Applied Mechanics and Materials Vols. 380-384

Abstract: How to find most relative factor in multifactor system and how to find the function between relative factors and result are very important. In this paper, Bayes is used to find the most relative factors, then the training dataset is clustered by AGNES(Agglomerative Nesting) , then the section polynomial fitting is used to formulate multiple function for those most relative factors. Finally, this method is applied in the eye sight analysis system, and satisfying results can be obtained.

Abstract: It is the problem how to search the main factors in various factors on accident. The gray correlation can not only improve the efficiency of the data which have existed, but also remedy the limitation of that carrying out systems analysis by mathematical statistics. From the overall perspective of human-machine-environment, accident prediction model is established and the influencing factors are analyzed of accidents in this paper. The grey correlation degree of the influencing factors is calculated. At last, prediction model of examples is examined. The result shows that the model is applicable and reliable in forecasting the main factors and the relations between them, thus providing reference for traffic administrative department to avoid traffic accidents.

Abstract: Battlefield situation assessment (BSA) has a positive significance on improving the efficiency of commanding decision-making, moreover, BSA cannot be made successfully without the support of some integrated and exact intelligence data. In this paper, basing on the demand of identifying the battlefield situation, the corresponding knowledge context database is first discussed; on this basic, construction of the intelligence data warehouses framework is explored. Then, from the view of a holistic conception, the study of data mining based on the intelligence data warehouse is made, and a detailed arithmetic is presented by making use of the tactic from data mining driven fishbone (DMDF).

Abstract: K-medoids clustering algorithm is an efficient algorithm in classifying cluster categories. Based on algorithm analysis, this paper first improves the selection of K center point and then sets up a web model of ontology data set object with the aim of demonstrating through experiment evaluation that the improved algorithm can greatly enhance the accuracy of clustering results under semantic web.

Abstract: Lots of research findings have been made from home and abroad on clustering algorithm in recent years. In view of the traditional partition clustering method K-means algorithm, this paper, after analyzing its advantages and disadvantages, combines it with ontology-based data set to establish a semantic web model. It improves the existing clustering algorithm in various constraint conditions with the aim of demonstrating that the improved algorithm has better efficiency and accuracy under semantic web.

Abstract: There is a shortcoming that particle swarm algorithm is ease fall into local minima. To avoid this drawback, this paper insert into a perception range that from Glowworm swarm optimization. according to domain to determine a perception range, within the scope of perception of all the particles find an extreme value point sequence. All the particles that in the perception scope find a extreme value point sequence, which apply roulette method, in order to choose a particle instead of global extreme value. So as to scattered particle, and avoid the local minima.

Abstract: The traditional training meshods of speaker codebook for speaker identification based on vector quantization are sensitive to the initial codebook parameters, and they often lead to a sub-optimal codebook in practice. To resolve this problem, this paper proposes a novel bi-group particle swarm optimizer (BPSO). It applies two sub-group particles with different particle update parameters simultaneously to explore the best speaker codebook, and the particles perform basic operations of particle swarm optimization (PSO) and conventional LBG algorithm in sequence, which can explore the solution space separately and search the local part in detail together. Information is exchanged when sub-groups are periodically shuffled and reorganized. Experimental results have demonstrated that the performance of BPSO is much better than that of LBG, fuzzy C-means (FCM), fast evolutionary programming (FEP), PSO, the impoved PSO algorithm consistently with higher correct identification rates and convergence rate. The dependence of the final codebook on the selection of the initial codebook is also reduced effectively.

Abstract: A construction window forecast method of large-scale tunnel elements is presented based on the group decision-making algorithm. Based on superiority method and satisfaction value method, multi-attributes group decision making algorithm for force and stability condition screening and window forecast algorithm is also provided, optimal evaluation function decision-making method for selection of construction window is also provided, these methods can take a quick solution for environment limit conditions. The whole construction window forecast method is applied to construction window forecast for the undocking, floating transport and mooring of tunnel elements, the results show that the method can give a good suggestion for engineers.

Abstract: In this paper, we propose a new discrimination method using image statistical characteristics is proposed which is designed to distinguish natural images from photorealistic computer graphics. Using Benford model as statistical basis, we conclude statistical properties of the MSD (most significant digit) of AC (Alternating Current) coefficients in DCT (Discrete Cosine Transform) domain of natural images and computer graphics, and then we constructed the detection model of the proposed algorithm. Experimental results show that this method can identify natural images and computer graphics effectively, compared with the existing algorithms this method has a higher recognition rate, which comes to 95.22%.

Abstract: The traditional cost function, minimization mean square prediction error is a second order statistic, and it is based on the error Gaussian distribution and linear assumption. But chaotic signals are non-Gaussian, so the optimization criterion is not suitable. Then we present using the robust optimization criterion, maximum correntropy to replace the popular minima mean square error criterion minimization error. In simulation, the algorithm shows an improved performance to a common three-order Volterra prediction.