Abstract: In this paper, a new face recognition method based on
PCA (principal Component Analysis), LDA (Linear Discriminant
Analysis) and neural networks is proposed. This method consists of
four steps: i) Preprocessing, ii) Dimension reduction using PCA, iii)
feature extraction using LDA and iv) classification using neural
network. Combination of PCA and LDA is used for improving the
capability of LDA when a few samples of images are available and
neural classifier is used to reduce number misclassification caused by
not-linearly separable classes. The proposed method was tested on
Yale face database. Experimental results on this database
demonstrated the effectiveness of the proposed method for face
recognition with less misclassification in comparison with previous
methods.
Abstract: Carbon Capture & Storage (CCS) is one of the various
methods that can be used to reduce the carbon footprint of the
energy sector. This paper focuses on the absorption of CO2 from
flue gas using packed columns, whose efficiency is highly dependent
on the structure of the liquid films within the column. To study the
characteristics of liquid films a CFD solver, OpenFOAM is utilised
to solve two-phase, isothermal film flow using the volume-of-fluid
(VOF) method. The model was validated using existing experimental
data and the Nusselt theory. It was found that smaller plate inclination
angles, with respect to the horizontal plane, resulted in larger wetted
areas on smooth plates. However, only a slight improvement in
the wetted area was observed. Simulations were also performed
using a ridged plate and it was observed that these surface textures
significantly increase the wetted area of the plate. This was mainly
attributed to the channelling effect of the ridges, which helped to
oppose the surface tension forces trying to minimise the surface area.
Rivulet formations on the ridged plate were also flattened out and
spread across a larger proportion of the plate width.
Abstract: Camera calibration is an important step in 3D
reconstruction. Camera calibration may be classified into two major types: traditional calibration and self-calibration. However, a calibration method in using a checkerboard is intermediate between traditional calibration and self-calibration. A self
is proposed based on a square in this paper. Only a square in the planar
template, the camera self-calibration can be completed through the single view. The proposed algorithm is that the virtual circle and straight line are established by a square on planar template, and
circular points, vanishing points in straight lines and the relation
between them are be used, in order to obtain the image of the absolute
conic (IAC) and establish the camera intrinsic parameters. To make
the calibration template is simpler, as compared with the Zhang Zhengyou-s method. Through real experiments and experiments, the experimental results show that this algorithm is
feasible and available, and has a certain precision and robustness.
Abstract: Extraction of Fe(III) from aqueous solution using Trin-
butyl Phosphate (TBP) as carrier needs a highly acidic medium
(>6N) as it favours formation of chelating complex FeCl3.TBP.
Similarly, stripping of Iron(III) from loaded organic solvents requires
neutral pH or alkaline medium to dissociate the same complex. It is
observed that TBP co-extracts acids along with metal, which causes
reversal of driving force of extraction and iron(III) is re-extracted
back from the strip phase into the feed phase during Liquid Emulsion
Membrane (LEM) pertraction. Therefore, rate of extraction of
different mineral acids (HCl, HNO3, H2SO4) using TBP with and
without presence of metal Fe(III) was examined. It is revealed that in
presence of metal acid extraction is enhanced. Determination of mass
transfer coefficient of both acid and metal extraction was performed
by using Bulk Liquid Membrane (BLM). The average mass transfer
coefficient was obtained by fitting the derived model equation with
experimentally obtained data. The mass transfer coefficient of the
mineral acid extraction is in the order of kHNO3 = 3.3x10-6m/s > kHCl =
6.05x10-7m/s > kH2SO4 = 1.85x10-7m/s. The distribution equilibria of
the above mentioned acids between aqueous feed solution and a
solution of tri-n-butyl-phosphate (TBP) in organic solvents have been
investigated. The stoichiometry of acid extraction reveals the
formation of TBP.2HCl, HNO3.2TBP, and TBP.H2SO4 complexes.
Moreover, extraction of Iron(III) by TBP in HCl aqueous solution
forms complex FeCl3.TBP.2HCl while in HNO3 medium forms
complex 3FeCl3.TBP.2HNO3
Abstract: α-Pinene is the main component of the most
turpentine oils. The hydration of α-pinene with acid catalysts leads to
a complex mixture of monoterpenes. In order to obtain more valuable
products, the α-pinene in the turpentine can be hydrated in dilute
mineral acid solutions to produce α-terpineol. The design of
separation processes requires information on phase equilibrium and
related thermodynamic properties. This paper reports the results of
study on liquid-liquid equilibrium (LLE) of system containing α-
pinene + water and α-terpineol + water.
Binary LLE for α-pinene + water system, and α-terpineol + water
systems were determined by experiment at 301K and atmospheric
pressure. The two component mixture was stirred for about 30min,
then the mixture was left for about 2h for complete phase separation.
The composition of both phases was analyzed by using a Gas
Chromatograph. The experimental data were correlated by
considering both NRTL and UNIQUAC activity coefficient models.
The LLE data for the system of α-pinene + water and α-terpineol +
water were correlated successfully by the NRTL model. The
experimental data were not satisfactorily fitted by the UNIQUAC
model. The NRTL model (α =0.3) correlates the LLE data for the
system of α-pinene + water at 301K with RMSD of 0.0404%. And
the NRTL model (α =0.61) at 301K with RMSD of 0.0058 %. The
NRTL model (α =0.3) correlates the LLE data for the system of α-
terpineol + water at 301K with RMSD of 0.1487% and the NRTL
model (α =0.6) at 301K with RMSD of 0.0032%, between the
experimental and calculated mole fractions.
Abstract: In this paper, a target signal detection method using
multiple signal classification (MUSIC) algorithm is proposed. The
MUSIC algorithm is a subspace-based direction of arrival (DOA)
estimation method. The algorithm detects the DOAs of multiple
sources using the inverse of the eigenvalue-weighted eigen spectra. To
apply the algorithm to target signal detection for GSC-based
beamforming, we utilize its spectral response for the target DOA in
noisy conditions. For evaluation of the algorithm, the performance of
the proposed target signal detection method is compared with that of
the normalized cross-correlation (NCC), the fixed beamforming, and
the power ratio method. Experimental results show that the proposed
algorithm significantly outperforms the conventional ones in receiver
operating characteristics(ROC) curves.
Abstract: Nowadays, precipitation prediction is required for proper planning and management of water resources. Prediction with neural network models has received increasing interest in various research and application domains. However, it is difficult to determine the best neural network architecture for prediction since it is not immediately obvious how many input or hidden nodes are used in the model. In this paper, neural network model is used as a forecasting tool. The major aim is to evaluate a suitable neural network model for monthly precipitation mapping of Myanmar. Using 3-layerd neural network models, 100 cases are tested by changing the number of input and hidden nodes from 1 to 10 nodes, respectively, and only one outputnode used. The optimum model with the suitable number of nodes is selected in accordance with the minimum forecast error. In measuring network performance using Root Mean Square Error (RMSE), experimental results significantly show that 3 inputs-10 hiddens-1 output architecture model gives the best prediction result for monthly precipitation in Myanmar.
Abstract: Our objectives were to evaluate the effects of sire
breed, type of protein supplement, level of supplementation and sex
on wool spinning fineness (SF), its correlations with other wool
characteristics and prediction accuracy in F1 Merino crossbred lambs.
Texel, Coopworth, White Suffolk, East Friesian and Dorset rams
were mated with 500 purebred Merino dams at a ratio of 1:100 in
separate paddocks within a single management system. The F1
progeny were raised on ryegrass pasture until weaning, before forty
lambs were randomly allocated to treatments in a 5 x 2 x 2 x 2
factorial experimental design representing 5 sire breeds, 2
supplementary feeds (canola or lupins), 2 levels of supplementation
(1% or 2% of liveweight) and sex (wethers or ewes). Lambs were
supplemented for six weeks after an initial three weeks of adjustment,
wool sampled at the commencement and conclusion of the feeding
trial and analyzed for SF, mean fibre diameter (FD), coefficient of
variation (CV), standard deviation, comfort factor (CF), fibre
curvature (CURV), and clean fleece yield. Data were analyzed using
mixed linear model procedures with sire fitted as a random effect,
and sire breed, sex, supplementary feed type, level of
supplementation and their second-order interactions as fixed effects.
Sire breed (P
Abstract: Analysis and visualization of microarraydata is veryassistantfor biologists and clinicians in the field of diagnosis and treatment of patients. It allows Clinicians to better understand the structure of microarray and facilitates understanding gene expression in cells. However, microarray dataset is a complex data set and has thousands of features and a very small number of observations. This very high dimensional data set often contains some noise, non-useful information and a small number of relevant features for disease or genotype. This paper proposes a non-linear dimensionality reduction algorithm Local Principal Component (LPC) which aims to maps high dimensional data to a lower dimensional space. The reduced data represents the most important variables underlying the original data. Experimental results and comparisons are presented to show the quality of the proposed algorithm. Moreover, experiments also show how this algorithm reduces high dimensional data whilst preserving the neighbourhoods of the points in the low dimensional space as in the high dimensional space.
Abstract: Pretreatment is an essential step in the conversion of
lignocellulosic biomass to fermentable sugar that used for biobutanol
production. Among pretreatment processes, microwave is considered
to improve pretreatment efficiency due to its high heating efficiency,
easy operation, and easily to combine with chemical reaction. The
main objectives of this work are to investigate the feasibility of
microwave pretreatment to enhance enzymatic hydrolysis of
corncobs and to determine the optimal conditions using response
surface methodology. Corncobs were pretreated via two-stage
pretreatment in dilute sodium hydroxide (2 %) followed by dilute
sulfuric acid 1 %. Pretreated corncobs were subjected to enzymatic
hydrolysis to produce reducing sugar. Statistical experimental design
was used to optimize pretreatment parameters including temperature,
residence time and solid-to-liquid ratio to achieve the highest amount
of glucose. The results revealed that solid-to-liquid ratio and
temperature had a significant effect on the amount of glucose.
Abstract: This paper proposes a novel approach to the question of lithofacies classification based on an assessment of the uncertainty in the classification results. The proposed approach has multiple neural networks (NN), and interval neutrosophic sets (INS) are used to classify the input well log data into outputs of multiple classes of lithofacies. A pair of n-class neural networks are used to predict n-degree of truth memberships and n-degree of false memberships. Indeterminacy memberships or uncertainties in the predictions are estimated using a multidimensional interpolation method. These three memberships form the INS used to support the confidence in results of multiclass classification. Based on the experimental data, our approach improves the classification performance as compared to an existing technique applied only to the truth membership. In addition, our approach has the capability to provide a measure of uncertainty in the problem of multiclass classification.
Abstract: The ability to predict an accurate temperature
distribution requires the knowledge of the losses, the thermal
characteristics of the materials, and the cooling conditions, all of
which are very difficult to quantify. In this paper, the impact of the
effects of iron and copper losses are investigated separately and
their effects on the heating in various points of the stator of an
induction motor, is highlighted by using two simple tests. In addition,
the effect of a defect, such as an open circuit in a phase of the stator,
on the heating is also obtained by a no-load test.
The squirrel cage induction motor is rated at 2.2 kW; 380 V; 5.2
A; Δ connected; 50 Hz; 1420 rpm and the class of insulation F, has
been thermally tested under several load conditions. Several
thermocouples were placed in strategic points of the stator.
Abstract: With a development of Hybrid Electric Vehicle(HEV),
A photovoltaic(PV) generation system is used for charging batteries in many cases. A dc/dc converter using PV power for a battery charger
requires a high efficiency. In this paper, A ZVS boost converter using the renewable energies for HEV charger is proposed. Through the theoretical analysis and experimental result, operation modes and characteristics of the proposed topology are verified.
Abstract: Lateral expansion is a factor defining the level of
confinement in reinforced concrete columns. Therefore, predicting
the lateral strain relationship with axial strain becomes an important
issue. Measuring lateral strains in experiments is difficult and only
few report experimental lateral strains. Among the existing analytical
formulations, two recent models are compared with available test
results in this paper with shortcomings highlighted. A new analytical
model is proposed here for lateral strain axial strain relationship and
is based on the supposition that the concrete behaves linear elastic in
the early stages of loading and then nonlinear hardening up to the
peak stress and then volumetric expansion. The proposal for the
lateral strain axial strain relationship after the peak stress is mainly
based on the hypothesis that the plastic lateral strain varies linearly
with the plastic axial strain and it is shown that this is related to the
lateral confinement level.
Abstract: Laminar natural-convective heat transfer from a
horizontal cylinder is studied by solving the Navier-Stokes and
energy equations using higher order compact scheme in cylindrical
polar coordinates. Results are obtained for Rayleigh numbers of 1,
10, 100 and 1000 for a Prandtl number of 0.7. The local Nusselt
number and mean Nusselt number are calculated and compared with
available experimental and theoretical results. Streamlines, vorticity -
lines and isotherms are plotted.
Abstract: Predicting protein-protein interactions represent a key step in understanding proteins functions. This is due to the fact that proteins usually work in context of other proteins and rarely function alone. Machine learning techniques have been applied to predict protein-protein interactions. However, most of these techniques address this problem as a binary classification problem. Although it is easy to get a dataset of interacting proteins as positive examples, there are no experimentally confirmed non-interacting proteins to be considered as negative examples. Therefore, in this paper we solve this problem as a one-class classification problem using one-class support vector machines (SVM). Using only positive examples (interacting protein pairs) in training phase, the one-class SVM achieves accuracy of about 80%. These results imply that protein-protein interaction can be predicted using one-class classifier with comparable accuracy to the binary classifiers that use artificially constructed negative examples.
Abstract: The chatter is one of the major limitations of the productivity in the ball end milling process. It affects the surface roughness, the dimensional accuracy and the tool life. The aim of this research is to propose the new system to detect the chatter during the ball end milling process by using the wavelet transform. The proposed method is implemented on the 5-axis CNC machining center and the new three parameters are introduced from three dynamic cutting forces, which are calculated by taking the ratio of the average variances of dynamic cutting forces to the absolute variances of themselves. It had been proved that the chatter can be easier to detect during the in-process cutting by using the new parameters which are proposed in this research. The experimentally obtained results showed that the wavelet transform can provide the reliable results to detect the chatter under various cutting conditions.
Abstract: In this paper, a new version of support vector regression (SVR) is presented namely Fuzzy Cost SVR (FCSVR). Individual property of the FCSVR is operation over fuzzy data whereas fuzzy cost (fuzzy margin and fuzzy penalty) are maximized. This idea admits to have uncertainty in the penalty and margin terms jointly. Robustness against noise is shown in the experimental results as a property of the proposed method and superiority relative conventional SVR.
Abstract: In this paper three basic approaches and different
methods under each of them for extracting region of interest (ROI)
from stationary images are explored. The results obtained for each of
the proposed methods are shown, and it is demonstrated where each
method outperforms the other. Two main problems in ROI
extraction: the channel selection problem and the saliency reversal
problem are discussed and how best these two are addressed by
various methods is also seen. The basic approaches are 1) Saliency
based approach 2) Wavelet based approach 3) Clustering based
approach. The saliency approach performs well on images containing
objects of high saturation and brightness. The wavelet based
approach performs well on natural scene images that contain regions
of distinct textures. The mean shift clustering approach partitions the
image into regions according to the density distribution of pixel
intensities. The experimental results of various methodologies show
that each technique performs at different acceptable levels for
various types of images.
Abstract: In this paper a novel, simple and reliable digital firing
scheme has been implemented for speed control of three-phase
induction motor using ac voltage controller. The system consists of
three-phase supply connected to the three-phase induction motor via
three triacs and its control circuit. The ac voltage controller has three
modes of operation depending on the shape of supply current. The
performance of the induction motor differs in each mode where the
speed is directly proportional with firing angle in two modes and
inversely in the third one. So, the control system has to detect the
current mode of operation to choose the correct firing angle of triacs.
Three sensors are used to feed the line currents to control system to
detect the mode of operation. The control strategy is implemented
using a low cost Xilinx Spartan-3E field programmable gate array
(FPGA) device. Three PI-controllers are designed on FPGA to
control the system in the three-modes. Simulation of the system is
carried out using PSIM computer program. The simulation results
show stable operation for different loading conditions especially in
mode 2/3. The simulation results have been compared with the
experimental results from laboratory prototype.