Papers by Keyword: Identification

Paper TitlePage

Authors: Jun Wei Ge, Ming Zhao, Yi Qiu Fang
Abstract: This paper presents a rapid identification model and analyses the behavioral characteristics which is different from non-P2P applications on link pattern through analysis on three P2P applications. This method classifies P2P applications in the background and improves the recognition efficiency through the effective combination of behavioral characteristics and valid flows filter on the premise of maintaining the recognition accuracy. In the packet processing, matching frequency parameter has been using to increase matching efficiency. The experimental results show that P2P traffic can be effectively identified by this method.
Authors: Yong Shi, Wen Tao Liu
Abstract: In order to identify the geometrical parameters of parallel kinematics machines tools (PKM), a new parameters identification method is presented. The identification method is proposed based on a pose discrepancy model, which is deduced from the error between the nominal and measurement relative distance of two different spatial locations of the moving platform. In the identification method, an error sensitivity matrix, which expresses the sensitivity between the pose error and geometrical structural parameters error of PKM, can be created with numerical methods. The results of different numerical methods are analyzed. A measurement method to get the precise lengths of legs is presented, which decrease the number of identified parameters. In an experiment, the error of PKM is reduced from 6.71mm to 1.144mm. Therefore, the identification method is verified effective and feasible.
Authors: Ming Lei Ju, Hamdi Jmal, Raphaël Dupuis, Evelyne Aubry
Abstract: Polyurethane foam is widely used in numerous applications. The quasi-static mechanical behaviour of polyurethane foam shows a nonlinear elasticity and a viscoelasticity in large compression deformation. In this paper, the uniaxial compression / decompression tests in three different strain rates are performed. The purpose of this paper is to propose three visco-hyperelastic models and to identify the parameters based on the experimental results. A difference-stress proper identification method is used to obtain the model parameters. The results show that the three models can predict the mechanical behaviour of polyurethane foam very well. Finally, the three models are compared according to the model results.
Authors: Gui Dong Yuan, Cui Ying Li
Abstract: This paper show a flexible chain conveying system with workpiece model marking, which could meet the need of the flexible manufacturing system for the processing different types products on a single product line. By using the hall switches and forming a various placement arrangement of the magnet sheets adhering to the workpiece carrier bracket on the conveyor belt; the identification results give a good correspondence one by one between the part and the type information. The workpiece carrier on the entire conveyor belt can be changed flexibility by changing the magnet sheets placement at the same time, because of the easy replacement of the magnet sheets. With the help of PLC control system, it turn to reality to complete automatically feeding work with discharge manipulator or unloading manipulator, to convey and unload automatic material delivery process. The use of chain transmission combined with the energy consumption braking can realize the accurate positioning of the belt when parking.
Authors: Hong Zheng Li, He Zhou, Yao Hong Jin
Abstract: In this paper, we introduced one kind of special compound noun in Chinese patent texts composed of verb and noun, and presented a rule-based method for Chinese-English patent Machine Translation (MT) to improve the identification of compound nouns, with the purpose of decreasing the possibilities that verbs may disturb the identification of core predicate verb. The system first tagged different weights on verbs then determined the properties of verbs and recognized the compounds according to the weights. We then conducted experiments with the method, which proved that the method could identify compound nouns efficiently.
Authors: Yukio Toyozawa, Naoto Sonoda, Hiroshi Harada, Hiroshi Kashiwagi
Abstract: This paper describes a practical method to identify mechanical parameters of two-inertia mechanical system such as a linear actuator driven by a ball screw mechanism. In the proposed method, mechanical system is represented as an inverse model and a low-pass filtered M-sequence is used as an input signal. The mechanical parameters are estimated by using steepest descent method with a dead band. From the results of the computer simulation, the identified mechanical parameters quickly approach to the true mechanical parameters.
Authors: Min Hu, Fang Fang Wu, Bo Zhu, Bo Lu, Jing Lei Pu
Abstract: It is important and difficult to identify the Hazard before a disaster happen because disaster often happens suddenly. This paper proposes a new method – State Transition Graph, which based on visual data space reconstruction, to identify hazard. The change process of the system state movement from one state to another in a certain period is described by some state transition graphs. The system state, which is safe or hazard, could be distinguished by its state transition graphs. This paper conducted experiments on single-dimension and multi-dimension benchmark data to prove the new method is effectiveness. Especially the result of stimulation experiments, based on the Yangtze River tunnel engineering data, showed that state transition graph identifies hazard easily and has better performances than other method. The State transition graph method is worth further researching.
Authors: Fan Li, Jian Qin Mao, Hai Shan Ding, Wen Bo Zhang, Hui Bin Xu, Cheng Bao Jiang
Abstract: In this paper, a new method which combines the least square method with Tree-Structured fuzzy inference system is presented to approximate the Preisach distribution function. Firstly, by devising the input sequence and measure the output, discrete Preisach measure can be identified by the use of the least squares method. Then, the Preisach function can be obtained with Tree-Structured fuzzy inference system without any special smoothing means. So, this new method is not sensitive to noise, and is a universal approximator of the Preisach function. It collect the merit and overcome the deficiency of the existing methods.
Authors: Yan Mei Li, Jian Luo
Abstract: Minor component analysis (MCA) plays an important role in signal processing and data analysis. In this paper, we propose a novel watermarking scheme using MCA neural networks. The proposed scheme can detect and localize any modification made to the image, which can be used for integrity verification and authentication. Another important property of this proposed scheme is that any gray scale watermark can be easily inserted into the image compared to some other watermarking method such as LSB (least significant bit) method. Furthermore, the proposed scheme can insert any watermark into one image without making any visible changes to the image. Since the minor component of each image is different, and it requires a user key during both the insertion and extraction procedures, it is not possible for an infringer to insert a new watermark or alter the existing one that could pass the test. Finally, various experiments also confirm such properties of the proposed scheme.
Authors: Gui Xin Su, Peng Ge, Jing Jing Song, Hong Mei Ding, Jie Lan, Jie Sun, Xu Guo Sun
Abstract: Objective: Employ a new a method using Microfluidic Cell Chip improve the recognition rate of cells in urine. This method will decrease errors caused by failure to distinguish cell in urine based on the overlapping cell morphology characteristics and cell parameters. Methods: (1) enroll 60 patients respectively of acute glomerulonephritis, acute pyelonephritis and acute cystitis, employ WB-2000 automated urine analyzer to detect 9 kinds of biochemical indexes urine protein, glucose, ketone, uric bilirubin and urobilinogen, nitrite, pH, erythrocytes and leukocytes. (2) Observe the overlap count of erythrocytes and leukocytes in the urine of three groups of patients, and calculate overlap rate of erythrocytes and leukocytes of each group of the patient's respectively. (3) After separating urine cells with Microfluidic Cell Chip technology, test the overlap rate of erythrocytes and leukocytes. Results: (1) The overlap rate of erythrocytes in acute glomerulonephritis patients is 8.53% ~ 8.72%, and the overlap rate of leukocytes is 15.51% ~ 17.18%; The overlap rate of erythrocytes in pyelonephritis patients is 3.64 ~ 4.95%, while the overlap rate of leukocytes is from 8.18 to 9.23%; The overlap rate of erythrocytes in cystitis patients is between 3.85% and 4.53%, and the overlap rate of leukocytes is 8.71% ~ 7.85%; In the glomerulonephritis group, the protein in urine is in the highest levels, the overlap rate of erythrocytes and leukocytes is higher than other groups significantly. (2) the Microfluidic Cell Chip technology can reduce the urinary cells overlap rate of three groups of patients, to levels of 0.22% ~ 0.28%. Conclusion: Microfluidic Cell Chip technology did reduce the overlap rate of erythrocytes and leukocytes in urine samples from patients with three different urinary tract disease.
Showing 1 to 10 of 294 Paper Titles