Applied Mechanics and Materials Vols. 380-384

Abstract: In the process of complex cloth modelling, the cloth tissues mobile disorderly, resulting in ineffective modeling and other problems. To address the problems mentioned above, a corresponding solution is put forward. Based on the HLA development platform and Vega development platform, a 3D cloth simulation method which improves the simulation algorithm of annealing is proposed. Some related processings are performed on the cloth image, and the moving cloth issues are simulated using the improved fast cloth moving particle model to complete 3D simulation of the cloth. Experimental results show that this method can describe the complex cloth issues accurately, quickly and completely, and achieves desirable results.

Abstract: An improved data clustering algorithm was proposed based on the Fuzzy C-Means (FCM) algorithm for the purpose of clustering the data precisely and effectively, through progressing the performance of the data clustering to afford the element work for the application of fault diagnosis and target recognition and so on. There was fatal weakness for the traditional FCM algorithm that the algorithm is sensitive to initial value and noise. The chaotic differential evolution FCM algorithm was proposed according to the efficient global search capability of differential evolution algorithm and the traversal characteristic of chaotic time series. The improved algorithm used the Logistics chaotic mapping to search for the optimal solution, and the chaos disturbance was introduced into the evolutionary population to make up for the defects of FCM algorithm. The new method can overcome the problems of initial value sensitiveness with FCM and local convergence with genetic algorithm. Because the new method. Three types of typical vibration data of faults engines was taken as the example for the research and application. The simulation and application result shows that the data clustering performance of the improved FCM algorithm is much better than the traditional FCM algorithm, and the accuracy rates of fault diagnosis in the application was increased by more than twenty percent, it shows good application prospect.

Abstract: The traditional motivation behind feature selection algorithms is to find the best subset of features for a task using one particular learning algorithm. However, it has been often found that no single classifier is entirely satisfactory for a particular task. Therefore, how to further improve the performance of these single systems on the basis of the previous optimal feature subset is a very important issue.We investigate the notion of optimal feature selection and present a practical feature selection approach that is based on an optimal feature subset of a single CAD system, which is referred to as a multilevel optimal feature selection method (MOFS) in this paper. Through MOFS, we select the different optimal feature subsets in order to eliminate features that are redundant or irrelevant and obtain optimal features.

Abstract: [Purpos In order to improve the accuracy of target tracking and reduce losing rate of target in the multiple target tracking, a new algorithm called Extended Probabilistic Data Association (EPDA) is presented in this paper. [Metho This paper defines joint association event based on the number of target and puts forward the EPDA for target tracking. [Result Experimental results show that this algorithm has higher accuracy of target tracking than the Probabilistic Data Association algorithm and costs much less time relative to the Joint Probabilistic Data Association algorithm. [Conclusion Consequently, EPDA is an effective algorithm to balance the accuracy and the losing rate in target tracking.

Abstract: With the rapid development of the Internet, database is getting more and more attention for its advantages of high concurrency, high scalability and high availability. In order to solve the problem that how to retrieving the data in the database efficiently and accurately, a model of fuzzy retrieval based on external index is proposed in this paper. By this external index, the efficiency of the NoSQL databases retrieval whose column not be appointed is greatly improved. As an example, the Cassandra database is adopted to store the data and the external index is stored in the national databases. A community information management system is utilized to show the feasibility of the model. The results show that the model can save a lot of time in retrieval whose column not be appointed. Moreover, this model can also be used to some other NoSQL databases.

Abstract: The paper presents a calculation method to get the probable key points in hands outline. First, use K-vector to calculate the K-slope of all points in outline and get a series of peak point sets. Then, cluster the peak point sets which have been obstained by K-medoids algorithm and get the location of every fingertips and probable points; At last, distinguish the fingertips and webbed points by vector cross product operation. The experiment shows that this method achieve the precise positionong of the fingertips and finger webbed points.

Abstract: This paper proposed a method of computing the hand direction vector. At first, using K-vector method, compute the K-slope of all the points on the contour to get a series of peak point sets; Then, cluster the sets which have been obtained through K-medoids algorithm; At last, detect the pointing direction of each fingertip from the clustering information, and get the hands direction by it. The experiment shows that the method achieves the precise positioning of hand direction in palm contours.

Abstract: In this paper, we compare and analyze the performances of nine unsupervised discretization methods, i.e., equal width, equal frequency, k-means clustering discretization, ordinal, fixed frequency, non-disjoint, proportional, weight proportional, mean value and standard deviation discretizations in the framework of continues entropy estimation based on 15 probability density distributions, i.e., Beta, Cauchy, Central Chi-Squared, Exponential, F, Gamma, Laplace, Logistic, Lognormal, Normal, Rayleigh, Student's-t, Triangular, Uniform, and Weibull distributions.

Abstract: In this paper, we propose an improved alpha-beta search algorithm, named trappy alpha-beta (simply), for game-tree in order to identify and set the potential traps in the game playing. can be regarded as an extension of the traditional alpha-beta search algorithm which ties to predict when the opponent might make a mistake and select such moves that can most likely lead the an opponent into the trap by comparing the various scores returned through iterative deepening technology. In our experiment, we test the performance of in comparison with three game-tree search algorithms, i.e., min-max, trappy minimax, and alpha-beta, by playing with four testing opponents, which are obtained form a typical Chinese chess computer game program-Xqwizard (http://www.xqbase.com). The comparative results show that our designedcan effectively find and set the traps in the playing with opponents.

Abstract: An improved Dijkstras shortest path algorithm based on search strategy is proposed in this paper. In order to solve the defects of the conventional algorithm, such as large redundancy of space and time, the proposed algorithm introduces a constraint function with weighted value ω for searching each position in the state space to guide the search forward to expected direction. Meanwhile, according to different complexity of map information, the weighted value ω can be flexibly changed to make the constraint function more reasonable to effectively improve the search efficiency. Experimental result shows that the number of nodes on search path and computation time is obviously reduced, and the improved algorithm can be fast to search out the goal nodes.