Search

Cloud Computing is becoming more and more popular for solving problems that need high concurrency and a lot of resources. Many traditional areas of research choose to solve their problems through the cloud, and workflow scheduling is one of them. Cloud computing brings many benefits, meanwhile, due to the almost “infinite” amount of resources for users, it also brings new challenges for scheduling and optimization, in which cost and makespan are the most concerned issues for workflow scheduling. Users want to obtain a low cost and fast makespan solution. This paper focuses on how to find an optimized solution to achieve better cost-makespan at the same time under the constraint of deadline. In order to solve this problem, an immune particle swarm optimization algorithm (IMPSO) is proposed, which effectively improves the quality and speed of the optimization. The proposed IMPSO overcomes the problem of slow convergence of PSO, which is easy to fall into local optimization. Experiments show the efficiency and effectiveness of the proposed approach.

For crystalline material with strong terahertz (THz) absorbance, the characterization with terahertz time-domain spectroscopy (THz-TDS) is typically performed on the mixture pellet where the analyte of interest is embedded within an inert polymer matrix. The polymer background significantly affects the phase delay of the propagating THz signal inside the sample, which makes the polymer matrix contribute to the overall dielectric response of the mixture sample to the probing THz waves. In this study, an analytical model was proposed to extract the target analyte’s intrinsic dielectric constants from the experimental dielectric spectra of the diluted mixture.

With the development of the financial sector, the growing complexity of financial transaction network, effectively identify a trading network of key trader has a important significance. Trading network abstraction for complex networks, traders abstraction for nodes, trandings between traders abstraction for the edges. The method of degree centrality, clustering coefficient, betweenness centrality, closeness centrality and the like is not sufficient to evaluate the importance of the node. Therefore, we propose a novel algorithm to evaluate the importance of nodes in undirected and unweighted network. We take the degree centrality and clustering coefficient of the nodes as the evaluation indicators, and combine the importance contribution of the nearest and the next nearest nodes. Through normalization and averaging, the benchmark ranking of node importance is obtained, which comprehensively considers the global and local features of the nodes. We used the real trading network data from Zhengzhou Commodity Exchange (ZCE) to conduct three comparative experiments and analyses. The experiment results show that our method has achieved better results, and can effectively identify key trading traders in ZCE.

Crowdsourcing web services has received much attention in recent years, which has been widely used in many fields, and provides important human computing services for the rapid development of AI. Crowdsourcing knowledge acquisition is one of the most important applications, which includes a series of work such as image annotation and picture classification. However, due to the differences in the difficulty of these tasks and the uncertainty of workers, how to make a reasonable task assignment while ensuring the completion of tasks becomes a big challenge. To this end, we introduce WordNet external knowledge base to help determine the difficulty of picture classification tasks. We refer to the e-sports rank mechanism and use dynamic update strategy to assess the actual ability of workers. A novel criterion affinity based on the weighted Euclidean distance with penalty factor is proposed to measure the suitability between tasks and workers. On this basis, the Kuhn-Munkres (KM) algorithm is used to solve the weighted bipartite graph matching problem. Through comparative experiments, the effectiveness of our proposed method is verified.

500kV autotransformer has split type and integrated type. The current transformer (CT) position of low voltage side is different for these two types of autotransformer. This paper analyzes circulating current in autotransformer’s delta winding when single-phase grounding fault takes place on high voltage side. Protection setting method considering circulating current in delta winding is proposed for split type autotransformer. Simulation results verify the position of two types of autotransformer’s CT influences the circulation current in low voltage side protection. Protection setting for an autotransformer used in Shandong electric power grid tests the correctness of the over-current protection setting with circulating current proposed in this paper.

Magnetization reversals in magnetic recording media are largely carried out by brute force: a field is applied opposite to the existing magnetization direction of some bit that has sufficient magnitude to nucleate a seed that then grows into an oppositely magnetized bit. The fields used are generally quite large, $\sim 10$ kG, requiring elaborate magnetic circuitry to keep the fields localized so they do not spill over onto neighboring bits. This situation is to be contrasted with the resonant magnetization reversals performed in NMR spin echo experiments in which r.f. fields of a few Gauss coherently reverse the magnetization in the presence of static fields of a few kG, by applying a so-called Pi pulse; two such pulses restores the original alignment.

Online simulation systems are mostly model-driven without considering about real-time interactive controlling with physical systems. In order to enhance the two-way interaction control capability of digital twins and physical systems, this paper proposes a digital twin-driven model based on Petri Net. It defines a state incremental update place for enhancing the ability of real-time simulation, and defines a feedback control structure to control the function of opening and closing the real industrial pipelines. Taking the industrial parameters of the suit processing line as experimental data, the feasibility of online synchronization communication and feedback control of the digital twin driven model and the actual production line was verified.

With the development of cloud computing, the cloud market is becoming more and more complicated. There are many cloud providers and different cloud instance types, which brings users some confusion when they select cloud instance types. In order to solve the cloud instance type selection problem in multi-cloud environment, a Cloud Instance Type Selection Algorithm based on Genetic Algorithm (CITSA-GA) is proposed. CITSA-GA mainly includes two-dimensional encoding with the constraint between adjacent genes, selection operation adopting the elite retention strategy and the roulette strategy, crossover operation using the first fit strategy, and mutation operation with mutation bounds. We perform some experiments to prove the effectiveness of the proposed CITSA-GA.

In big data mining analysis, the data records in the dataset are randomly retrieved. The distributed storage modes, such as BigTable, HBase, provide the cache policy for file blocks in retrieval operations. Since these records are scattered in different file blocks, the block cache does not have a high hit rate. To deal with the above problem, we propose an LRU-based double queue K-frequency cache method (DLK). The method presents a double queue storage structure, applying different storage and eviction rules for the data with varying access frequency (i.e., high/low access frequency). While the method divides the memory into data area and list area and adopts different data structure to reduce the time of data retrieval and data processing. The experimental results show that proposed method can reduce retrieval time by 30% with the cache mechanism. Compared with existing methods DLK can improve the hit rate by 60.1% and reduce the retrieval time by 43.5%. While applying in smaller cache capacity, our method outperforms other algorithms.

With the rapid development and popularization, Internet is becoming the most convenient way to publish and obtain information, which causes an extremely increasing quantity and variety of data. It is difficult to find out potentially valuable information from these data, which is the primary problem of data mining. Mining company hot events from Internet news can effectively reflect how its business works. Thus, we propose a method for discovering and obtaining hot events from Internet news. In the proposed method, we use Gaussian kernel to update clustering center instead of global cluster to modify Single-Pass clustering algorithm. It is a dynamic incremental clustering algorithm which does not need to initialize the number of clusters. Then, Top-N hot events can be obtained through the clustering centers. Experimental comparison shows that the improved algorithm has higher clustering efficiency than the classic algorithm. Case studies from Shanghai pilot free-trade zone (FTZ) also show the effectiveness of our proposed method.

High efficiency pulse current transmission is critical important for inertial confinement fusion (ICF), dynamic material property research and other high energy density physics experiments, in which current density in several MA/cm will achieved on the electrode surface near the load. Independent on power generation, high current pulse transmission to load has to deal with structural transition in transmission line, massive electrons/ions insulation, ~MA/cm current density, megagauss (MG) magnetic field and its diffuse into conductor. The current loss across transition and the penetration of MG magnetic fields into conductors may cause anode-cathode gap closure, and even transmission failure. In order to understand the pulse transmission property in MA/cm range, two types of experiments were carried out on the Primary Test Stand. In the stainless steel electrode transmission experiments, the expansion velocity of conductor at several MG magnetic field was recorded with displacement interferometer system and the evolution process of dense plasma were recorded by shadowgraphy system. One dimensional magneto-hydrodynamics (MHD) simulations to the experiments were carried out and a threshold magnetic field of 3.3 MG for thermal plasma formation was inferred for stainless steel. In the coaxial-conical transition experiment, current transmission efficiency under MA/cm current density was derived for different states. In this paper, the pulsed power platform, the experimental designs and results will be presented.

In this age of information, data plays an increasingly important role. Nowadays, with the development and popularization of cloud computing, more and more enterprises and personal users choose to put their applications and data into the cloud. However, most of them tend to use single cloud because of convenience and cheap storage price in the short term. This may lead to some challenges and risks, such as vendor lock-in, data low-availability, and data privacy risks. Recently, multi-cloud has attracted considerable interests. In this paper, an ant colony algorithm-based approach is proposed for data hosting in multi-cloud environments, which is subjected to optimization objectives including cost and availability. Extensive experiments are performed to evaluate the proposed method, which are driven by real-world cloud services information.

To organize process of a number of events containing common topics, the paper proposed an event information organization algorithm by two text clustering, topic clustering which combines Latent dirichlet allocation model (LDA model) and TF-IDF model and event Hierarchical Clustering based on timeline. Topic clustering is a topic clustering through dynamic K-means in which the document vector distance is calculated. It can be known that events sets have a common topic by this clustering method. And another clustering method event hierarchical clustering is a clustering events phase based on timeline. In a word, this paper organizes event information by combining the above two clustering methods.

In this paper, we propose a novel method for classroom teaching evaluation by combining with speech analysis. The real time voice signals are transformed into discrete digital signals through hardwares, and we can access and deal with them in MATLAB. During the procedures, a lot of experiments and simulations have been made. The enrichment of rhythm, tone fluctuation and passion are judged by setting some evaluation criteria, and a preliminary conclusion about the teaching effects is provided.

High quality teaching has always been the pursuit of universities, but the automatic and real time evaluation on the quality of classroom teaching has not been achieved. To solve this problem, a real-time processing of classroom video through face detection technology and a software system to provide a basis for judging the quality of teaching is proposed in this paper. The main application of machine learning and deep learning is to build the face detection algorithm, and the application of C# and Hikvision SDK are for the second development building software interface and nestification of software systems.

This paper presents a novel two-stage Hough transform (HT) algorithm and how to use it on a lane detection system for vehicles and industrial robot based on TMS320DM6437. As is known to all, white and yellow are the basic color of lane mark, which can be easily extracted through YCbCr color space. In order to improve detection performance, the two-stage HT is proposed to act on a binary image which is produced by setting threshold for ‘Y’ component and ‘Cb’ component. As lane detection algorithm usually does, HT is used on a edge image first. Then the two-stage Hough transform use the same accumulator with the first time. Finally, an Enhanced Hough Transform Graph (EHTG) is generated. The EHTG contains the information both edge and color. Thus, we can utilize it to visualize lane and transplant the algorithm on DSP. After a series of optimization in programming, the result shows that the proposed algorithm give an outstanding performance on DM6437 platform despite of lane mark embedded in severe noise environment.

Air quality forecasting is an effective way to protect public health by providing an early warning against harmful air pollutants. Traditional methods mainly focus on neural network model. However, relationship between parameters is neglected in the learning process and model structure also limits the accuracy of forecasting. In this paper, a two-phase neural network (2-NN) model based on full connected neural network is proposed to forecast air pollutants (PM10, SO2, NO2, et al) concentrations over a relatively short period of time. The model 2-NN takes mapping, tendency, and correlation into consideration to improve forecasting accuracy. For the first phase, considering the mapping relation, five meteorological parameters are used as inputs, which are fed to a back propagation neural network to get the prediction of the first phase. For the second phase, considering the tendency and correlation, some historical time series of air pollutants and one of the most influential meteorological parameters are included to forecast the prediction of the second phase. The experimental results show that 2-NN model has better forecasting performance for the air pollutants than traditional methods like SVM, Neural Network, and KNN.

The study of community search problem subjected to find densely connected subgraphs that satisfy the query conditions in social and information networks has attracted many researchers. In most real life networks a query node normally involves in several communities, therefore, most likely the same query results into communities of different densities. In this paper, based on the concept of k-truss, which is a type of cohesive subgraph, we propose linear-time algorithms to find the top-r k-truss communities containing a query node in large graphs. We adapt some optimal index construction strategies which can support efficient search of top-r k-truss communities with a linear cost to work for top-r community search. We carry out extensive experiments on real-world large networks, and the results signify the efficiency and effectiveness of the proposed algorithms.

There exists massive information on the Internet, consequently information extraction has attracted more and more attention. Open relation extraction aims to get relational tuples from text without defining relation types in advance. This paper focuses on identifying relation phrases and associated arguments in Chinese arbitrary sentences, and proposes a domain-adaptive method to get entity relations from large amount of related Chinese texts in a supervised way. The experiments show that the proposed method can get both good accuracy and recall rate in the domain of Chinese listed companies.

Wave energy power generation with non-polluting, renewable, reserves and other advantages, to solve the increasingly serious energy shortage problem is very important. This paper designs a set of ocean buoy power generation system based on wave energy. The permanent magnetic linear reciprocating generator as a wave energy conversion device, the permanent magnetic linear generator used in wave power field, the vertical linear reciprocating movement of the permanent magnet rotor for generator excitation, can be directly related to the wave of ups and downs movement characteristics coincide, simplifies the wave energy conversion process, eliminating the various transmission device and the application of Matlab to closed-loop control of Buck-Boost converter circuit and wave energy generation system simulation. The theoretical analyses and the simulation results could be used for engineering applications of the conversion system of ocean buoys.

In this paper, we formulate the distributed optimization of multi-building energy systems as a cost minimization problem with spatially and temporally coupled constraints. The problem is divided into subproblems of each building at each time slot via Lagrangian dual decomposition. The strong duality of the primal and dual problem is proved, and distributed algorithms are proposed based on the subgradient projection method to converge to the global optimum. Specifically, the local temperature control signals are designed for handling the temporally coupled constraints of each building, while an information sharing platform is designed to update and broadcast the congestion price to autonomous buildings to handle the spatially coupled constraints. Numerical results are provided to illustrate the convergence and effectiveness of the proposed method.

Under the real-time electricity pricing environment in smart grid, building owners are faced with the indoor temperature control problem to minimize the daily electricity cost. Taking the heating scenario as an example, an intuitive strategy is to maintain the building's indoor temperature always at the lower bound of the predetermined comfort range. However, this strategy may not always achieve the lowest electricity bill, especially with the significant fluctuation of electricity prices. On the other hand, the cost minimization problem can be optimally solved one day ahead in a temporally- coupled manner, but the challenge lies in that the building owner needs to acquire the accurate information of electricity prices and outdoor temperatures of the next day, which may not be available. In this paper, we equivalently decouple the cost minimization problem into subproblems at each hour. Each subproblem can be temporally decoupled and optimally solved, only requiring the next-hour electricity price. Besides, the temporally-decoupled algorithm explicitly indicates when to take advantage of pre-heating/cooling for electricity cost reduction. It is demonstrated with the real data that our proposed algorithm could result in considerable economic savings compared with the intuitive strategy, paving the way towards practically applicable cost-effective smart buildings.

Topologies with strong invulnerability and high reliability has become an important issue in the development of Wireless Sensor Networks (WSNs). In this paper, WSNs is modeled as a weighted network which the load of nodes is considered as the weight of nodes. Thus, the problem of balancing the load turns into the problem of balancing the weight of the network. A load-aware and energy-aware local world (LAEAL) model which considers the node fitness, node energy and node weight for WSNs topology evolution is proposed. Both analysis and simulations indicate that LAEAL model enhances connectivity and network lifetime, achieves high robustness against malicious attacks, and at the same time maintains the invulnerability of random attacks.

Plant-wide disturbance such as oscillations are common in large-scale complex controlled processes whose effects propagate to many units and may deteriorate overall control performance. It is important to capture the major causal relationship within the plant and diagnose the root cause along with complete propagation paths. This paper presents a novel multilevel Granger causality framework for root cause diagnosis. The high level is regarded as a group-wise analysis, which is clustered by dynamic time warping-based K-means method and investigated using group Granger causality. The low level is individual causal reasoning within each group where a partial least squares modified Granger causal test is developed to overcome multicollinearity issue. The proposed causality analysis framework is validated through a benchmark industrial case study to show its effectiveness and superiority.

Underwater target recognition is very difficult because of the complexity of the ocean environment. The existing recognition technologies are mostly based on time domain, frequency domain, and time-frequency domain. Beam forming technology is a method of target recognition that has been researched and has developed rapidly in recent years. This paper applies beam forming and spatial spectrum analysis technology to an underwater echo signal in order to capture target echo and highlight spot and dimension features in the time-space domain. Image processing is used to extract the dimension and experimental data also verifies the feasibility of this method of beam forming.

In this paper, we study the sensor selection problem for parameter estimation in wireless sensor networks (WSNs) to minimize the energy consumption with guaranteed quality of estimation. We adopt the Best Linear Unbiased Estimator (BLUE) as the fusion algorithm, and the Mean Squared Error (MSE) criterion to measure the quality of estimation. We formulate the problem as a constrained optimization problem, and prove that it is NP-Complete. A greedy algorithm and a heuristic algorithm are proposed based on the variances of the noises and the energy-accuracy ratio (EAR) of the sensors respectively. Extensive simulation results show that the heuristic algorithm can achieve superior performance than the greedy algorithm in both the energy consumption and the quality of estimation.

Pulsed power researches for military and civil applications have been conducted at China Academy of Engineering Physics (CAEP) for more than fifty years. The pulsed power research activities include development of pulsed power components such as different kinds of high voltage switches, series of pulsed power sources and pulsed X-ray machines, high current accelerators for Z-pinch and flash X-ray radiography as well as medical application, electro-magnetic launch and so on. The most recent progress of pulsed power researches at CAEP will be presented.

Sensor scheduling is essential to collaborative target tracking in wireless sensor networks (WSNs). Most existing sensor scheduling schemes are implemented via the single-step mechanism. To improve the overall performance on energy efficiency and tracking accuracy, in this paper, we propose a novel multi-step sensor scheduling approach using the branch-and-bound algorithm. The approach adopts the Extended Kalman Filter (EKF) to predict the tracking accuracy over a multi-step horizon, and an energy consumption model to calculate the corresponding energy cost. Simulation results demonstrate that, compared to the single-step sensor scheduling approach, the proposed approach can decrease the energy consumption and increase the tracking accuracy significantly.

In real-time imaging applications, the auto-focus algorithm and hardware structure must be adapted to the limited resources for fast and accurate real-time auto-focusing. In this paper, to obtain more precision, a new clarity evaluation function base on image's edge energy is proposed. Then, we present the corresponding FPGA hardware implementation which has simple structure. Additionally, utilizing the prior knowledge during the climbing process, a new climbing algorithm for the auto-focus which has the random start point is proposed. The experiment results show the proposed auto-focus algorithm can reduce the steps during the climbing and accelerate the auto-focus speed simultaneously.

Wireless based data acquisition solutions are widely used in many fields like structure monitoring, transportation or environmental studies. This paper presents a General Packet Radio Service (GPRS) based remote data acquisition and forecasting system, which monitors the geological disasters along the natural gas pipeline under the ground. The field data terminal units (DTU) collect data from the sensors installed on the pipe, and transmit the data by GPRS module to the server in the control center. The server processes the data, broadcasts on the website, and forecasts abnormal variation in time. Three years' operation in site shows the system is effective and feasible.

Flame photometion analysis technics possess many advantages. In past several decades, flame photometric analysis technics have draw much attention to the inspectoion workers, and has obtained considerable development, particularly on its fusion technique. The main elements of toxicants are the materials which contain P, S, Cl, which has a great meaning for detect them. The textual research expresses. The multi-sensor technology can achieve the detect of harmful materials, especially those contains P, S, Cl.