Search results for: third order block method

Authors: H. Ghanbari, H. Nikbakht, A. Zahedi, M. Ghanbari

Abstract:

This paper is focusing on designing a control system for wind turbine which can control the speed and output power according to arbitrary algorithm. Reference Tracking Method is used to control the turbine spinning speed in order to increase its output energy.

Keywords: Wind Turbine, Simulink, Reference Tracking Method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1072

Authors: T.C. Manjunath, B. Bandyopadhyay

Abstract:

This paper features the proposed modeling and design of a Robust Decentralized Periodic Output Feedback (RDPOF) control technique for the active vibration control of smart flexible multimodel Euler-Bernoulli cantilever beams for a multivariable (MIMO) case by retaining the first 6 vibratory modes. The beam structure is modeled in state space form using the concept of piezoelectric theory, the Euler-Bernoulli beam theory and the Finite Element Method (FEM) technique by dividing the beam into 4 finite elements and placing the piezoelectric sensor / actuator at two finite element locations (positions 2 and 4) as collocated pairs, i.e., as surface mounted sensor / actuator, thus giving rise to a multivariable model of the smart structure plant with two inputs and two outputs. Five such multivariable models are obtained by varying the dimensions (aspect ratios) of the aluminum beam, thus giving rise to a multimodel of the smart structure system. Using model order reduction technique, the reduced order model of the higher order system is obtained based on dominant eigen value retention and the method of Davison. RDPOF controllers are designed for the above 5 multivariable-multimodel plant. The closed loop responses with the RDPOF feedback gain and the magnitudes of the control input are observed and the performance of the proposed multimodel smart structure system with the controller is evaluated for vibration control.

Keywords: Smart structure, Euler-Bernoulli beam theory, Periodic output feedback control, Finite Element Method, State space model, SISO, Embedded sensors and actuators, Vibration control, Reduced order model

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2032

Authors: Marzieh Dosti, Alireza Nazemi

Abstract:

Recently, it is found that telegraph equation is more suitable than ordinary diffusion equation in modelling reaction diffusion for such branches of sciences. In this paper, a numerical solution for the one-dimensional hyperbolic telegraph equation by using the collocation method using the septic splines is proposed. The scheme works in a similar fashion as finite difference methods. Test problems are used to validate our scheme by calculate L2-norm and L∞-norm. The accuracy of the presented method is demonstrated by two test problems. The numerical results are found to be in good agreement with the exact solutions.

Keywords: B-spline, collocation method, second-order hyperbolic telegraph equation, difference schemes.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1809

Authors: S. Z. Moussavi, A. Badri, F. Rastegar Kashkooli

Abstract:

Wind farms (WFs) with high level of penetration are being established in power systems worldwide more rapidly than other renewable resources. The Independent System Operator (ISO), as a policy maker, should propose appropriate places for WF installation in order to maximize the benefits for the investors. There is also a possibility of congestion relief using the new installation of WFs which should be taken into account by the ISO when proposing the locations for WF installation. In this context, efficient wind farm (WF) placement method is proposed in order to reduce burdens on congested lines. Since the wind speed is a random variable and load forecasts also contain uncertainties, probabilistic approaches are used for this type of study. AC probabilistic optimal power flow (P-OPF) is formulated and solved using Monte Carlo Simulations (MCS). In order to reduce computation time, point estimate methods (PEM) are introduced as efficient alternative for time-demanding MCS. Subsequently, WF optimal placement is determined using generation shift distribution factors (GSDF) considering a new parameter entitled, wind availability factor (WAF). In order to obtain more realistic results, N-1 contingency analysis is employed to find the optimal size of WF, by means of line outage distribution factors (LODF). The IEEE 30-bus test system is used to show and compare the accuracy of proposed methodology.

Keywords: Probabilistic optimal power flow, Wind power, Pointestimate methods, Congestion management

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1897

Authors: Abdolvahab Ehsani Rad, Mohd Shafry Mohd Rahim, Alireza Norouzi

Abstract:

Medical image analysis is one of the great effects of computer image processing. There are several processes to analysis the medical images which the segmentation process is one of the challenging and most important step. In this paper the segmentation method proposed in order to segment the dental radiograph images. Thresholding method has been applied to simplify the images and to morphologically open binary image technique performed to eliminate the unnecessary regions on images. Furthermore, horizontal and vertical integral projection techniques used to extract the each individual tooth from radiograph images. Segmentation process has been done by applying the level set method on each extracted images. Nevertheless, the experiments results by 90% accuracy demonstrate that proposed method achieves high accuracy and promising result.

Keywords: Integral production, level set method, morphological operation, segmentation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 4241

Authors: Thomas Kampke

Abstract:

Equations with differentials relating to the inverse of an unknown function rather than to the unknown function itself are solved exactly for some special cases and numerically for the general case. Invertibility combined with differentiability over connected domains forces solutions always to be monotone. Numerical function inversion is key to all solution algorithms which either are of a forward type or a fixed point type considering whole approximate solution functions in each iteration. The given considerations are restricted to ordinary differential equations with inverted functions (ODEIs) of first order. Forward type computations, if applicable, admit consistency of order one and, under an additional accuracy condition, convergence of order one.

Keywords: Euler method, fixed points, golden section, multi-step procedures, Runge Kutta methods.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1456

Authors: Shimal Das, Dibyendu Ghoshal

Abstract:

The effect of directional search using iterated functional system has been studied on four images taken from databases. The images are portioned successively towards smaller dimension. Presented method provides the faster rate of convergence with respect to processing time in the flat region, but the same has been found to be slower at the border of the images and edges. It has also been revealed that the PSNR is lower at the edges and border portions of the image, and it is found to be higher in the uniform gray region, under the same external illumination and external noise environment.

Keywords: Iterated functional system, fractal compression, structural similarity index measure, fractal block coding, affine transformations.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 929

Authors: Mir Hadi Seyedafsari, Iraj Hasanzadeh

Abstract:

Like other external sorting algorithms, the presented algorithm is a two step algorithm including internal and external steps. The first part of the algorithm is like the other similar algorithms but second part of that is including a new easy implementing method which has reduced the vast number of inputoutput operations saliently. As decreasing processor operating time does not have any effect on main algorithm speed, any improvement in it should be done through decreasing the number of input-output operations. This paper propose an easy algorithm for choose the correct record location of the final list. This decreases the time complexity and makes the algorithm faster.

Keywords: External sorting algorithm, internal sortingalgorithm, fast sorting, robust algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2201

Authors: M.Sushanth Babu, P.Krishna, U.Venu, M.Ranjith

Abstract:

In this paper, we evaluate the choice of suitable quantization characteristics for both the decoder messages and the received samples in Low Density Parity Check (LDPC) coded systems using M-QAM (Quadrature Amplitude Modulation) schemes. The analysis involves the demapper block that provides initial likelihood values for the decoder, by relating its quantization strategy of the decoder. A mapping strategy refers to the grouping of bits within a codeword, where each m-bit group is used to select a 2m-ary signal in accordance with the signal labels. Further we evaluate the system with mapping strategies like Consecutive-Bit (CB) and Bit-Reliability (BR). A new demapper version, based on approximate expressions, is also presented to yield a low complexity hardware implementation.

Keywords: Low Density parity Check, Mapping, Demapping, Quantization, Quadrature Amplitude Modulation

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1741

Authors: Shengbiao Wu, Weihua Cao, Min Wu, Can Liu

Abstract:

This paper presents an intelligent tuning method of microwave filter based on complex neural network and improved space mapping. The tuning process consists of two stages: the initial tuning and the fine tuning. At the beginning of the tuning, the return loss of the filter is transferred to the passband via the error of phase. During the fine tuning, the phase shift caused by the transmission line and the higher order mode is removed by the curve fitting. Then, an Cauchy method based on the admittance parameter (Y-parameter) is used to extract the coupling matrix. The influence of the resonant cavity loss is eliminated during the parameter extraction process. By using processed data pairs (the amount of screw variation and the variation of the coupling matrix), a tuning model is established by the complex neural network. In view of the improved space mapping algorithm, the mapping relationship between the actual model and the ideal model is established, and the amplitude and direction of the tuning is constantly updated. Finally, the tuning experiment of the eight order coaxial cavity filter shows that the proposed method has a good effect in tuning time and tuning precision.

Keywords: Microwave filter, scattering parameter (s-parameter), coupling matrix, intelligent tuning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1326

Authors: Aakash Subramanian, Vansh Pal Singh Makh, Abhijit Mitra

Abstract:

A new digital transceiver circuit for asynchronous frame detection is proposed where both the transmitter and receiver contain all digital components, thereby avoiding possible use of conventional devices like monostable multivibrators with unstable external components such as resistances and capacitances. The proposed receiver circuit, in particular, uses a combinational logic block yielding an output which changes its state as soon as the start bit of a new frame is detected. This, in turn, helps in generating an efficient receiver sampling clock. A data latching circuit is also used in the receiver to latch the recovered data bits in any new frame. The proposed receiver structure is also extended from 4- bit information to any general n data bits within a frame with a common expression for the output of the combinational logic block. Performance of the proposed hardware design is evaluated in terms of time delay, reliability and robustness in comparison with the standard schemes using monostable multivibrators. It is observed from hardware implementation that the proposed circuit achieves almost 33 percent speed up over any conventional circuit.

Keywords: Asynchronous Communication, Digital Detector, Combinational logic output, Sampling clock generator, Hardwareimplementation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2218

Authors: C. Ardil

Abstract:

Uncertainty set theory is a generalization of fuzzy set theory and intuitionistic fuzzy set theory. It serves as an effective tool for dealing with inconsistent, imprecise, and vague information. The technique for order preference by similarity to ideal solution (TOPSIS) method is a multiple-attribute method used to identify solutions from a finite set of alternatives. It simultaneously minimizes the distance from an ideal point and maximizes the distance from a nadir point. In this paper, an extension of the TOPSIS method for multiple attribute group decision-making (MAGDM) based on uncertainty sets is presented. In uncertainty decision analysis, decision-makers express information about attribute values and weights using uncertainty numbers to select the best stealth combat aircraft.

Keywords: Uncertainty set, stealth combat aircraft selection multiple criteria decision-making analysis, MCDM, uncertainty decision analysis, TOPSIS

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 167

Authors: R. Anitha, R. S. Lekshmi

Abstract:

Graph decompositions are vital in the study of combinatorial design theory. Given two graphs G and H, an H-decomposition of G is a partition of the edge set of G into disjoint isomorphic copies of H. An n-sun is a cycle Cn with an edge terminating in a vertex of degree one attached to each vertex. In this paper we have proved that the complete graph of order 2n, K2n can be decomposed into n-2 n-suns, a Hamilton cycle and a perfect matching, when n is even and for odd case, the decomposition is n-1 n-suns and a perfect matching. For an odd order complete graph K2n+1, delete the star subgraph K1, 2n and the resultant graph K2n is decomposed as in the case of even order. The method of building n-suns uses Walecki's construction for the Hamilton decomposition of complete graphs. A spanning tree decomposition of even order complete graphs is also discussed using the labeling scheme of n-sun decomposition. A complete bipartite graph Kn, n can be decomposed into n/2 n-suns when n/2 is even. When n/2 is odd, Kn, n can be decomposed into (n-2)/2 n-suns and a Hamilton cycle.

Keywords: Hamilton cycle, n-sun decomposition, perfectmatching, spanning tree.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2271

Authors: Ahmed Mahmoudi, Imen Mejri, Mohamed A. Abbassi, Ahmed Omri

Abstract:

In this paper, the 1-D conduction-radiation problem is solved by the lattice Boltzmann method. The effects of various parameters such as the scattering albedo, the conduction–radiation parameter and the wall emissivity are studied. In order to check on the accuracy of the numerical technique employed for the solution of the considered problem, the present numerical code was validated with the published study. The found results are in good agreement with those published

Keywords: Conduction, lattice Boltzmann method, planar medium, radiation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2586

Authors: Kyosik Koo, Hyungtai Cha

Abstract:

Generally, in order to create 3D sound using binaural systems, we use head related transfer functions (HRTF) including the information of sounds which is arrived to our ears. But it can decline some three-dimensional effects in the area of a cone of confusion between front and back directions, because of the characteristics of HRTF. In this paper, we propose a new method to use psychoacoustics theory that reduces the confusion of sound image localization. In the method, HRTF spectrum characteristic is enhanced by using the energy ratio of the bark band. Informal listening tests show that the proposed method improves the front-back sound localization characteristics much better than the conventional methods

Keywords: HRTF, 3D sound, Psychoacoustics, Localization

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2031

Authors: Zhengsheng Wang, Jing Qi, Chuntao Liu, Yuanjun Li

Abstract:

The harmonic Arnoldi method can be used to find interior eigenpairs of large matrices. However, it has been shown that this method may converge erratically and even may fail to do so. In this paper, we present a new method for computing interior eigenpairs of large nonsymmetric matrices, which is called weighted harmonic Arnoldi method. The implementation of the method has been tested by numerical examples, the results show that the method converges fast and works with high accuracy.

Keywords: Harmonic Arnoldi method, weighted harmonic Arnoldi method, eigenpair, interior eigenproblem, non symmetric matrix.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1561

Authors: Delaram Jarchi, Reza Boostani

Abstract:

Linear Discrimination Analysis (LDA) is a linear solution for classification of two classes. In this paper, we propose a variant LDA method for multi-class problem which redefines the between class and within class scatter matrices by incorporating a weight function into each of them. The aim is to separate classes as much as possible in a situation that one class is well separated from other classes, incidentally, that class must have a little influence on classification. It has been suggested to alleviate influence of classes that are well separated by adding a weight into between class scatter matrix and within class scatter matrix. To obtain a simple and effective weight function, ordinary LDA between every two classes has been used in order to find Fisher discrimination value and passed it as an input into two weight functions and redefined between class and within class scatter matrices. Experimental results showed that our new LDA method improved classification rate, on glass, iris and wine datasets, in comparison to different versions of LDA.

Keywords: Discriminant vectors, weighted LDA, uncorrelation, principle components, Fisher-face method, Bootstarp method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1533

Authors: Chinwendu. B. Eleje, Udechukwu P. Egbuhuzor

Abstract:

One of the difficulties encountered in solving nonlinear Boundary Value Problems (BVP) by many researchers is finding approximated solutions with minimum deviations from the exact solutions without so much rigor and complications. In this paper, we propose an approach to solve a two point BVP which involves a combination of Taylor series expansion method and Newton Raphson method. Furthermore, the fourth and sixth order approximated solutions are obtained and we compare their relative error and rate of convergence to the exact solution. Finally, some numerical simulations are presented to show the behavior of the solution and its derivatives.

Keywords: Newton Raphson method, non-linear boundary value problem, Taylor series approximation, Michaelis-Menten equation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 345

Authors: G.Mehdiyeva, M.Imanova, V.Ibrahimov

Abstract:

Taking into account that many problems of natural sciences and engineering are reduced to solving initial-value problem for ordinary differential equations, beginning from Newton, the scientists investigate approximate solution of ordinary differential equations. There are papers of different authors devoted to the solution of initial value problem for ODE. The Euler-s known method that was developed under the guidance of the famous scientists Adams, Runge and Kutta is the most popular one among these methods. Recently the scientists began to construct the methods preserving some properties of Adams and Runge-Kutta methods and called them hybrid methods. The constructions of such methods are investigated from the middle of the XX century. Here we investigate one generalization of multistep and hybrid methods and on their base we construct specific methods of accuracy order p = 5 and p = 6 for k = 1 ( k is the order of the difference method).

Keywords: Multistep and hybrid methods, initial value problem, degree and stability of hybrid methods

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1605

Authors: Takayuki Shiina

Abstract:

We consider power system expansion planning under uncertainty. In our approach, integer programming and stochastic programming provide a basic framework. We develop a multistage stochastic programming model in which some of the variables are restricted to integer values. By utilizing the special property of the problem, called block separable recourse, the problem is transformed into a two-stage stochastic program with recourse. The electric power capacity expansion problem is reformulated as the problem with first stage integer variables and continuous second stage variables. The L-shaped algorithm to solve the problem is proposed.

Keywords: electric power capacity expansion problem, integerprogramming, L-shaped method, stochastic programming

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1839

Authors: Ali Raad Hassan, G. Thanigaiyarasu, V. Ramamurti

Abstract:

Natural frequencies and dynamic response of a spur gear sector are investigated using a two dimensional finite element model that offers significant advantages for dynamic gear analyses. The gear teeth are analyzed for different operating speeds. A primary feature of this modeling is determination of mesh forces using a detailed contact analysis for each time step as the gears roll through the mesh. Transient mode super position method has been used to find horizontal and vertical components of displacement and dynamic stress. The finite element analysis software ANSYS has been used on the proposed model to find the natural frequencies by Block Lanczos technique and displacements and dynamic stresses by transient mode super position method. A comparison of theoretical (natural frequency and static stress) results with the finite element analysis results has also been done. The effect of rotational speed of the gears on the dynamic response of gear tooth has been studied and design limits have been discussed.

Keywords: Natural frequency, Modal and transientanalysis, Spur gear, Dynamic stress.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3666

Authors: Delowar Hossain, Genci Capi

Abstract:

This paper concerns with the problem of deep learning parameters tuning using a genetic algorithm (GA) in order to improve the performance of deep learning (DL) method. We present a GA based DL method for robot object recognition and grasping. GA is used to optimize the DL parameters in learning procedure in term of the fitness function that is good enough. After finishing the evolution process, we receive the optimal number of DL parameters. To evaluate the performance of our method, we consider the object recognition and robot grasping tasks. Experimental results show that our method is efficient for robot object recognition and grasping.

Keywords: Deep learning, genetic algorithm, object recognition, robot grasping.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2149

Authors: Hee-Jeong Ahn, Kee-Won Kim, Seung-Hoon Kim

Abstract:

The knowledge of the relationship between characters can help readers to understand the overall story or plot of the literary fiction. In this paper, we present a method for extracting the specific relationship between characters from a Korean literary fiction. Generally, methods for extracting relationships between characters in text are statistical or computational methods based on the sentence distance between characters without considering Korean linguistic features. Furthermore, it is difficult to extract the relationship with direction from text, such as one-sided love, because they consider only the weight of relationship, without considering the direction of the relationship. Therefore, in order to identify specific relationships between characters, we propose a statistical method considering linguistic features, such as syntactic patterns and speech verbs in Korean. The result of our method is represented by a weighted directed graph of the relationship between the characters. Furthermore, we expect that proposed method could be applied to the relationship analysis between characters of other content like movie or TV drama.

Keywords: Data mining, Korean linguistic feature, literary fiction, relationship extraction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1804

Authors: Jieer Wu, Zheshu Ma

Abstract:

Nondestructive testing in engineering is an inverse Cauchy problem for Laplace equation. In this paper the problem of nondestructive testing is expressed by a Laplace-s equation with third-kind boundary conditions. In order to find unknown values on the boundary, the method of fundamental solution is introduced and realized. Because of the ill-posedness of studied problems, the TSVD regularization technique in combination with L-curve criteria and Generalized Cross Validation criteria is employed. Numerical results are shown that the TSVD method combined with L-curve criteria is more efficient than the TSVD method combined with GCV criteria. The abstract goes here.

Keywords: ill-posed, TSVD, Laplace's equation, inverse problem, L-curve, Generalized Cross Validation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1484

Authors: Tamás Hartványi, Ferenc Tóth

Abstract:

European telecommunication distribution center performance is measured by service lead time and quality. Operation model is CTO (customized to order) namely, a high mix customization of telecommunication network equipment and parts. CTO operation contains material receiving, warehousing, network and server assembly to order and configure based on customer specifications. Variety of the product and orders does not support mass production structure. One of the success factors to satisfy customer is to have a proper aggregated planning method for the operation in order to have optimized human resources and highly efficient asset utilization. Research will investigate several methods and find proper way to have an order book simulation where practical optimization problem may contain thousands of variables and the simulation running times of developed algorithms were taken into account with high importance. There are two operation research models that were developed, customer demand is given in orders, no change over time, customer demands are given for product types, and changeover time is constant.

Keywords: CTO, aggregated planning, demand simulation, changeover time.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 795

Authors: M. Saeedi, R. Wuchner, K.-U. Bletzinger

Abstract:

In order to study the aerodynamic performance of a semi-flexible membrane wing, Fluid-Structure Interaction simulations have been performed. The fluid problem has been modeled using two different approaches which are the vortex panel method and the numerical solution of the Navier-Stokes equations. Nonlinear analysis of the structural problem is performed using the Finite Element Method. Comparison between the two fluid solvers has been made. Aerodynamic performance of the wing is discussed regarding its lift and drag coefficients and they are compared with those of the equivalent rigid wing.

Keywords: CFD, FSI, Membrane wing, Vortex panel method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2323

Authors: Tarek Fatouh, Chehab Elbelehy, Alaa Abdelsalam, Eman Elakkad, Alaa Abdelshafie

Abstract:

This paper aims to explore the possibility of time compression in Engineer to Order production networks. A case study research method is used in a Norwegian shipbuilding project by implementing a value stream mapping lean tool with total cycle time as a unit of analysis. The analysis resulted in demonstrating the time deviations for the planned tasks in one of the processes in the shipbuilding project. So, authors developed a future state map by removing time wastes from value stream process.

Keywords: Engineer to order, total cycle time, value stream mapping, shipbuilding.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 605

Authors: Vijay Kumar Kukreja, Ravneet Kaur

Abstract:

In this paper, one dimensional advection diffusion model is analyzed using finite difference method based on Crank-Nicolson scheme. A practical problem of filter cake washing of chemical engineering is analyzed. The model is converted into dimensionless form. For the grid Ω × ω = [0, 1] × [0, T], the Crank-Nicolson spatial derivative scheme is used in space domain and forward difference scheme is used in time domain. The scheme is found to be unconditionally convergent, stable, first order accurate in time and second order accurate in space domain. For a test problem, numerical results are compared with the analytical ones for different values of parameter.

Keywords: Consistency, Crank-Nicolson scheme, Gerschgorin circle, Lax-Richtmyer theorem, Peclet number, stability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 773

Authors: Deokho Kim, Wonwoo Lee, Jaewoong Lee, Teresa Ng, Gun-Ill Lee, Jiwon Jeong

Abstract:

This paper presents a method for improving object search accuracy using a deep learning model. A major limitation to provide accurate similarity with deep learning is the requirement of huge amount of data for training pairwise similarity scores (metrics), which is impractical to collect. Thus, similarity scores are usually trained with a relatively small dataset, which comes from a different domain, causing limited accuracy on measuring similarity. For this reason, this paper proposes a deep learning model that can be trained with a significantly small amount of data, a clustered data which of each cluster contains a set of visually similar images. In order to measure similarity distance with the proposed method, visual features of two images are extracted from intermediate layers of a convolutional neural network with various pooling methods, and the network is trained with pairwise similarity scores which is defined zero for images in identical cluster. The proposed method outperforms the state-of-the-art object similarity scoring techniques on evaluation for finding exact items. The proposed method achieves 86.5% of accuracy compared to the accuracy of the state-of-the-art technique, which is 59.9%. That is, an exact item can be found among four retrieved images with an accuracy of 86.5%, and the rest can possibly be similar products more than the accuracy. Therefore, the proposed method can greatly reduce the amount of training data with an order of magnitude as well as providing a reliable similarity metric.

Keywords: Visual search, deep learning, convolutional neural network, machine learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 838

Authors: Yan Ma, ShunBao Li

Abstract:

A new approach is adopted in this paper based on Turk and Pentland-s eigenface method. It was found that the probability density function of the distance between the projection vector of the input face image and the average projection vector of the subject in the face database, follows Rayleigh distribution. In order to decrease the false acceptance rate and increase the recognition rate, the input face image has been recognized using two thresholds including the acceptance threshold and the rejection threshold. We also find out that the value of two thresholds will be close to each other as number of trials increases. During the training, in order to reduce the number of trials, the projection vectors for each subject has been averaged. The recognition experiments using the proposed algorithm show that the recognition rate achieves to 92.875% whilst the average number of judgment is only 2.56 times.

Keywords: Eigenface, Face Recognition, Threshold, Rayleigh Distribution, Feature Extraction

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1501