revealed that the buckling pressure or the material failure pressure would restrict the design pressure of different thickness of pressure shell. Application of composite materials for deep-water pressure shell had more reserve buoyancy, which would result in miniaturizing the size of underwater vehicles. Effects of different type of ring-stiffened configuration were analyzed. Mid-stiffened configuration can enhance stability significantly. End-stiffened configuration can minimize the effect of stiffness mismatch. Results of this study provided a valuable reference for designers of underwater vehicles. The paper suggest that new way of enhancement, such as variable thickness, composites ribs would be used to solve the conflict of buckling pressure and material failure pressure.

This paper is devoted to the buckling behavior of the composite cylindrical shell subjected to hydrostatic pressure. Both critical buckling pressure and buckling shape are taken into account in the paper. The critical buckling pressure has been calculated using analytic formulas and validated by finite element method. The influence of type of materials, ply orientation angle, and overall thickness of shells on critical buckling pressure and buckling shape has been investigated numerically and demonstrated by several examples. Three type of common composite materials, carbon/epoxy, boron/epoxy and glass/epoxy, are chosen in the paper to make a comparison.

Results of a numerical study into buckling behavior of cylindrical laminated shell with, and without mid-strengthening subjected to uniform hydrostatic pressure are discussed. Compared with the conventional ring-stiffened cylinder shell, it can greatly improve pressure resistant capabilities of the shell with only small increase the weight. The finite element method based on ABAQUS used to simulate and analyze the buckling behavior of the pressure shell. The model was about 300mm inner diameter, 700mm long and had uniform thickness of 4mm. The buckling pressures were in the range from 0.75 to 1.82Mpa under different parametric of the reinforced structure. The main work and conclusions are mentioned below: (1) Mid-strengthening structure can improve the critical buckling load of cylindrical shell significantly, buckling mode and maximum amount of deformation will both change; (2) The critical buckling load of cylindrical shell with mid-strengthening gradually increases at first and then decreases with increase in ply angles of reinforced layers; (3) When the width of the reinforced stiffener is inversely proportional to its thickness, the critical buckling load also increases at first and then decreases with the increase in widths of reinforced layers. The results of this study can provide a reference for designing similar composite cylindrical shell under external pressure.

The goal of semi-supervised learning is to improve the performance of supervised learning tasks using unlabeled data. Deep learning has a high demand for making use of large-scale unlabeled data. We propose a simple and novel method of utilizing unlabeled data for semi-supervised learning to improve the performance of the deep learning model. First, we train a Gaussian process classifier (GPC), and use its output on unlabeled data as a pseudo-label. Then, we pre-train the deep learning model with the pseudo-labeled data to initialize the parameters of the model. Finally, we fine-tune it with the labeled data. We apply this method to five classes of video from the UCF101 data set which contains a small amount of labeled data. The experiment results show significant performance improvements compared to those of the deep learning model without pre-training. We further study the pre-training effect using pseudo-labeled data with a different probability, and give some advice for practical application. Moreover, we propose to train GPC with stratified sampled data (SGPC) to reduce the computation time of GPC when there are large amounts of data. Finally, we generalize proposed pseudo-labeling method to any classifier having good performance and give some advice for pseudo-labeling method selection.

Image analogy is the process of creating an image filter that precisely reflects the characteristics contained in the training data. Recently, the image analogy problem was generally handled by deep neural network (DNN) with the development of a deep learning technology. Generally, DNN suffers from a fatal problem in that it requires large amounts of data for training. However, as pairs of images with the same relationship are needed for an image analogy, it is hard to collect sufficient data for image analogy using DNN. In order to solve this problem, we propose an image analogy method using a Gaussian process. In this method, a Gaussian process regression is used instead of DNN regression to adjust the feature vectors which will be used in creating filtered image. Additionally, in order to accelerate the training speed of Gaussian process, we also propose novel sampling methods that select salient instances from a given dataset. Our experiment result demonstrates that the proposed image analogy method using a Gaussian process with salient instance sampling performs significantly better than DNN in environments with small dataset size.

NanoSQUIDs with Dayem nanobridge weak-link junctions were fabricated on single layer niobium film by focus ion beam(FIB) milling. The dimension of the nanobridge is less than 60 nm and the loop scale is from 400×400 nm to 2.5×20 μm. The working temperature range for this device was researched. The voltage-flux transfer properties were measured at different working temperatures.

In this paper we report the progress on single quantum Hall devices for the resistance standards in NIM. We use the NIM-made device and the LEP-made device distributed to NIM by BIPM to calibrate a Tinsley 100 Ω transfer resistor using a cryogenic current comparator (CCC) at the filling factor k=2. The deviations of resistance values from the nominal value were −4.4808×10−6±3.5×10−9 by NIM-1 and −4.4987×10−6±3.5×10−9 by BIPM-2 separately. We assume the relative difference 1.8×10−8 is due to the drift of the 100 Ω transfer resistor. More stable resistor will be used in future experiments.

For the architectural design projects have the characteristics as comprehensiveness, relevancy, and etc., it is of great significance to make a classification study on all the related stakeholders. Combining with the research on documents and questionnaire survey, also making using of the SPSS statistical software, there have been ten kinds of key stakeholders to be identified, based on the three dimensions-interests, power and importance. According to the idea about the similarity between the relevant behavior among project stakeholders and the operation tendency of atoms, the ten kinds of key stakeholders could be divided into the core layer, the electronic layer and the outer core layer, through establishing the three dimensional atom atlas in architectural design projects. Here, we could get the results through analysis: there exists a link of harmony and loyalty between the stakeholders in different levels, which reflects specifically the relationships as coordination, centripetal force, leap and centralized group. And each of them is related and depended on the other. The atom structure would finally achieve the independence and adaptability of projects through the impetus of the stakeholders in the core layer and the expansion of four kinds of network structure relations. And the achievements in the study could provide the design enterprises with the new theoretical support to formulate the right business strategy.