Advanced Materials Research Vols. 850-851

Abstract: In this paper, a predatorprey model with discrete and distributed delays is investigated. The necessary and sufficient of the stable equilibrium point for this model is studied. Further, analyzed the associated characteristic equation. And, it is found that Hopf bifurcation occurs when τ crosses some critical value. Last, an example showed the feasibility of results.

Abstract: The paper uses digital image processing technology, technology of face pattern recognition and traditional database retrieval technology, integrate image retrieval technology based on version and content, and avoid the complexity of matching image process. The experiments with 200 human samples, correctly retrieved for 155 people, exactly matched to 125. The recognition rate of the system is 75.55%, the average time of search is less than 0.1s. Experiments indicate this method has strong robustness. The semantic face image retrieval system using this method has the characteristics of fast, efficient, practical.

Abstract: Handwritten numeral recognition is an important branch in the field of pattern recognition, has broad application prospects. This article presents a method of using BP Neural Network to implement programme for recognition of free handwritten numerals. Scanned handwritten numeral image after preprocessing and feature extraction, classificated and recognized by the BP Neural Network. Through Matlab simulation experiments it shows that the recognition method is effective and has high recognition rate.

Abstract: To improve the performance of LMS algorithm ,a modified algorithm is proposed .decoration is used and orthogonal vectors is applied to replace weigh of the filter ,forgetting factor is also introduced to control the step .A new model is built and experiment show that convergence speed is improved and steady-state error is reduced. The performance is improved.

Abstract: Considering the problem that vibration signals which are generated by three-axis accelerations will impact turbulence signal, an improved motion compensation algorithm based on discriminant analysis is proposed. The characteristics of three axis acceleration signals are got by using cross validation discriminant analysis method to test the performance of de-noising among different filtering orders which are got by three-axis accelerations vibration signals. Finally the discriminant function is found to get the optimal filter order. The results of simulation and real sea experiment data statistical analysis shows that the algorithm improves the de-noising performance and can be better applied to ocean turbulence signal de-noising processing.

Abstract: A new passive target motion analysis (TMA) based on method a single seismic-filed vector sensor is designed using pitching angle and multi-path delay is studied. The state equation and measurement equation are established and analyzed systemically by improvement particle filter based on the measurements of both bearings and time delay that are collected by the sensor. The Monte Carlo simulation experiment results demonstrate that the algorithm has some advantages of fast convergence, high precision and stability.

Abstract: With respect to the multiple attribute decision making problems based on grey correlation, sensitivity analysis is proceeded to attribute values, and analysis model as well as search algorithm are given to compute the attribute variation interval which maintain the original rank. Finally, a numerical example was presented.

Abstract: According to the theory of the acoustic scattering field of underwater targets, the theoretical derivation of C-Matrix method is given, and the mathematical model for solving acoustic scattering field is established. By means of MATLAB, the accurate expression method for the figure of acoustic scattering field is studied, and the characteristic and difference of the simulation curve in various frequency and azimuth are analyzed. It is shown by the application that, to a certain extent, the conclusion has practical value and guide meaning for the detection and identification of the hiden objects.

Abstract: This paper investigates topology identification and adaptive synchronization of uncertain complex networks with fractional-order systems nodes. In comparison with those of the integerorder chaotic system nodes, this method is more efficiently which can enlarge the key space. Moreover, the uncertain network topological structure is identified simultaneously in the process of synchronization. Illustrative examples are presented to demonstrate the application of the theoretical results.

Abstract: For EMC simulation, the vector fitting model is transformed into time domain state equation model. Then the system is balanced. Then the reduced model can be obtained by removing the states corresponding to the small HSV (Hankel Singular Values). The order of the reduced model is determined by the singular curvature spectrum. Finally, balanced truncation model reduction method is used on a sample chip PDN (Power Distribution Network) and compared with the performance of the AWE (Asymptotic Waveform Evaluation) method. Simulation results show that the proposed method can operate in a wide frequency range and has smaller error and faster speed.