Applied Mechanics and Materials Vols. 530-531

Abstract: During the evaluation of multi-source information system, uncertainty exists widely. In order to use and process the uncertainty and improve the creditability of system evaluation, this paper based on the cloud theory, studied the mapping method between nature variable and quantity variable. At the same time, the paper promoted a novel evaluation method of system performance based on the cloud theory. Finally, an example validates the process and feasibility of the method. The method exploits new application field for the cloud theory.

Abstract: In order to more effectively classify the science and technology intelligence text, the idea that classifying science and technology intelligence text categorization based on different classifiers is proposed. The experiment is done with two thousand Chinese texts based on three different classifiers in this paper. Among these classifiers, the rate of correctly classified instances with NaiveBayes Classifier is 96.95 percent and J48 Classifiers is 97.59. The highest of three classifiers is SMO Classifier and its correct rate is 98.65 percent. According to the analysis of experimental results, it is proved that the idea proposed is applicable to science and technology intelligence text categorization and it is able to meet the needs of text categorization.

Abstract: In order to reduce the time of fuzzy inference, the relevant matrices and the relationship matrices are used to constitute the fuzzy-valued concept networks. The elements of a relevant matrix represent the relevant degrees between concepts. The elements of a relationship matrix represent the relevant relationships between concepts. Fuzzy positive association relationship or fuzzy negative association relationship are used for formulating users queries in order to increase the flexibility of fuzzy information retrieval systems. Expanding the fuzzy-valued concept network architecture to the Internet environment, we propose a fuzzy information retrieval method based on the network-type fuzzy-valued concept network and it can be relatively more effective information retrieval in the distributed network

Abstract: Travelling salesman problem is a fundamental combinatorial optimization model studied in the operations research community, and yet, there is surprisingly little literature that addresses stochastic uncertainties and multiple objectives in it simultaneously. This paper is devoted to a novel TSP variation called stochastic multiobjective TSP (SMOTSP) with random variables on the arc, and a new solution approach is proposed to obtain Pareto efficient route in it, whose validity is proved finally.

Abstract: Using the neural network to deal with complex data, because the pending sample with many variables, aiming at this nature of the pending sample and the structure properties of the BP neural network, in this paper, we propose the new BP neural network algorithm base on principal component analysis (PCA-BP algorithm). The new algorithm through PCA dimension reduction for complex data, got the low-dimensional data as the BP neural networks input, it will be beneficial to design the hidden layer of neural network, save a lot of storage space and computing time, and conductive to the convergence of the neural network. In order to verify the validity of the new algorithm, compared with the traditional BP algorithm, through the case analysis, the result show that the new algorithm improve the efficiency and recognition precise, worthy of further promotion.

Abstract: The success of supervised learning approaches to word sensed disambiguation (WSD) is largely dependent on the representation of the context in which an ambiguous word occurs. In practice, different kernel functions can be designed according to different representations since kernels can be well defined on general types of data, such as vectors, sequences, trees, as well as graphs. In this paper, we present a composite kernel, which is a linear combination of two types of kernels, i.e., bag of words (BOW) kernel and sequence kernel, for WSD. The benefit of kernel combination is that it allows to integrate heterogeneous sources of information in a simple and effective way. Empirical evaluation shows that the composite kernel can consistently improve the performance of WSD.

Abstract: It has been an important subject that how to apply transformer condition information efficiently and scientifically to judge the transformers state and predict the remaining life. This paper, in view of the distribution transformer test information, analyzed the factors affecting its residual life evaluation, and established the evaluation model of distribution transformer. Based on the existing research results and the introduction of improved Markov prediction method, it finally established a dynamic variable metric improved Markov distribution transformer life prediction model. Actual data verified the effectiveness and reasonability of the model.

Abstract: The MUSIC algorithm cannot deal with the problem of DOA estimation of coherent sources, this paper proposes the USTC (unitary spatio-temporal correlation matrices)-MUSIC algorithm using single vector hydrophone to solve this problem, by utilizing the unitary spatio-temporal correlation matrix instead of the covariance matrix. The simulation results demonstrate that the USTC-MUSIC algorithm has a better ability to distinguish the coherent sources from different directions than the spatial smoothing MUSIC algorithm.

Abstract: To solve the problem of sliding-type scattering center and the problem of angle limitation of single radar, a kind of method for extracting three-dimensional micro-motion characteristics of ballistic target based on netted radars was proposed. Firstly, A new feature extraction method based on half period delay multiplication was introduced which establishes the relationship of sliding-type scattering center and ideal scattering center, then making use of the expand hough transformation, the parameters of echo signals were extracted. At last, the three-dimensional micro-motion features and the structure features were obtained by solving nonlinear multivariable equation systems which were established by the multi-view of netted radar. Simulation validated the method can get both the high estimating precision and the three-dimensional micro-motion parameters.

Abstract: The random noise is the kind of noise with wide frequency band in seismic data detected by the optical acceleration sensors. The noises influence and destroy the useful signal of the seismic information. There are a lot of methods to remove noise and one of the standard methods to remove the noise of the signal was the fast Fourier transform (FFT) which was the linear Fourier smoothing. In this paper, the novel denoising method based on wavelet analysis was introduced. The denoising results of seismic data with the noise with FFT method and wavelet analysis method, respectively. SNRs of the signal with noise, FFT denoisng and wavelet analysis denoising are-8.69, -1.13, and 8.27 respectively. The results show that the wavelet analysis method is prior to the traditional denoising method. The resolution of the seismic data improves.