Applied Mechanics and Materials Vols. 513-517

Abstract: This paper presents an overview of the new possible concept for the application of smart mobile devices (smartphones, tablets) in defined environments. Practical work that has been done (control of smart solar house) is presented in two planes. The first plane is observe a defined environment, smart solar house with all the necessary functional requirements, hardware elements and communication modules. The second plane is the Android application that is a 3D interface to communicate with mobile devices and smart components of solar home. The work represents an excellent experimental basis in the teaching of programming and robotics. The proposed solutions are based on the work and research using the latest smart mobile devices, communication modules and software.

Abstract: A new high-precision method to orient the ultrasonic linear array signals was proposed in this paper. First, the FastICA algorithm was used to de-noise the ultrasonic linear array signals; then the TCT algorithm is used to focus the ultrasonic linear array signals to greatest extent to cumulate the position information of broadband signals, the FastDOA is used to estimate the direction-of-arrival (DOA) of focused narrowband signals to speed up the calculation after focusing; Then, the application of partial discharge detection using this method was studied, the estimated accuracy of the method was simulated with different SNR(Signal to Noise Ratio) at first, the results show that the errors of DOA were about 2°;At last, the results from some experiments show that the errors were about 3.5°, which indicates that this estimation method is with a good accuracy.

Abstract: A novel texture classification approach based on neighborhood estimated local binary patterns (NELBP) is proposed. In the proposed approach, the local surrounding values of neighborhood estimated are introduced to operate binary patterns. Moreover, two different and complementary descriptors (average-based descriptor and differences-based descriptor) are extracted from local patches. Contrast experiments on Outex database and CUReT database demonstrate that the proposed NELBP is more robust to Gaussian noise than the conventional LBP for texture classification. In addition, the results also show that the combined complementary descriptor playes an important role in texture classification.

Abstract: With the development of Chinas economy, credit scoring has become important. The general credit scoring model is to solve the two classification problems, but in real life we often encounter multiple classification problems. This paper proposes a multi-class support vector machine based on genetic algorithm, which can solve multiple classification problems in the behavior assessment model.

Abstract: Natural scene classification is a challenging pattern classification problem nowadays. The description of image plays a crucial role in the process of recognition. Many different approaches and feature extraction methodologies concerning scene classification have been proposed and applied in the last few years. This paper proposed a novel method of natural scene recognition based on graph edit distance (GED) in which scene images are represented by attributed graph. The vertex label is the features of regions and edge label is the features of public area of adjacent regions. This method used local representation as well as global way, realized the cooperation of global and local mechanisms. The proposed method approaches satisfactory categorization performances on the well-known scene classification datasets with 8 scene categories.

Abstract: Hydraulic conductivity (HC) is deduced indirectly from soil water retention curves (SWRC) by Mualem model, but the mathematical calculation in the Mualem model is sensitive to integration interval near saturation. After the van Genuchten model fit actual SWRC, the van Genuchten-Mualem (VGM) HC is integrated simply by an analytical function for unsaturated soils. However the analytical solution on VGM HC results in the rapid change of HC near saturation. For unsaturated soils sampled in Korea, SWRCs and unsaturated HCs were obtained by experiments. The HC experiments were compared with the HC models from the SWRCs. As a result, VGM models of HC function underestimate the unsaturated HC and show rapid reduction near saturation. It is found that a modification in VGM model should be required to predict accurate HC functions.

Abstract: In this paper, a vector control strategy with UVM for seven-phase induction motor drives is presented and tested.Based on the spatial decomposition model of 7-phase induction motor, the d-q fundamental spatial components are controlled using vector control strategy. However the harmonic components of currents are generated by the harmonic components in the 3rd and 5th harmonic subspaces.Combined with voltages in harmonic subspaces set to be zero, the UVM with harmonic current elimination is proposed.Based on the simulation, experiment results verify the feasibility and effectiveness of the strategy proposed on the 7-phase IM experimental setup with DSP core TMS320F28335.

Abstract: The mass event happens frequently, and once dealt not well, the event will lose control and the result is unimaginable. This paper, based on the analysis of the theory of the emotional contagion, proposes leveled emotional model based on the importance of leaders. By improving the interaction algorithm of the participants, introducing the factors of followers, emotional decline, emotional un-adaption, moving around, we also build an emotional affection model of the mass event. Finally we explore the general rule of the mass event through simulation in the AnyLogic software, which can provide policy when facing mass event.

Abstract: In a voice localization system, the nonlinear equation of voice sources coordinates were established according the accepted information, so the algorithm for solving nonlinear equations is the key problem to the voice source localization syetem. In this paper, the Newton - Raphson method (N-R) is applied to solve the nonlinear equations, by setting the initial solution of equations,then calculating the unbalance vector and Jacobian matrix (J) so as to attain corrected vetor, after modification and iteration the initial solution,until meet the precision numerical solution. Test results show that, application of N-R method in voice localization system have advantange of less number of iterations, saving chip resources and high precision, can meet the precision requirements of voice localization system.

Abstract: According to the test result about class insulator string of 32 units, a method of calculating 50% ac flashover voltage U₅₀ based on radial basis function network is put forward in this paper. Inputs of the model are equivalent salt deposit density ρ_ESDD and nonsoluble deposit density ρ_NSDD ,the output of the model is 50% ac flashover voltage of polluted insulator strings, the Orthogonal least squares algorithm was used to select right hide layer neurons and center vectors, and the weight from hide layer neurons to output layer neurons are determined by pseud-inverse method. The calculating result is in concordance with the test result, it is indicated that the model can give some references to access the pollution condition of external insulation.