Abstract:Sparse grid is one of Non-uniform grid forms with special properties of hierarchical interpolation. These is an algorithm of discrete Fourier transform on sparse grid, which is called Hyperbolic Cross FFT. This algorithm effectively reduce the number of sampling points, andminimize exponential time complexity of d-dimensions DFT algorithms to O(Nlog^dN). Hexagonal lattice is another kind of grid which has some special propertys such as number of sampling points less and higher sampling efficiency etc. This research focus on combining hexagonal lattice with sparse grid to design FFT algorithm on hexagonal sparse grid. We successfully achieve FFT algorithm on hexagonal sparse grid by defining ahexagon and square lattice conversion between subccript. And the new algorithm is proved to be effective by numerical experiments.

Abstract:With the rapid development and popularization of social network site, how to achieve efficient friend recommendation has become a hot issue. Currently, Matrix Factorization algorithm is widely used method by industry. Although the traditional Matrix Factorization algorithm could bring a good results, but there are still some problems. First, this model does not take full advantage of structural relationship between users in social network; Secondly, this algorithm is dependent on the user-rating matrix, which only has secondary scoring and cannot fully express the user's preferences. In order to solve these two problems, a Matrix Factorization model with social network regularization was proposed in this paper, modeling use of social network users in the model the relationship between neighbors. And as an auxiliary information fusion to the matrix Decomposition Model. This?model?can?solve?the?problems?that?traditional?Matrix Factorization model cannot?solve. Though the contrast experiments on tencent weibo data set, verify that our proposed method could obtain a higher mean average precision than other traditional methods.

Abstract:For solving a kind of multi-agent collaboration task planning problems with limited communication environment, this paper proposes a multi-agent online planning method based on MAXQ-OP. And in the Robocup soccer simulation 2D Wall Station and many players pass the ball to the algorithm in question did an experiment. The experiment shows intelligent agent in this method than the traditional algorithm has increased significantly in the scenario which cooperation is needed.

Abstract:It is unrealistic to configure the different parameters and test analysis in a real cloud computing data centers. Because of the multi-dimensional factors such as the application service are effected from the complexity components, configuration, and deployment; it is most important to use Simulation tools to effective for resources and task scheduling strategy simulation experiment. Based on these, we introduce three major cloud computing simulation platform in this paper:CloudSim, CloudAnalyst, CloudReports, analysis and comparion on the characteristics of its own, separately proceed the three kinds of simulation experiment of simulation platform by the usual task scheduling strategy, evaluates and analyzes its performance. The existing defects are improved and optimized.

Abstract:According to the specific characteristic of the micro-blog user's interests changing with time, a potential interest discovery method based on fuzzy association rules(PIDFAR) is proposed. It could express micro-blog theme distribution by using LDA topic model, and then use the time weighted method to calculate user's recent interest subject distribution, for mining fuzzy associate rules. That set of association rules to express and discover the general rule of user's interests change over time. Finally, after calculating the similarity between the the associated rule of association rule in the collection and the topic distribution of user's interest, take the high similarity of the latter collection in association rules set to constitute the user's potential interests. Experiments show that the method of PIDFAR can make the process of discover the user's potential interest break away from the limit of user's friends group, and improve the accuracy of the discovery of potential interests of micro-blog users obviously than the traditional method of discovery potential interest based on collaborative filtering technology.

Abstract:Aiming at the deficiency of identify semantic in traditional keyword tosearch characters the way the picture when you use the query statement to query the relationship between people. The paper designs and achieves an image search architecture based on character relationships. We offer an approach to extract the latent relationships between persons from news corpus. Domain ontology library creates pictures of people and in the ontology contains a total of 174 relatives, friends, colleagues and other kinds of character relationship with the hierarchy properties. An ontology-driven query interface for the sentence which provides a query-oriented relationship of one sentence, first We extract the critical component of the query statement for the relative persons which comsolidation rules based on dependency grammar tree from the dependency syntax tree, then generate the SPARQL sentence with the help of triple supplement algorithm, and use SPARQL sentence to query the image ontology library. The semantic retrieval was realized. Finally, the experiment results were given to verify the feasibility and effectiveness.

Abstract:Pub/Sub systems are the middleware which provides the distributed event detection, they can support a variety of application development. Considering application of the quality of services provided often depend on the guarantee capability of quality of service(QoS) in the Pub/Sub systems, the paper builds a QoS aware Pub/Sub system named Phoenix, which improves the QoS from the architecture, the reliability guarantee mechanism and the timeliness guarantee mechanism. As a result, Phoenix can successfully deal with link failure, broker failure and cluster churn to guarantee that the applications can run normally. Moreover, it manages to satisfy the timeliness specified by the users with its best effort. The experimental results show that Phoenix is fairly reliable and supports the timeliness requirements of users well.

Abstract:Along with the rapid development of Beijing, focused on the present situation of the discontinuity and deficiency of urban culture and history, this paper tries to construct an urban memory resources platform represented by "Beijing Urban Memory" project. Through the design of "1+1+N" overall platform framework, introducing the layered construction of digital resources library, "Beijing Urban Memory" system designed several subsystems such as resource management subsystem, resources visualization subsystem, professional retrieval subsystem, system uses a LAOP architecture. In addition, this paper designed the corresponding network and server scenari and developed system platform for both general and professional users. At present, the platform has been developed and online, which is a beneficial practice of Beijing culture "going out" strategy.

Abstract:According to the need of vehicle roadway test, a fuel consumption measuring and velocity testing intelligent system was designed and realized. This system used the ethernet port of the computer to create virtual serial ports and chosen RS-485 Means of communication in order to improve the efficiency of serial ports. Besides, for the sake of achieving the real-time monitoring of automobile comprehensive performance tests, this paper programmed on serial ports based on VC++6.0, and designed the software operation interfacefor receiving and displaying the data real-timely, which were measured by the lower computer. The experimental results of the real vehicle showed that fuel consumption measuring and velocity testing intelligent system had the advanced and reliable measurement method as well as the high accuracyon the premise of considering the cost.

Abstract:For the traditional two factor authentication technology, the paper analyzes its deficiency and puts forward using personal identity file as the second factor in the two factor authentication. Therefore, we designed the unique system which identification at server and authentication at client and realized this system by J2EE technology. Experiments show that compared with other two factor authentication technology, it has the characteristics of easy use, low cost, good performance, protect the authentication security effectively.

Abstract:With the rapid development of Internet technology and computer network technology, Android platform system has been widely applied to intelligent mobile terminal to make remote multi video chat anytime and anywhere possible thus improving work efficiency. Such multi video chat system based on Android platform has been concerned by many companies and research institutions. In our system Websocket is used as a communication signal, and WebRTC technology is used to transmitting video and audio in multi video chat system, which includes the server and the Android client. Remote multi video chat is realized by Android client within the same room. This paper focuses on the use of Websocket signaling and WebRTC technology to achieve multi video communication.

Abstract:In a video conference system, business management system is a core component of the entire business environment. Based on the demand of the complexity of the design of the system structure, operation mechanism and the business logic, this paper designs a new architecture of business management system. Analysis of the overall structure of this system and study of the design of business module in detail are also given. On the basis of analyzing business management system of video conference system, this paper puts forward the business management system as a system model based on the layered architecture and this model reduces the complexity of the system. Simulation results show that this scheme effectively improves the efficiency of business processing of business system.

Abstract:Android, an operating system based on Linux kernel, has been widely applied in mobile and other devices, while its development on the function of wire telephones is still under exploration. This research goes through a comprehensive analysis and application of software and hardware, including Calling Identity Delivery(CID) circuits, Android system kernel driver, JNI invoking and message driving mechanism of Looper and Handler. Eventually it realizes a system to display caller information in the form of an Android APP by combining Android and CID. Tests have been conducted in the real system, and the experimental results prove Android's ability of receiving and processing messages from the wired telephone which contributes to its CID function. In summary, it enriches the function and application of Android.

Abstract:Enterprise information system plays a crucial role in business management. But in the development process of enterprise information systems, there are still many issues such as low efficiency, inconsistent system architecture, long development cycle and so on. A code generation tool based on J2EE architecture was designed to solve these problems. With the tool, system developers only need to design models in the visual designer and the tool will complete the code generation. At present, the code generation tool has been applied to the actual software development platform. Practice shows that the use of the code generation tool can unify systems architecture, reduce the development of repetitive code, reuse and share the design, improve the quality and efficiency of enterprise information system development.

Abstract:English preposition error correction system is to help English language learners to correct automatically the common mistakes of English prepositions. First, to classsify and count up the preposition errors in the marked corpus files, sum up 21 kinds of common prepositions, with English wiki corpus and use of computer algorithms automatically error interpolation algorithm to get the training set, and then based on the training set, by using a classification based on the maximum entropy model chosen, including context, prepositions complement other features, training model on the training set, and then use the model to predict the input sentence and correct use of the presence of errors. In NUCLE corpus experiment, given corpus processing, model features, size, number of iterations to test the effect of the impact of training data set, and compare the results of the Naive Bayes model, and finally to the F value 27.68 in the test data with respect to the shared task CoNLL2013 best results have slightly improved.

Abstract:The direct application of network coding is vulnerable to pollution attack. In order to solve this security problem, we propose a network coding scheme to prevent it from pollution based on CRC check code. First, this protocol employs a fast parallel CRC check code and message timestamp to effectively resist pollution attack and replay attack. Secondly, by combining with the homomorphic properties of RSA signature algorithm, we ensure the security of CRC and timestamp. Security analysis and efficiency analysis show that applying network coding, parallel CRC check code and message timestamp together can greatly reduce the computational cost of the node and improve the network throughput.

Abstract:To improve the efficiency of business process and the objectivity of process modeling, we analyse four basic structures of structural workflow net under the premise of the existence of noise in the log file, furthermore, mining algorithm of a structured process model from the log file was researched. This paper expatiates in four parts, includeinggetting information from log file, extracting all simple paths, mining structured branch model and merging final model, finally it shows the model in the form of Petri nets. On the basis of that, the analysis of resource is proposed, it offers the supporting decision tools for decision-makers in resource planning and improves the efficiency of resource utilizing, eventually it increases the efficiency of enterprises.

Abstract:Resource scheduling in cloud computing has always been the focus of research, thus the cuckoo algorithm is introduced in this paper to solve the problem of resource allocation. The resource model of cloud computing is described at first. Then, aiming at the problem that this algorithm is easy to cause local optimal value with fast local convergence speed, this model is improved from the following three aspects. Firstly, the variable-length factor is adopted to make the adjustment and reduce difference between the quality of solutions. Secondly, differential mutation strategy is used to update the bird nest's location. Thirdly, chaotic global search and local search are used based on Coelho to avoid the random disturbance of Levy. It is shown through the test functions that the algorithm proposed in this paper has superior performance than the basic cuckoo algorithm, and the Cloudsim simulation platform shows that algorithm in this paper has obvious advantages in the consumption of time, costs and users' satisfaction.

Abstract:With the development of petroleum industry and the improvement of oil recovery technology, the fault diagnosis of pumping unit is particularly important in the production process. The deficiencies in diagnosis of traditional indicator diagram towards pumping unit are analyzed in the paper. The main insufficiencies are focused on the diagnosis and analytic methods which are the qualitative analysis, and the diagnosis dimension is too limited. The faults of pumping unit are summarized in this paper. Meanwhile, the fuzzy neural network is introducedand the reasoning process is achieved by introducing a step-changing BP algorithm based on gold-segmentation is given. From the different dimensions to solve the problems of pumping unit fault it creates a diagnosis scheme referring to the characteristics of the fuzzy neural network, and the indicator diagram to comprehensively evaluate the faults. Finally, the theoretical feasibility is verified through the simulation experiments.

Abstract:This paper presented a multi-OD dynamic user equilibrium model based on the road network, in order to enable users to minimize travel time in the network. It uses Ant Colony algorithm to solve the problem of dynamic user equilibrium. By designing a numerical example and employing simulation we obtain traffic flow distribution data which compared traffic flow distribution data of Frank-Wolfe algorithm in the road network. The conclusion is that Ant Colony algorithm has certainly advantage in solving the problem of dynamic traffic user equilibrium.

Abstract:In this paper, a face retrieval method based on deep neural networks with relevance feedback has been presented to solve the problem of large-scale human face retrieval. Firstly, convolutional neural networks has been used for feature extracting, then, traditional search methods will be used in face retrieval. A feedback model is added after the retrieval stage. According to the users' feedback, result samples are divided into positives and negatives, which are be used for training the feedback model. Experiments show that this method can significantly improve the accuracy of face retrieval.

Abstract:This paper analyses the main factors affecting aircraft performance degradation in QAR data, uses the optimal weighted data fusion algorithm and the extended Calman filter algorithm for the estimationof state parameter which are related to aircraft performance parameters. When there is less external environment effect, the convenience of the optimal weighted fusion algorithm is verified, and when the effect is overlarge, the precision of this algorithm is limited. Thus we use the extended Calman filter algorithm, take Gauss noise into account, and improve the accuracy of state estimate for the airline to improve the operation of the aircraft provides a reference.

Abstract:In CPS related embedded system design area, time predictability is becoming increasingly important. The system of time predictability can be divided into two level:the programming level based on the traditional RTOS or platform and the architecture level extended with time behavior. In this paper, based on the research between the predictable architecture and the predictable programming model, and a new programming model based on both PRET(PRecised Timed machine) was proposed. Finally an implementation based on this new model was exhibited, through this exhibition we can find that this new model had advantages to the Giotto model.

Abstract:In view of the present parametric modeling method of parts based on UG secondary development, the paper presents two more simple and rapid modeling. The first one takes the parametric modeling of inner gear as an instance, by first establishing a template model part, then in its dialog callback function, code change parameters, last updated parameterized parts. The second one takes ultrasonic transformer as an example, using JA file recording function in UG modeling environment to record the modeling process. In the callback procedure of the dialog box, the recorded code is used to replace the part's modeling program code which originally developed by API or grip. It can realize parametric part after the input values of the dialog box are associated with the parameters of the model in the code. The application of the two methods can effectively reduce duplication designers to improve design efficiency.

Abstract:The dimensions are higher after combining MFCC with LPCC, which affect the performance of recognition. To solve the problem, genetic algorithm parameters for initial feature dimension reduction, to improve recognition performance. Firstly, we extract MFCC and LPCC of the speech signal. Then we reduce the dimensionality based on genetic algorithm. Finally, recognition is presented based on SVM with the low dimensional data. Simulation results show that compared to traditional PCA, the genetic algorithm recognition can increase the recognition rate by 12.2%. Meanwhile, recognition rate was reduced by 1.23% compared with the initial feature,but the recognition time increased 4.5 times.

Abstract:OpenMP regulates a series of compilation guidance, environment variables and runtime routines, having the advantages of simple operation, portability and supporting incremental parallel. But at the same time, the use of frequent thread management overhead FORK-JOIN model is one of the bottlenecks caused by OpenMP program performance constraints.. This article discusses how to use the merge and extend parallel zone, reconstruct the parallel section, and on this basis, using global process analysis ability which is provided by IPA, one of the Open64 optimization components, realize the parallel block mergeing which acrosses process boundaries. The final experimental results show that the method improves the performance of OpenMP programs effectively.

Abstract:Currently one of core issues of link prediction is how to rationally and efficiently combine link attributes, node attributes and other relevant information for forecasting purposes. Aim at this problem, we propose a link prediction algorithm based on influence and interest, which mainly consists of quantizing the influence of nodes by topology structure information and simulating users' interests by text information. These two types of information are aggregated to give the relation score to the node pairs. High-scoring pair of nodes which represents a strong link. Experiments on real datasets show the method proposed in this paper is feasible.

Abstract:Range-free localization technology in wireless sensor networks is of great current interest for researchers. In order to improve the accuracy of range estimation between the sensors and anchors, we introduced a dynamic reliable anchor-based location discovery algorithm in anisotropic wireless sensor networks with holes, which was range-free. Based on the differences of average distances of one hop along the shortest path between different anchor pairs, sensor nodes managed to get two sets of direct reliable reference anchors and indirect reliable reference anchors. Then sensor nodes were localized by using the reliable reference anchors set. Simulation results show that the positioning error is decreased compared with previous algorithms.

Abstract:In order to add the kinds of the sort algorithms which have the complexity of O(n²) and then serve teaching and daily application better, the paper proposes a new sort algorithm-bidirectional selection sort algorithm. Through mathematical method that the time complexity of the new algorithm is O(n²) and its space complexity is O(1). It is known by empirical method that the running time of the new algorithm is averagely 27 percent of bubble sort, and 62 percent of simple selection sort and 88 percent direct insertion sort in same condition.

Abstract:The detection tools can judge whether the virtual machine is under attack or not through detecting the hidden files. The traditional file detection tools reside in the monitored virtual machine, which are vulnerable to attack by the malicious software. According to the virtual machine introspection, a modularized virtual machine file detection method(FDM) is designed and implemented. With the operating system kernel knowledge, FDM can parse the physical hardware and build the semantic view of the files. Then FDM can identify the hidden files by comparing with the internal file list. Meanwhile, parsing hardware status and obtaining semantic information are implemented in different modules. FDM has not only the tamper resistance of the virtual machine introspection, also has a modular architecture, portability and efficiency The experimental results show that the FDM can quickly and accurately detect the hidden filesinside virtual machine.

Abstract:Power system intelligent alarm is mainly concentrated on the rule reasoning domain. This method does not effectively use the analysis of historical failure to analyze and handle the accident which is currently happening in the system. In this paper, after a preliminary study on fault diagnosis algorithm with case-based reasoning, a matching algorithm is proposed based on vector calculation. In this algorithm, the fault information in the system is abstracted and the similarity to the cases within system case library is then calculated to try to match the fault information to the case information one by one. All the history cases are obtained by manual analysis of historical failure information in the system case library. The fault properties and the cause of similar cases are helpful for current failure analysis, diagnosis and exception handling.

Abstract:For time-driven data collection methods, a new optimal cluster head selection(OCHS) algorithm is proposed based on the principle of fuzzy logic, considering different data collection period and different initial energy for different nodes in heterogeneous Wireless Sensor Networks. In the choice of cluster head node, OCHS algorithm uses a competitive approach to select the optimal cluster heads considering the factor that nodes have different data collection period, different initial energy and distribution density. By adopting the method of the Pythagorean theorem, the member node selects a path with minimal energy consumption to join the cluster. Simulations were conducted using MATLAB software and the results show that OCHS algorithm can effectively balance the network nodes energy consumption, prolong the lifetime of the sensor network and improve the stability of the entire network.

Abstract:In software testing, the key to a successful test is a fast and efficient testing-case generation. Genetic algorithm is an algorithm to search for the optimal solution by simulating the natural process of evolution. The algorithm guides the of search direction through the selection, crossover and mutation operations. to reach the global optimal solution step by step. Traditional genetic algorithm is widely used in the test case generation by many scientific researchers due to its better global search ability. But the genetic algorithm can easily lead to convergence to a local optimal solution because of its inherent defects "premature convergence". In order to solve this problem, the author proposed an adaptive genetic algorithm in this paper. The crossover operator and mutation operator of the proposed algoritym can be adjusted automatically according to the change of the program. The improved algorithm is then applied in the test case generation process. The test results show that this algorithm is better than the traditional search algorithm and common improved algorithm in efficiency and effectiveness of testing-case generation.

Abstract:To meet the collaborative management requirement of multi-core network enterprise group on SaaS-based collaborative platform of automobile industry value chain with automobile parts industry value chain, a multi-tenants and multi-level authorization model based on the core of a leading enterprise is proposed on the basis of the traditional RBAC model. The scheme effectively controls users' access rights.and achieves analysis of authority by using menu dynamic generating technology and URL address resolution algorithm. The plan and the implementation are tested on the SaaS-based collaborative platform of industrial-chain. The results show that it can solve the problem of authority management of multi-tenants for SaaS-based Collaborative Platform of industrial-chain.

Abstract:In order to solve the overlap between frames and improve the self-adaptability of the speech signal, an improved speech recognition system based on hidden Markov model(HMM) and genetic algorithm neural network is proposed in this paper. The major improvement is the adoption of wavelet neural networks in the training of Mel frequency cepstral coefficients(MFCC). And by using HMM models time series of speech signal, the speech's score on the output probability of HMM is calculated. The results will be used as the input of genetic neural network, the information of the speech recognition and classification can then be obtained. The experimental results show that, the improved system has better noise robustness than the pure HMM and the performance of the speech recognition system is also improved

Abstract:The problem of a traditional experimental instrument for the course of computer principles is that it can only be used for the functions of realization and vertification and cannot meet the requirement for students' independent design and innovation. A method for improving the experimental board of computer organization is proposed. The method uses a complex programmable logic device(CPLD) chip to replace the traditional discrete small-scale IC chip as the carrier. By using electronic design automation(EDA) tools and very-high-speed integrated circuit hardware description language(VHDL) it can achieve a variety of hardware logic, and connection with the input and output hardware interface circuits. The final composition with a complete computer system can be used for experimental teaching. Experiments show that the new experimental board can achieve the experimental requirements of original experimental instrument, enhance the function of the experimental system, improve flexibility and reduce test costs.

Abstract:In this paper, wavelet transform and clustering method are combined in order to detect the task stimulation-caused activation areas of a human brain in fMRI. According to the idea of multi-step decision, firstly, we perform fuzzy-c-means clustering in the fMRI image to solve the ill-balanced problem of the data. Secondly, we use the stationary wavelet transform to decompose the data, and extract those data of interested frequency, and then the improved k-means clustering algorithm is proposed to analyze these data in the wavelet domain. The experimental results with several subjects show that the visual stimulation-caused activation areas of human brain can be detected. Compared with the popular SPM method, the proposed method in this paper is more reasonable, and has directive significance and practical value on the functional connectivity detection of human brains.

Abstract:This paper analyzed the banknote bundling problems in current bank system and designed an ARM-based automatic banknote packaging machine, which integrates with bundling and plastic enveloping. In this paper, the composition and structure of banknote packaging machine, as well as its working principles, software and hardware implementationare described in detail. The actual operating results show that it can accomplish all the motion control in the banknote packaging processes, including sending notes, bundling notes, cutting and sealing plastic film. The new banknote packaging machine has improved the efficiency of banknote packaging.

Abstract:The dynamic response of continuous box girder bridge when vehicle crossing was analyzed with a finite element calculation program. But the pre-post processing interface and graphic function are weak. In view of this, a mixed programming method was proposed by Fortran language with powerful computing function combined with Visual Basic language which has the good functions of visualization design. A finite element system for the calculation of bridge dynamic responseis worked out. This software not only can solve the complex engineering problems, but also has a good interactive interface. Some knotty problems encountered in the development process about mixed programming are described, and some corresponding solutions are given.

Abstract:In ray tracking, acceleration structure is indispensable for reducing computation time of ray-geometry intersection. Finding an efficient and high quality acceleration structure is still an active area of research. The top-down approach is widely used in building acceleration structure, such as BVH. However, due to its limitation, the result tree is not always optimal. This paper presents a bottom-up approach to build an effective BVH which is a hybrid between BVH and Space Partitioning. An improved surface area heuristic cost function and multiple parameters have also been used to optimize this model. Experiments show that the approach mentioned above is more effective than the conventional BVH and KD-Trees. It can realize 5%-10% performance improvement. Furthermore, the way to reduce the building time has been investigated in this paper.

Abstract:Considering the low contrast and cell adhesion of cell image, this paper proposes a method to combine the pattern recognition ability of both computer and the human to complete cell image segmentation. It takes human-computer interaction to obtain dynamic threshold and shows the result with different colors. Experimental results show that it has the advantages of easy operation, good segmentation, with better usage value compared with the Otsu method, Kapur method and Niblack method.