Search

Although the deep learning-based indoor image localization has made significant improvement in terms of accuracy, efficiency, and storage requirement of large indoor scenes, the need for collecting huge labeled training data severely limits its practical application. Recently, the synthetic images rendered from widely available 3D models have shown promising potential to relieve the data collection problem. However, due to the dramatic differences between the synthetic and real images, the localization accuracy of approaches trained on synthetic images is not comparable to the methods trained on real images. In this paper, we propose a domain adaptation-based approach to address this issue. Specifically, the proposed approach mainly contains a model consisting of a multi-level constrained pose regression network and a feature-level discriminator network. The discriminator network forces the pose regression network to generate domain-invariant features from real and synthetic images by adversarial learning and thus reduces the performance gaps. In addition, the multi-level constraints further enhance the localization accuracy of pose regression. We perform extensive experiments on open-source rendering images in different settings. The results show that the proposed method significantly improves the performance. The code for the proposed work is available at https://github.com/lqing900205/BIM_domainadaptation.

Deep neural networks (DNNs) have been applied in various machine learning tasks with the success of deep learning technologies. However, they are surprisingly vulnerable to adversarial examples, which can easily fool deep neural networks. Due to this drawback of deep neural networks, numerous methods have been proposed to eliminate the effect of adversarial examples. Although they do play a significant role in protecting deep neural networks, most of them all have one flaw in common. They are only effective for certain types of adversarial examples. This paper proposes an ensemble denoiser based on generative adversarial networks (GANs) to protect deep neural networks. This proposed method aims to remove the effect of multiple types of adversarial examples before they are fed into deep neural networks. Therefore, it is model-independent and cannot modify deep neural networks’ parameters. We employ a generative adversarial network for this proposed method to learn multiple mappings between adversarial examples and benign examples. Each mapping behaves differently for different types of adversarial examples. Therefore, we integrate these mappings as the ultimate method to defend against multiple types of adversarial examples. Experiments are conducted on the MNIST and CIFAR10 datasets. We compare this proposed method with several existing excellent methods. Results show that this proposed method achieves better performance than other methods when defending against multiple types of adversarial examples. The code is available at https://github.com/Afreadyang/ensemble-ape-gan

Neural tube defects (NTDs) remain one of the most life-threatening birth defects affecting infants. Most patients with a NTDs eventually develop lifelong disability, which cause significant morbidity and mortality and seriously reduce the quality of life. Therefore, identifiable of novel pathogenic strategies for NTDs patients is urgently required. Increasing evidence indicates that METTL3-mediated m6A modification is present in many physiopathological processes of cell apoptosis and survival. Our results demonstrate that SAM play not only a compensatory role, SAM can also lead to m6A modification changes in neural tube development and regulation. This research provides a good theoretical basis for further research on folic acid deficiency leading to NTDs, reveals the important role of SAM in the development of NTDs, and provides new clue for clinical researchers and clinical work.

Impact sensors are widely used to detect grain losses in harvesters. Using polyvinylidene fluoride (PVDF) films as sensing elements is a promising way to improve sensor performance due to their high sensitivity, stability, and flexibility. However, the overlap of collision signals significantly reduces the accuracy of a sensor. To solve this problem, a novel impact sensor with two crossed PVDF films was designed and investigated. This sensor has two orthogonal layers of sensing elements that both respond to impacts, which creates positioning information for the impacts. Because of the sensor’s structure, a signal processing method was designed based on multisensor fusion theory. Tests were performed to verify the performance of the proposed impact sensor. The average signal-to-noise ratios (SNRs) for impacted PVDF films were 34.79 and 20.23 dB, respectively, for the upper and lower layers, while the average signal-to-clutter ratios (SCRs) for nonimpacted films were 21.90 and 10.05 dB, respectively. The sensor also has an extremely high detection efficiency of at least 1528 collisions per second and can identify particles that impact at the same time. Keywords: Grain loss detection, Impact sensors, Multisensor fusion, Particle impact tests, PVDF films.

Light detection and ranging (LiDAR) technologies, the active imaging technology, show a great potential in applications like threedimensional imaging. Combined with the frequency-modulated continuous-wave (FMCW) technology, LiDAR can provide the velocity of the target based on the Doppler effect. The linearization of the laser frequency sweep is crucial for higher resolution and lower signalto-noise ratio in FMCW LiDAR. In this paper, a proportional integral derivative (PID) control method based on a back propagation (BP) neural network is used to pre-calibrate the DFB laser to realize the high-linearity frequency sweep. The laser sweeping frequency is 12.5 kHz with a laser frequency excursion of 3.7 GHz, and the residual nonlinearity is 0.0015%. The linearized DFB laser is applied to the FMCW LiDAR system for two-dimensional imaging with a maximum distance of 70 meters.

Introduction: Long noncoding RNAs (lncRNAs) have been reported to be involved in the occurrence and development of various diseases. This study was to investigate the role of lncRNA-H19 in the transition from acute kidney injury (AKI) to chronic kidney disease (CKD) and its underlying mechanism. Methods: Bilateral renal pedicle ischemia-reperfusion injury (IRI) was used to establish the IRI-AKI model in C57BL/6 mice. The expression levels of lncRNA-H19, miR-196a-5p, α-SMA, collagen I, Wnt1, and β-catenin in mouse kidney tissues and fibroblasts were determined by quantitative real-time PCR and Western blotting. The degree of renal fibrosis was evaluated by hematoxylin and eosin staining. The interaction between lncRNA-H19 and miR-196a-5p was verified by bioinformatics analysis and luciferase reporter assay. Immunohistochemistry and immunofluorescence were used to evaluate the expression of α-SMA and collagen I in kidney tissues and fibroblasts of mice. Results: lncRNA-H19 is upregulated, and miR-196a-5p is downregulated in kidney tissues of IRI mice. Moreover, miR-196a-5p is a direct target of lncRNA-H19. lncRNA-H19 overexpression promotes kidney fibrosis and activates fibroblasts during AKI-CKD development, while miR-196a-5p overexpression reversed these effects in vitro. Furthermore, lncRNA-H19 overexpression significantly upregulates Wnt1 and β-catenin expression in kidney tissues and fibroblasts of IRI mice, while miR-196a-5p overexpression downregulates Wnt1 and β-catenin expression in kidney tissues and fibroblasts of IRI mice. Conclusion: lncRNA-H19 induces kidney fibrosis during AKI-CKD by regulating the miR-196a-5p/Wnt/β-catenin signaling pathway.

Dynamic resistance can reflect the ablation of arc contacts in circuit breaker at the moment of opening and closing operation, and thus indirectly reflect the condition of the circuit breaker and its subsequent breaking reliability, providing a new idea for the study of the reliability to the electrical life of circuit breaker. Based on the long-term tracking and data recording of a 550 kV circuit breaker, this paper studies the relationship between the equivalent breaking times and the dynamic resistance in the whole electrical life of the circuit breaker with the measurement of the dynamic resistance. The results show that the equivalent breaking times are strongly related to the split-point resistance with a monotonic increasing trend.

In recent years, deep reinforcement learning (DRL) achieves great success in many fields, especially in the field of games, such as AlphaGo, AlphaZero, and AlphaStar. However, due to the reward sparsity problem, the traditional DRL-based method shows limited performance in 3D games, which contain much higher dimension of state space. To solve this problem, in this paper, we propose an intrinsic-based policy optimization (IBPO) algorithm for reward sparsity. In the IBPO, a novel intrinsic reward is integrated into the value network, which provides an additional reward in the environment with sparse reward, so as to accelerate the training. Besides, to deal with the problem of value estimation bias, we further design three types of auxiliary tasks, which can evaluate the state value and the action more accurately in 3D scenes. Finally, a framework of auxiliary intrinsic-based policy optimization (AIBPO) is proposed, which improves the performance of the IBPO. The experimental results show that the method is able to deal with the reward sparsity problem effectively. Therefore, the proposed method may be applied to real-world scenarios, such as 3-dimensional navigation and automatic driving, which can improve the sample utilization to reduce the cost of interactive sample collected by the real equipment.

Research in word representation shows that isotropic embeddings can significantly improve performance on downstream tasks. However, we measure and analyze the geometry of pre-trained BERT embedding and find that it is far from isotropic. We find that the word vectors are not centered around the origin, and the average cosine similarity between two random words is much higher than zero, which indicates that the word vectors are distributed in a narrow cone and deteriorate the representation capacity of word embedding. We propose a simple, and yet effective method to fix this problem: remove several dominant directions of BERT embedding with a set of learnable weights. We train the weights on word similarity tasks and show that processed embedding is more isotropic. Our method is evaluated on three standardized tasks: word similarity, word analogy, and semantic textual similarity. In all tasks, the word embedding processed by our method consistently outperforms the original embedding (with average improvement of 13% on word analogy and 16% on semantic textual similarity) and two baseline methods. Our method is also proven to be more robust to changes of hyperparameter.

This paper proposes a improved method for measuring the focal length of a telephoto lens. In this paper, a photoelectric theodolite with high accuracy of angle measurement is used. Besides, a CMOS camera is used to take a position image of the center point of the theodolite, and the corresponding azimuth displayed by the theodolite is recorded. Leveling the optoelectronic theodolite on the open field, Polaris is used as a point light source outdoors. Adjusting the point target on the CMOS image, by which the target is placed in the view center of the field (crosshair center). And then, the pitch and azimuth data of the device are recorded. Adding disturbance to the photoelectric theodolite, the CMOS images is recorded. Adapting methods of target feature segmentation, sub-pixel positioning is performed on the point target image which captured by the CMOS camera. And the point target coordinates is obtained. Finally, the result based on the theodolite angle offset, sub-pixel offset and elevation angle data is calculated. The perturbation data is averaged to calculate the focal length of the telephoto lens. Compared with equipment testing result of focal length, experiments have achieved high measurement accuracy. And this method is suitable for detecting the focal length of an optoelectronic device in an outdoor environment.

The increasing peak-to-valley load difference in China pose a challenge to long-distance and large-capacity hydropower transmission via high-voltage direct current (HVDC) lines. Considering the peak shaving demands of load centers, an optimization model that maximizes the expected power generation revenue is proposed here for the long-term operation of an interprovincial hydropower plant. A simulation-based method was utilized to explore the relationships between long-term power generation and short-term peak shaving revenue in the model. This method generated representative daily load scenarios via cluster analysis and approximated the real-time electricity price of each load profile with the time-of-use price strategy. A mixed-integer linear programming model with HVDC transmission constraints was then established to obtain moving average (MA) price curves that bridged two time-coupled operations. The MA price curves were finally incorporated into the long-term optimization model to determine monthly generation schedules, and the inflow uncertainty was addressed by discretized inflow scenarios. The proposed model was evaluated based on the operation of the Xiluodu hydropower system in China during the drawdown season. The results revealed a trade-off between long-term energy production and short-term peak shaving revenue, and they demonstrated the revenue potential of interprovincial hydropower transmission while meeting peak shaving demands. A comparison with other long-term optimization methods demonstrated the effectiveness and reliability of the proposed model in maximizing power generation revenue.

The space cable plays an important role in signal transmission and power distribution in spacecraft, and is an important part of connecting the subsystems and units. The number and weight of the space cables of spacecraft are increasing with the increasingly complex payload function of the spacecraft, which leads to a large increase in the cost of the whole satellite. The copper-clad aluminum (CCA) wire is 40% lighter than copper wire, which was used by Chinese satellite before. Some Europe and America communication satellite manufacturer use copper-clad aluminum (CCA) wire in order to reduce the satellite weight. Based on the study of crimping technology of light-weight copper-clad aluminum and electrical connector terminal which made in china, the tensile strength, sealing, metallographic and contact resistance of the crimping joints were analyzed. It was found that the hardness of the connector terminal, the position of the observation hole, and the match of wire gauge and terminal directly affect the crimping joints quality.

To investigate the diagnostic significance of lingual flange protrusion for liver metastasis of patients with malignant neoplasia.The data of 191 patients with malignant neoplasia were analyzed. All photos of patients' tongue image were recorded and lingual flange protrusion was the positive standard. χ2 test for paired data and Kappa test were used to determine the diagnostic value of lingual flange protrusion for metastatic liver cancer. Mann-Whitney U test was used to compare the levels of liver serological markers. The area under receiver operating characteristic curve (ROC) and logistic regression model were used to analyze the predictive values of lingual flange protrusion, alkaline phosphatase (ALP), and lactate dehydrogenase (LDH) levels.Patients with lingual flange protrusion had a higher risk of liver metastasis than those without it (P0.001). There was no significant difference in diagnosis of liver metastasis between lingual flange protrusion and traditional diagnostic criteria (P = 0.541). Kappa was 0.738 (P0.001). Lingual flange protrusion was significantly correlated with increased serum ALP and LDH levels (P0.01). Comparison of ROC curves showed that the diagnostic value of lingual flange protrusion is better than ALP, LDH and the combination of ALP and LDH (P0.01). Furthermore, the combined diagnostic values of lingual flange protrusion and ALP, lingual flange protrusion and LDH, and lingual flange protrusion, ALP and LDH are not better than lingual flange protrusion alone (P0.05).Lingual flange protrusion is a potential diagnostic marker for liver metastasis of patients with malignant neoplasia.

3D localization of point source is widely used in many fields, such as bioimaging and autonomous driving fields. However, the localization is hard to perform under scattering conditions because of the diffuse effect of the scattering. We propose a novel method for 3D localization of point source under scattering conditions based on light field imaging and deep learning by only one shot. First, we introduce the description of the point source in a light field wise and how to localize a point source by its light field. On the basis, we elaborate on the effect of scattering on a light field and how to retrieve the location of a point source from a light field with scattering. Then, the effect of aberration on a light field will be introduced. We also build an artificial scene and a deep learning framework to perform a 3D localization practically, and the feasibility and accuracy of our method have been evaluated.

The combination of the passivated emitter and rear cells (PERC) technology and the laser selective emitter (SE) preparation process was one of the effective methods to improve the efficiency of the PERC cells developed in recent years. However, in the SE preparation process, the phosphorosilicate glass (PSG) layer was prone to ablation, which led to the degradation of solar cell performance. By studying the effects of laser wavelength, laser power, pulse frequency and scanning speed on the surface morphology, surface sheet resistance and solar cells performance of silicon wafers, the preparation method of high efficiency and low damage SE solar cells was explored. The experimental results show that the optimun cell performance can be obtained under the condition of laser wavelength of 1064 nm, laser power of 2 W and pulse frequency of 30 Hz. The conversion efficiency of SE solar cells is up to 22.01%, which is about 0.51% higher than that of the solar cells prepared by traditional process.

We experimentally demonstrate a $1\times 128$ optical phased array, and an improved genetic algorithm is proposed to efficiently increase the side lobe suppression ratio (SLSR). The measured SLSR can achieve 20 dB with a divergence angle of 0.16°. © 2020 The Author(s)

Project selection is a crucial decision-making process in many organizations. By adopting a project-based learning perspective, this study sets out to develop a framework to integrate organizational knowledge development with project selection. We utilize various knowledge management models to create a structured evaluation metric to measure project contribution to organizational knowledge. A project selection model, which involves project knowledge contribution as one of the evaluation perspectives, is proposed. Results of a focus group study effectively validate the proposed evaluation metric. The article concludes with an empirical implementation of the model in an electronic component manufacturing company.

The wind turbine emulator(WTE) uses the motor to simulate the output characteristic of the wind turbine(WT) under the real wind field environment, and can effectively verify the various control strategies of the WT. This paper analyzes the working principle of the traditional WTE, discusses the torque and speed characteristics of the drive shaft, adopts the control strategy of acceleration compensation, and gives the motor torque that needs to be simulated. The shortcomings of the traditional WTE are pointed out. The range of moments of inertia of the WT is extended by adding a low-pass filter to the forward channel, and its stability and accuracy are analyzed. Finally, the effectiveness of the scheme is verified by simulation and experiment.

Hedging rules for hydropower operation provide optimal decisions to hedge against extreme system loss from an energy shortfall in the future at the price of a small current loss. Aiming at ...

Most of the existing person re-identification (ReID) methods use a classification network pre-trained on external data as the backbone and then fine-tune it, which results in a network architecture that is fixed and dependent on pre-training of external data. There are also some methods that are specifically designed by human experts for ReID, but manual network design becomes more difficult as network requirements increase and often fails to achieve optimal settings. In this paper, we consider using emerging neural architecture search (NAS) technology as a tool to solve above problems. However, most of NAS methods deal with classification tasks, which causes NAS to not be directly extended to ReID. In order to coordinate the inconsistency between the two optimization goals, we propose to establish an objective function with the assistant of the triplet loss to guide the direction of architecture search. Finally, it is no longer dependent on external data to automatically generate a ReID network with excellent performance using NAS directly on the target dataset. The experimental results on three public datasets validate that our method can automatically and efficiently find the network architecture suitable for ReID.

Objective: Patients with difficult-to-control asthma have difficulty breathing almost all of the time, even leading to life-threatening asthma attacks. However, only few diagnostic markers for this disease have been identified. We aimed to take advantage of unique Chinese medicine theories for phenotypic classification and to explore molecular signatures in difficult-to-control asthma. Methods: The Chinese medicine syndrome differentiation algorithm (CMSDA) is a syndrome-scoring classification method based on the Chinese medicine overall observation theory. Patients with difficult-to-control asthma were classified into Cold- and Hot-pattern groups according to the CMSDA. DNA methylation and metabolomic profiles were obtained using Infinium Human Methylation 450 BeadChip and gas chromatography-mass spectrometer. Subsequently, an integrated bioinformatics analysis was performed to compare those two patterns and identify Cold/Hot-associated candidates, followed by functional validation studies. Results: A total of 20 patients with difficult-to-control asthma were enrolled in the study. Ten were grouped as Cold and 10 as Hot according to the CMSDA. We identified distinct whole-genome DNA methylation and metabolomic profiles between Cold- and Hot-pattern groups. ALDH3A1 gene exhibited variations in the DNA methylation probe cg10791966, while two metabolic pathways were associated with those two patterns. Conclusions: Our study introduced a novel diagnostic classification approach, the CMSDA, for difficult-to-control asthma. This is an alternative way to categorize diverse syndromes and link endotypes with omics profiles of this disease. ALDH3A1 might be a potential biomarker for precision diagnosis of difficult-to-control asthma.

This paper focuses on the monthly operations of an interprovincial hydropower system (IHS) connected by ultrahigh voltage direct current lines. The IHS consists of the Xiluodu Hydropower Project, which ranks second in China, and local plants in multiple recipient regions. It simultaneously provides electricity for Zhejiang and Guangdong provinces and thus meets their complex operation requirements. This paper develops a multi-objective optimization model of maximizing the minimum of total hydropower generation for each provincial power grid while considering network security constraints, electricity contracts, and plant constraints. The purpose is to enhance the minimum power in dry season by using the differences in hydrology and regulating storage of multiple rivers. The TOPSIS method is utilized to handle this multi-objective optimization, where the complex minimax objective function is transformed into a group of easily solved linear formulations. Nonlinearities of the hydropower system are approximatively described as polynomial formulations. The model was used to solve the problem using mixed integer nonlinear programming that is based on the branch-and-bound technique. The proposed method was applied to the monthly generation scheduling of the IHS. Compared to the conventional method, both the total electricity for Guangdong Power Grid and Zhejiang Power Grid during dry season increased by 6% and 4%, respectively. The minimum monthly power also showed a significant increase of 40% and 31%. It was demonstrated that the hydrological differences between Xiluodu Plant and local hydropower plants in receiving power grids can be fully used to improve monthly hydropower generation.

ObjectiveTo test the synergistic effects of the aqueous extract of Tubeimu (Rhizoma Bolbostemmatis) and Fuzi (Radix Aconiti Lateralis Preparata) on MDA-MB-231 and SKBR3 breast cancer cells.MethodsA combined index was created for the effects of Tubeimu (Rhizoma Bolbostemmatis) and Fuzi (Radix Aconiti Lateralis Preparata) extracts. Cell proliferation was performed by trypan blue exclusion and 3-(4,5-dimethylthiazol-2-yl)-5-(3- carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assays. Flow cytometry was used to assess cell cycle distribution and apoptosis. Cell migration was determined by wound-healing and transwell assays. Confocal microscopy was used to detect E-cadherin and actin filaments.ResultsThe aqueous extract from Tubeimu (Rhizoma Bolbostemmatis) and Fuzi (Radix Aconiti Lateralis Preparata) exerted synergetic effects on the growth of MDA-MB-231 cells and G1 phase arrest. When exposed to extracts at concentrations of 62.5:62.5 and 62.5:31.3 µg/mL, the combination index was 0.83 and 0.74, respectively. Interestingly, 62.5: 31.3 µg/mL of combined drugs enhanced the inhibitory effect of Tubeimu (Rhizoma Bolbostemmatis) on the migration of SKBR3 cells and reduced the stimulative effect of Fuzi (Radix Aconiti Lateralis Preparata) (P < 0.01), in which cells showed an increased expression of E-cadherin and reorganization of actin filaments (P < 0.001). 62.5:62.5 µg/mL extract also synergistically induced apoptosis of MDA-MB-231 cells (P < 0.05 and P < 0.001). Acting as the main active ingredients in the extract, tubeimoside I and acetylbenzoylaconine at 10: 10 µg/mL and 5: 2.5 µg/mL also produced inhibitory effects on the proliferation and migration of cells (P < 0.01).ConclusionTubeimu (Rhizoma Bolbostemmatis) and Fuzi (Radix Aconiti Lateralis Preparata) extracts had synergic effects on MDA-MB-231 and SKBR3 cells.

Ongoing advances in digital-array technology have led to the development in fields such as radio astronomy, radar, and deep-space exploration. In this paper, we introduce the world's largest synthesis radio telescope, the Square Kilometer Array (SKA), and new-generation radio-telescope devices. We analyze the characteristics and performance advantages of digital-array system architecture in radio telescopes. We discuss in detail three kinds of low-frequency aperture-array system architecture based on the SKA work package in KLAASA, and we present a practical system scheme.

Using the band gap properties of the periodic structure, the truss structure of an offshore platform is designed as a one-dimensional periodic conduit structure, and the band gap characteristics are used to realize the vibration control of the truss structure of the offshore platform. The finite element method is used to simulate the vibration characteristics of the periodic conduit, and the influence of the structural parameters (length and wall thickness) of the periodic conduit on the vibration band gap of the periodic conduit is analysed. The frequency response characteristics of the periodic truss structure of an ocean platform are calculated and compared with those of a traditional truss structure. The results show that the periodic truss exhibits better suppression of vibration transmission than the traditional truss, especially for the vibration in the band gap, which is greatly attenuated. This study shows that the periodic structure has broad application prospects in vibration isolation design of truss structures for offshore platforms. The research provides a new technical approach for vibration control of truss structures on offshore platforms.

The research content of this paper is to improve the high temperature fault problem of the wind power inverter and optimize the processing process of the high temperature fault problem. In order to solve the high temperature fault problem, the water-cooling system of inverter is analyzed in detail by using the function analysis, cause and effect analysis and object field analysis in the invention problem solving theory (TRIZ theory). At the same time, combined with the invention principle of TRIZ theory and the standard solution, seven different solutions are finally obtained. Based on the ECRS principle, the general process of high temperature fault problem treatment is simplified through continuous optimization and improvement.

This work focuses on designing a self-sufficient trigeneration system for low-grade heat resource applications. For this purpose, a novel transcritical CO2 combined cooling, heating, and power (CCHP) system is proposed, which integrates a Rankine power cycle and an ejector refrigeration cycle (ERC). To evaluate the feasibility of the proposed system, the mathematical model of the combined cycle is built and validated. The effects of seven key parameters on system performance are investigated from the thermoeconomic viewpoint. Furthermore, multi-objective optimization is performed for the system when it generates both cooling and power or simultaneously produces heating and power. The results show that when using low-grade heat resource, the proposed plant not only has desirable net power output under all considered conditions, but also achieves adjustable output for cooling and heating in a broad range. The exergy efficiency of the proposed system and the coefficient of performance (COP) of ERC under base case conditions are respectively improved by 13.3% and 167.7% compared to the reference cycle. Under optimal conditions, the total useful energy and exergy efficiency of the system respectively are 127.3 kW and 22.7% for the combined cooling and power (CCP) mode, while corresponding values are 126.2 kW and 43.6% for the combined heating and power (CHP) mode, respectively. The cost per unit of exergy products for the system in the CCP mode is 3.4 times more than that in the CHP mode.

With the rapid growth of Internet of Thing(IoT) and mobile internet(4G/5G), there will be a huge volume of data generated by things that are immersed in our daily life, and hundreds of applications will be deployed at the edge to consume these data. Cloud computing as the widely used high performance processing center is not efficient enough to support these applications emerging in IoT and mobile Internet. It is mainly shown in the following two aspects, 1) Centralized cloud center is hard to deal with explosive growth of edge data, 2) Transmitting massive data from edge devices to cloud centers causes network load and service delay increasing dramatically. Therefore, edge computing for massive edge data appears, which collaboration with cloud computing to solve the above problems and provide low latency service. Edge storage as one of the main ways of resource management in edge computing is an important support for low latency service, which is based on the prediction of resources according to user's needs, and basing on user's location and the state of the edge nodes, the resources are prestored in different edge servers to reduce the edge load and improve the response time. In this paper, a edge storage method based on multi-view learning is proposed, it predicts the resources that need to be prestored through scenario demand based on multi-feature linear discriminant analysis, and multiple edge servers cooperate to store the above resources based on their real time performance and user's location. The experiments are carried out based on edge computing environment which distributed at three universities, and experimental results show that the proposed method is more effective than cloud storage when dealing with part of the mobile application.

With the large-scale promotion of cloud computing, intrusion detection is a necessary way to guaranteen cloud security. However, because of the lack of adaptive model, detection accuracy is still a challenge issue in intrusion detection. In our works, based on attribute significant expressing cooperative deep belief network (CDBN) was proposed for specific attack in intrusion detection. Firstly, a specific attack multi-view division method was proposed to extract the significant features of specific attack. secondly, an adaptive coding mechanism based on multi-view encoding was described to denoise and compress the attack features.finally, based on cumulative prospects, an cooperative decision-making deep belief network was proposed for cooperative intrusion detection. Thought testing and verified on the NSL-KDD data set, it proved that our proposed method has good applicability and high detection rate compared to the current general model.

Xiluodu Hydropower Plant with 13.86GW is the second largest in China and the third largest in the world. 18 hydropower units are equally installed in its left and right sides, which respectively provide electricity for Zhejiang Power Grid and Guangdong Power Grid. Thus, the plant is operated by two different dispatching departments. This is a novel hydropower operation problem facing huge hydropower reservoir. The paper develops a method for coordinating the operation of Xiluodu and many hydropower plants in the receiving power grids. In this method, the complex minimax objective function is transformed into an easily solving linear formulation. Moreover, an ideal point algorithm is used to cope with multi-objective optimization. The developed method has been demonstrated by a case study about a practical engineering. The results show that both the hydropower firm generations of Guangdong Power Grid and Zhejiang Power Grid during dry season are effectively enhanced. It is demonstrated that the hydrological differences between Xiluodu Plant and local hydropower plant in receiving power grids can be fully used to improve the monthly hydropower generation.

Monitoring networks in underground mining environment should provide many of mobile devices with the same time multi-access communications, high speed data transmission, and low transmit power capability. Bandwidth limitations, propagation loss, noise, interference, time variance, and frequency selective fading caused by multipath propagation can bring high symbol error ratio in high rate transmission. Communication systems can enhance their performance by using anti multi-path fading techniques only when they could make precise and real time estimation of fading coefficients of multipath channel. The purpose of this paper is for optimizing the wireless network system designs, and providing theoretical guides and technology programs for promoting the anti-multipath fading performance of wireless emergency networks.

The Sloan Digital Sky Survey (SDSS) has released the latest data (DR13) which is the first data release for the MaNGA survey. The massive spectra produced by SDSS is available for large sample research and also data source of rare and special objects like white dwarf main-sequence star (WDMS). Many techniques have been proposed to solve SDSS automatic classification problems with massive spectra and have important significance for improving the rate of scientific output of telescope. In this paper, an improved Fisher classifier is implemented for the classification of WDMS spectra in SDSS DR13. The inspiring experimental results show that this method is practical. In addition, this method can also be applied to the automatic classification of other specified objects in massive spectra without human involvement.

To improve the dewatering of coal slurry, Zheng Gangfeng and his colleagues at Anhui University of Science and Technology in China have developed and built a new design of bowl dehydrator with a spiral screen. Here, they discuss its operating principles and salient features, and present test results that demonstrate its potential for industrial-scale application.

This temperature measurement system is based on the control of AT89S52 single chip microcontroller, with temperature display function. Temperature information is changed into digital signal directly by a single bus digital temperature sensor DS18B20, and the digital signal is sent into microcontroller for real-time acquisition. The results are displayed through LCD module 1602 in real time. The system is with the characteristics of low cost, high precision, stable performance and convenient operation. The system can be extended to multi-point temperature detecting system easily, and adding temperature transfinite alarm function.

The authors designed a set of adherent self-circulation evaporative cooling system for a general desktop computer, and also tested its performance.The results show that: this cooling system can effectively dissipate heat of the CPU, the temperature will be kept below 50°C while operating at full load; For this application,we remove the fan of the CPU and achieve zero-noise and zero-power of computer equipment basically.

In the premise of introducing existing seven kinds of commonly used robot architectures at home and abroad and analyzing their strengths, weaknesses as well as applications, this paper shows the CPS-R architecture which is applied in the internal of robots to solve the problem in unknown dynamic environment and deeply compliant operation in detail where the current robot architecture cannot meet. Through describing the function and data flow of each component in CPS-R such as external environment in real-world, sensors, actuators, device interfaces in the physical layer, message generation and task allocation, network security authentication and prioritization, information collection, analysis, decision-making and sharing in the information layer. This paper analyzes the challenges of CPS-R architecture such as feasibility, real-time, security, reliability, intelligence, hardware/software standardization of information sharing in the end.

Compared with the traditional 2D electronic map navigation, the visual navigation has large amount of data. Especially when the viewpoint changes with high speed, navigation devices can’t display smoothly. Based on the above problem, this paper carried out the research from three aspects which could influence the speed of terrain visualization, that were the storage index of terrain data, data scheduling and occlusion culling. By adopting appropriate storage method, putting forward a new data scheduling strategy, and improving occlusion culling algorithm, the efficiency of the terrain rendering had been improved effectively. Experiments show that this method can meet the basic requirements of visual navigation.

Radial ventilation channel is a key component of large turbo generator stator, which affects the cooling of stator core and bars. The design of radial ventilation channel need simultaneously consider the flow resistance and cooling capacity. In this paper, the width of radial ventilation channel, the structure and material of support parts will be discussed to improve the cooling capacity of large air-cooled turbo generator stator. The theories of Computational Fluid Dynamics (CFD) and heat transfer are used to calculate the flowing field and temperature field, and the simulation results will be compared with experimental data of a 150MW air-cooled turbo generator.

In this paper, the attenuation characteristics (attenuation vs microwave pulse power) of the microwave attenuator with wedge-type carbonyl iron (CI) under high power microwave (HPM) pulses are investigated. A reliable experimental platform was set up and used to measure the attenuation under HPM pulses. A database describing the attenuation dependence of the injected microwave pulse power (MPP) up to 100kW on various pulses width of 200ns to 1us and repeating frequency of 10Hz to 500Hz is created. The results indicate that (I) the attenuation decrease with the increasing of the injected MPP; (II) with the attenuation increasing, the variation of attenuation in whole injected MPP range increase; (III) the effects of different pulses width and repeating frequency on attenuation characteristics can be ignored.

Inner evaporative cooling way of stator bars is generally affected by circulating system and coolant. In this paper, we focus those factors to improve the capability of cooling system. Based on the flowing resistance distribution of system, the structure of components is redesigned to improve the stability and reliability of circulating flow. In the same time, different kinds of coolant are used into the inner evaporative cooling experiment model. The experiment results show that boiling point and viscosity of coolant are the main factors of improvement. The cooling efficiency of hollow strands is improved by new experiment model and coolant, and the structure of improved model and coolant ZF-2 can perfectly satisfy the cooling requirement of 300MW turbo generator.

Ensemble techniques have been successfully applied in the supervised machine learning area to increase the accuracy and stability of base learner. Recently, analogous techniques have been investigated in unsupervised machine learning area. Research has showed that, by combining an ensemble of multiple clusterings, a superior solution can be attained. In this paper, we solve the cluster combination problem in term of finding a "best" subspace and formulate it as an optimization problem. Then, we get the solution according to basic concept and theorem in linear algebra whereupon a novel cluster combination algorithm is proposed. We compare our algorithm with other common cluster ensemble algorithms on real-world datasets. Experimental results demonstrate the effectiveness of our algorithm.

Big data storage is the foundation of big data processing and analysis. By researching and summarizing main processing technology of data storage, this paper respectively investigates and analyzes the following four aspects: distributed file system, NoSQL database, database appliance and new-type data storage technology of MPP architecture. In addition, this paper gives some recommendations applicable to different environments in favor of grasping the development states of data storage technology from different angles. This paper summarizes file segmentation, appropriate scenarios and merits and faults of distributed file system, and mainly analyzes and summarizes the theories and appropriate scenarios of four data storage models of NoSql database. Furthermore, this paper investigates and concludes the developments and features of database appliance minutely. At the same time, outline MPP (Massively Parallel Processing) architecture, a new data storage technology. At last, the research trends of storage technology are prospected, providing references to the research of big data storage technology.

3D system-in-package has recently been considered a major enabler for high density and heterogeneous microsystem integration. We hereby proposed the concept of a unified micromachining and packaging platform based on LTCC (low temperature cofired ceramic) material system and process, which is implemented by first enhancing an existing LTCC hybrid IC fabrication line and then integrating different LTCC micromachining process modules one by one. Hence, the unified process flow can be accomplished within just one single package-test house. The platform has been capable of micromachining basic 3D MEMS (micro electromechanical system) microstructures into LTCC laminates and using them as a packaging substrate for mounting IC/MEMS from other process platforms, realizing self-contained and versatile microsystems of high density. The 3D microstructures formation process consisting of green tape machining, lamination and cofiring are demonstrated. The designing, analysis and fabricated samples of various micro functional structure enabled by the platform are illustrated, including embedded cooling microchannels (capable of lowering substrate temperature by more than 50K), microaccelerometer for harsh environment, micro Pirani gauge for in-situ vacuum level monitoring and THz (tera hertz) vacuum microelectronic devices. Samples of overall packaged MEMS and IC chips with micromachined LTCC substrate are displayed, showing ultra-low leakage (< 5×10−11 Pa·m3/s) vacuum packaging capability and significantly enhanced device performance. In addition, the platform has demonstrated the potential of stacking several laminates with mounted chips into a 3D frame-like microsystem. In comparison, 3D integration purely based on Si micromachining, e.g. anodic-bonding based in-situ wafer encapsulation, may only support a very limited spectrum of devices/materials and integration density and is somehow too expensive for many MEMS researchers.