Advanced Materials Research Vols. 734-737

Abstract: Microkernel-based virtualization is one of several popular virtualization technologies aiming for embedded devices and efficient communication between virtual machines should be provided as other virtualization solutions. Long data communication is one of main two communication methods. This paper proposes EMLDC: an efficient method of long data communication based on shared memory in the L4 Microkernel-based virtualization environments. Especially EMLDC adopting lock-free mechanism can reduce performance loss and priority inversion in the real-time virtual machines against traditional blocking algorithm. Benchmark programs have been conducted to demonstrate that EMLDC can support long data communication efficiently.

Abstract: The graph vertices design into classes, for each vertex in the design of the abscissa, ordinate and in-degree members, realizes the dynamic demonstration graph traversal algorithm. Around two window synchronization of animation, " traversal sequence " list box list dfs (bfs ) traversal sequence ,with thick line in the left window drawing traversed through the vertices and edges, and in "traversed stack (queue) changes in" real-time text box displays the current traversed through vertices, in the right box demo the process of algorithm dynamic execution.

Abstract: The tracking of ground targets using aerial images was studied. A improved Kalman filter was derived for the tracking of ground targets. The novel feature of this improved filter were that the grey prediction equations and the road information have been incorporated to improve the accuracy of state estimates.The GM(1,1) (Grey Model) was introduced into Kalman prediction equations.The next value was forecasted by using few forward estimated values with a grey differential equation,which was baesd on the correction of mesurement covariance matrix. The tracking of the targets shows a satisfactory result.

Abstract: Study on a new type of fingerprint identification algorithm and its application in intelligent home electric control lock problem. The traditional fingerprint recognition algorithms on fingerprint minutiae matching accuracy is low, difficult to accurately extract details, leading to lock malfunction or could not be opened. In order to overcome this problem, improved Point pattern fingerprint recognition matching algorithm, joined the matching weight coefficient on the base of pattern matching algorithm, and gives the hardware structure of fingerprint identification system, the improved algorithm is successfully applied in smart home applications, the example shows that, the improved algorithm can effectively improve the recognition rate , reduce false positives, has a certain practical value.

Abstract: In the past few years many scientists have tried to develop models of networks and to investigate the mechanisms that determine the topology of complex networks. We define a class of graphs called (k,m)-dragon graphs and uniformly (k,m)-dragon graphs as referenced models of complex networks. These dragon graphs have some properties such as its admits (k,d)-odd-graceful labeling, graceful labellings, odd-graceful labellings and so on. In this paper, we proposed and defined the notion of some labelings of dragon graphs, the odd-graceful, graceful and total labellings of edges magic of these dragon graphs have been investigated in detail, and our methods can be algorithmic. Through the definition of dragon graphs and proofs of Theorems, we get some simple but usefulconclusions.

Abstract: According to the basic theories of Logit regression analysis and support vector machine, this article involves improved multi-classification combination algorithm. When applying this model, there are some innovations. First, choose optimized composite indicator as a variable through principal component analysis and get more information. Second, introduce Logit parameter model to the quadratic to increase prediction accuracy. Third, put forward a multi-classification combination model of improved Logit model with SVM to increase prediction accuracy.

Abstract: In granular materials processing, how to efficiently recognize and remove foreign bodies is very important. Training data including H,I mean and I standard deviation of target material are distinguished to be foreign bodies or not by training model, which is established using FWOCSVM method, while taking into account the characteristics of foreign body detection. A way to introduce the weight value reflecting the importance degree of attribute with its mean square deviation is developed to solve the problem, that is, attribute weights are not considered in OCSVM. The results show that FWOCSVM has more excellent performance than that of threshold or OCSVM .And color features adopted in the paper have excellent identification performance in above foreign bodies detection.

Abstract: With the development of information technology, the development of wisdom travel to represent the general trend, which is not only conducive to the development of tourism enterprises, but also conducive to the development of the tourism industry. Jiangxi wisdom tourism e-commerce platform construction is a systematic work, the relevant units should attach great importance to this work, and actively carry out cooperation, coordination, project group should be in line with the service of tourism enterprises, scenic area, the principle of mutual benefit, do research, development and function of the system extension work.

Abstract: Session replication is the common user demand of web application. Good session management program can enhance the user experience. At the same time it doesnt increase the burden of system operation. The paper mainly analyses on the five commonly used session management program. And it compares the topological structure, advantages and disadvantages and the difficulty level of realization of five session management program.

Abstract: Due to disadvantages of nonlinear, complexity and uncertainty of fermentation process, a research on cell concentration prediction of erythromycin fermentation process was carried out. Combining the optimization ability of ant colony algorithm and the regression ability of support vector machine, an ACO-SVM model is built. Case study shows that, the model is more accurate and more effective for the cell concentration prediction than ANN and SVM model.