Search

Pansharpening is a very debated spatio-spectral fusion problem. It refers to the fusion of a high spatial resolution panchromatic image with a lower spatial but higher spectral resolution multispectral image in order to obtain an image with high resolution in both the domains. In this article, we propose a novel variational optimization-based (VO) approach to address this issue incorporating the outcome of a deep convolutional neural network (DCNN). This solution can take advantages of both the paradigms. On one hand, higher performance can be expected introducing machine learning (ML) methods based on the training by examples philosophy into VO approaches. On other hand, the combination of VO techniques with DCNNs can aid the generalization ability of these latter. In particular, we formulate a $\ell _2$ -based proximal deep injection term to evaluate the distance between the DCNN outcome, and the desired high spatial resolution multispectral image. This represents the regularization term for our VO model. Furthermore, a new data fitting term measuring the spatial fidelity is proposed. Finally, the proposed convex VO problem is efficiently solved by exploiting the framework of the alternating direction method of multipliers (ADMM), thus guaranteeing the convergence of the algorithm. Extensive experiments both on simulated, and real datasets demonstrate that the proposed approach can outperform state-of-the-art spatio-spectral fusion methods, even showing a significant generalization ability. Please find the project page at https://liangjiandeng.github.io/Projects_Res/DMPIF_2020jstars.html .

Neurological injuries headed by hemiplegia are often the leading cause resulting in coordination decay. Bimanual robotic systems have demonstrated promising efficacy in recovering coordination of bilateral limbs. The main principle that bimanual robots work for physical therapy is through delivering motor tasks to resemble activities of daily life (ADLs). Current evidence also indicates that bimanual training outcomes can be improved by integrating the senses of haptics. However, majority of robotic systems have not been developed with the integration of haptic feedbacks, especially when required to meet the workspace of ADLs. This study sought to address this issue by developing a new haptic-integrated robotic system capable of delivering bilateral tasks. The system is implemented with robotic motion control and a physical engine integrated in Unity3d. Human users are expected to be able to perform bimanual trainings through interacting with dual robotic handles. Experiments were conducted to examine position tracing performance of motion control and haptic transparency of bimanual tasks rendered by the robot. The experimental results led us to believe that the developed robotic system can deliver bilateral physical tasks with haptics integration in an ADL-required workspace. Future work will examine its haptic performance in terms of real force feedback.

This paper proposes $l_{2}$ norm, $l_{1}$ norm and $\iota _{\infty }$ norm of clustering optimization algorithms based on dictionary learning. By solving an optimization problem to assign each feature to a cluster and solving another optimization problem to re-calculating the vectors representing the clusters, each algorithm keeps iterating until it converges. Computer simulation experiments show that the three algorithms have good clustering results and the convergence is confirmed. The runtime of l2 norm clustering optimization algorithm is much faster than h norm and $ l\infty $ norm clustering optimization algorithms.

Artificial Neural Network is an effective technique of prediction needing a lot of time for training. Benefitting from data parallelization and structure parallelization, training time of network has been greatly reduced. In this paper, a hybrid parallel algorithm is proposed to further optimize training process. The experimental results indicate that the parallel algorithm owns ideal expansibility and great speed-up ratio up to 52.9 on KNL. When optimized with SIMD instruction set and MCDRAM, the maximum speedup ratio comes to 124.21.

Urbanization and enlargement of cities contributes to different functional regions in urban area. In this paper, we investigate several main patterns of human online behavior to discover different functional areas in cities. To that end, we conclude a method for automatic identification of urban land use from the information obtained by mobile data traffic. We consider data traffic as reflection of human activities due to its timeliness and pervasiveness. During the process we also propose two possible indicators from traffic flows to build bridge between data traffic and human activities. We apply our technique to real-world mobile data traffic collected by major mobile operators in China. Our results prove the existence of several typical patterns in large-scale data traffic of city and link them to specific urban land uses.

With the in-depth development of state grid and the improvement of technical means, the standardization, correctness and uniqueness of substation monitoring information have become increasingly demanding. In order to ensure the practical operability of information sharing between the source side and main stations of substations, the state grid monitoring message needs to achieve precise location of information and the strategy of self-integration and late optimization of data capacity lightweight. In this paper, Map-Reduce and ant colony algorithm is applied in a joint innovation. On the one hand, the monitoring information at the source side of substations is reduced by Map-Reduce technology; on the other hand, by the ant colony algorithm, the message is optimized the message after the light-weight processing, and the accurate monitoring information point is obtained. Through the verification, it not only improves the reading efficiency of the massive monitoring information message, but also deepens the understanding of the message information itself.

In rehabilitation, passive control mode is common used at early stages of the post-stroke therapy, when the impaired limb is usually unresponsive. The simplest is the use of a proportional-integral-derivative (PID) feedback control which usually regulates the position or the interaction force along a known reference. Nonetheless PID method cannot achieve an ideal tracking performance due to dynamical uncertainties and unknown time-varying periodic disturbances from the environment. In order to minimize steady-state error with respect to uncertainties in exoskeleton passive control, Iterative Learning Control(ILC) and Neural PID control are proposed to improve the control effective of conventional linear PID. In this paper, two different control algorithms are introduced. Moreover, an experimental study on a 5-DOF upper limb exoskeleton with them is addressed for comparison.

This paper proposes a chaotic filter bank based computer cryptographic system for wireless applications. The signal is first decomposed via an analysis filter bank. Second, a pair of chaotic modulators is employed for performing the encryption on the subband signals. Third, the encrypted signals are upsampled and filtered so as to obtain the oversampled signals. Fourth, the oversampled signals are quantized via a multi-dimensional sigma delta modulator. Fifth, the quantized signals are encoded via a low density parity check code scheme. The encoded signals are transmitted via a wireless channel. At the receiver side, first the encoded signals are decoded via a low density parity check decoder. Then, a multi-dimensional filter is employed for separating the signals and the quantization noises. Next, the filtered signals are downsampled to obtain the same number of samples. After that, a pair of chaotic demodulators is employed for performing the decryption. Finally, a synthesis filter bank is used to reconstruct the original signal. In this paper, the filters in both the filter bank system and the multidimensional sigma delta modulator are designed. Computer numerical simulation results show that the proposed system outperforms the existing computer cryptographic systems.

Currently, pedestrian dead reckoning (PDR) is widely used in indoor positioning. Its principle is to recursively update the location of pedestrians by using the step count, step length and heading. To estimate the pedestrian heading, a common method in PDR is to utilize magnetometer measurements to estimate the heading of a smartphone. However, magnetometer is extremely vulnerable to magnetic interferences in indoor environments, resulting that magnetometer measurements are unreliable for heading estimation. Moreover, there is an offset angle between the device heading and the pedestrian heading, and the device heading may continuously change while the pedestrian heading keeps fixed. Therefore, the pedestrian heading cannot be simply replaced by the device heading. In this paper, we propose a step-wise method to accurately estimate the heading of a pedestrian who carries a smartphone at different positions. To be specific, the proposed method fuses the signals from the accelerometer, gyroscope and magnetometer of the smartphone to transform the device reference frame to the Earth reference frame, and then utilizes the horizontal angular velocities to establish an optimization problem, the analytical solution of which is obtained. Extensive simulations are carried out and confirm that the proposed method is more robust and more accurate than the existing algorithms.

This article addressed the problem of delay-dependent stability criteria for linear time system with interval time-delay. Firstly, we proposed a new inequality based on the Fourier Theory, more precisely on the Wirtingers inequalities, which is proved to encompass the Jensens inequality. Secondly, a less conservative stability condition is established in terms of linear matrix inequalities by employing delay-partition idea and Lyapunov functional method. Finally, in order to illustrate the potential gain of employing this new inequality with respect to the Jensens one and the effectiveness of the proposed results,an application to time-delay analysis is provided.

This article primarily discusses the problem about delay-dependent stability condition for systems with time-delays. Firstly, we propose a new inequality referred to the Wirtingers inequality, which is proved to encompass the Jensens inequality. Secondly, some less conservative stability conditions are established by constructing a new LKF. Finally, the availability of the provided result is stated by a classical example from the literature.

In this paper, we have given four confidence intervals estimator for the relative difference in matching pairs under inverse sampling. The performance of the proposed methods is evaluated with respect to their coverage probability, expected widths and RNCP(the ratios of the mesial non-coverage probability to the non-coverage probability). Our empirical results show that the confidence intervals based on the wald statistic produce similar coverage probabilities with the pregiven confidence level and smaller confidence interval widths.

Instance selection also named sample selection is an important preprocessing step for pattern classification. Almost all of the existing instance selection methods are developed for specific classifiers, such as nearest neighbor (NN) classifier, support vector machine (SVM) classifier. Few of them are designed for single hidden layer feed-forward neural networks (SLFNs) classifier. Based on sample entropy, this paper presents an instance selection method for efficient data classification with extreme learning machine (ELM), which is used to train a SLFN. The proposed method is compared with four state-of-the-art approaches by a series of experiments. The experimental results show that the proposed method can provide similar generalization performance but lower computation time complexity.

Numerical study is performed for an effective suppression of temperature rise in AlGaN/GaN high electron mobility transistors (HEMT) using our developed algorithm based on finite element method. Such a HEMT can be fabricated on SiC or other substrates with graphene inserted between the AlGaN and the GaN layers. The effects of both geometrical and physical parameters of the structure on the maximum temperature rise at the top surface of AlGaN layer are examined and compared. It is shown that as the semiconducting graphene has extreme large thermal conductivity in comparison with other semiconductor counterparts, it can be integrated into AlGaN/GaN HEMTs and other SOI active devices as an effective heat removal layer. Under such circumstances, their electrical performance and reliability can be enhanced greatly. Therefore, we are able to say that graphene has great potential in challenging thermal managements of some active devices operating at high temperature, which must be used for the development of some miniaturized system in package (SiP).

It is regarded that building energy efficiency service market (BEESM) in China is still in the government pushing and demonstration process after about ten years of development. Although Government implemented a lot of incentive policies, as the marekt supply main body, energy service company (ESCo) doesn't really develop rapidly. To anylyse the contradiction between the management policy and the practice, the paper establishes options gambling model and analyses influencing factor of ESCo behavior choice of entering market. The analysis result shows that the intertemporal characteristic of the contract energy management (CEM) project results in ESCO's wait and see attitude of implementing the contract energy management project, which enables the CEM investment to have the economy risk. When the income of CEM project cannot meet ESCO requirements, ESCo hesitates greatly whether to develop the CEM project or not. CEM project's anticipated income and building energy efficiency cost are the key factors. Therefore more income, less cost and controlling CEM risk is helpful to implement CEM project for ESCo.

This paper explores HRM system features used to build an entrepreneurial team which fits with strategy and contingencies. Our findings include: recruitment channels of personal social network; selecting through interviews and paying great attention to the years of employment, position and achievement; no trainings during the initial stage of entrepreneurship; no short-term performance assessment system but still having the requirements of strategy goals and performance; the overall compensation structure featuring low basic salary, more share, no bonus and welfare.

The space analysis model of the Runyang Cable-stayed Bridge and Suspension Bridge are established to perform modal analysis at different simplified environment temperature and to compare with the analysis result of the collected data from the safety monitoring of the Runyang Bridge when identification of the modal frequency are greatly affected by change of environment temperature. The results of the study show as follows: The relationship between temperature and modal frequency of both Runyang Cable-stayed Bridge and Suspension Bridge is negative correlation. Frequency of vertical bending model decreases gradually as the temperature grows. Frequency of high order model of the mid-span vertical bending of the Runyang Suspension Bridge have more notable influence than that of low order model while the Runyang Cable-stayed Bridge decreases. The related conclusion is beneficial to eliminate the effect of temperature on identification of the modal frequency.

Based on case engineering Runyang Bridge, the temperature gradient of flat steel box girder in suspension bridge was continuously observed. According to the analysis of measured datum, the temperature gradient on flat steel box girder was proposed. With three temperature gradient models (JTG D60-2004 specification, BS5400 specification and the temperature gradient which was proposed in this paper), the control sections stress was calculated by finite element program ANSYS. The calculated result indicates that the temperature gradient that was put forward in JTG D60-2004 specification and BS5400 specification for calculating stress of flat steel box girder was not suitable to apply to flat steel box girder and inclines to unsafe. The temperature gradient on flat steel box girder which was proposed in this paper was reasonable.

Unlike ordinary refined products, high waxy crude oil has a potential serious safety problem during transportation because of its non-Newtonian characteristic. Further, the operating parameters of the pipeline and physical properties of the crudes are varied. Hence a common practice for the waxy crude pipeline operator is to overheat the transported crude in order to avoid any potential safety problems. However, this could lead to high energy consumption and not necessarily guarantee the safety of the pipeline because the entire system is a non-linear system and the parameters contains a large amount of uncertainly. Therefore flow assurance which means ensuring the safety of hydro carbon transmission is of great importance for pipeline company in China. In the paper, the mathematical model for simulating non-Newtonian flow is briefly presented, then the flow assurance technique and its process are generally explained. The evaluation results are obtained by using the knowledge and the results show that with the flow assurance technique, the safety is assured in all cases and the energy consumption is significantly reduced by about 20% in one typical case. Further, the content of the flow assurance technique is based on the real field data and the experiment data of the transported crudes. Therefore the technique could be generally applied to all the hydro carbon pipelines if the data accuracy is guranteed.

Although the new type of network security incidents continue to occur, most security incidents are similar, the response methods have in common, so CBR (Case Based Reasoning) technology can be used to describe the successful experience of the past incident response. Based on past examples of how to develop rapid response strategy is the key to incident responses. Automated planning method can greatly improve the efficiency and level of decision making. According to the characteristics of incident responses, combined with automatic planning method, CBR technology and ontology technology, a novel approach of getting incident response methods is presented.

We report on a new iterative method for regularizing a nonlinear operator equation in Hilbert spaces. The proposed IIGRA algorithm is combination of nonlinear implicit iterative method and a gradient method for minimizing a modified Tikhonov functional. Under some not very strict assumptions, we will give a convergence result. Finally we present two numerical results for integral equation and parameter identification.

A new approach of fast color character extraction was proposed. Clustering algorithm was adopted in our method to differentiate between objective character regions and back- ground regions on the premise that character regions are nearly monochromatic. However, the key point of this approach was how to select suitable elements' features based upon the original image information and character's information; it directly determined extraction results. We brought forward a means for feature selection to resolve clustering's key problem, and proved the means is effective. Furthermore, in order to overcome the non-ignorable computational overhead of clustering, parallel clustering algorithm was designed and realized by CUDA. Experiments illustrated color character could be extracted satisfactorily even if these characters are surrounded by complex background; meanwhile, parallel improved method that executed by assist of GPU was more efficient than former method that executed by CPU only.

The growth of research staff is the interactive effect of research leaders, cultural environment of research organization and characteristics of research staff. Using the Theory of Servant Leadership (SL) and Altruism-Oriented Culture (AOC) for reference, we built a theoretical model on the Growth Performance (GP) of research staff. The statistics results of the survey indicated that: �7 The promotion of SL was very evident. �8 GP dimensions were significantly varied as for different combinations of AOC. �9 Because the combinations of AOC were different, the impacts of SL had been indicated as three patterns, which were total/ alternative/ opposite patterns. The study proved the adjusting effects of AOC and revealed the microcosmic mechanism and application conditions of SL's effect.

International entrepreneurship strategy is one of the most important operations of manufacturing industry nowadays. World Automotive enterprises all accelerate their international entrepreneurial process. Chinese automotive enterprises have also entered the international market actively. By reviewing Chinese automotive enterprises present situations, the authors put forward international entrepreneurial opportunity—orientation— performance model of auto enterprise. We collect data from main Chinese auto manufacturers to make empirical study by means of questionnaire. The study provides instructive significance for the international entrepreneurship management of automotive enterprises.

This paper sponsored by the NSFC 70702020, summarized previous opportunity recognition literature with two main approaches. According to the integrated approach, the authors provided a new comprehensive model of the entrepreneurial opportunity recognition characterized as one sort of entrepreneurial innovation process. Four stages comprised in the new model were described in details. In the end, several contextual factors were mentioned in order to increase the holistic validity of the model.

When a fault such as unbalance occurs in a turbogenerator set, sensors should be put on its bearing to detect vibration signals for extracting fault symptoms and then diagnose faults. But the relationships between faults and fault symptoms are too complex to get enough accuracy for industry application. In this paper, a new diagnosis method based on fuzzy neural network is proposed and a fuzzy neural network system is structured by associating fuzzy set theory with neural network technology. Especially, an effective fuzzy organization method for training samples is presented, fault symptoms are discretized by a focusing quantization method and are then fuzzified to obtain fuzzy sets. In addition, the standard fault data which is confirmed by application is added to standard fault case database in order to improve accuracy of diagnosis system. Finally, a vibration fault diagnosis system for 600 MW turbo-generator set is designed and realized by the proposed system structure based on fuzzy neural network, its running results showed that the new system could satisfy fault diagnosis requirement of large turbo-generator set, its accuracy varied from 92 percent to 98 percent.

Optical Coherence Tomography (OCT) is a new technique mainly used in biomedical imaging. Here we present a Particle-Fixed Monte Carlo (PFMC) simulation for OCT. The scattering particles of the sample are assumed to be temporarily fixed randomly in the simulation process of backscattering light. The new model explains the exponential decay signal at the interface of different media layers in OCT experimental measurement. To extend the application area of OCT, an ultrasonic motor of 1mm in diameter is used to implement the circular imaging scanning in OCT endoscopic probe.