2012
|
Altun, Kerem; Barshan, Billur Pedestrian dead reckoning employing simultaneous activity recognition
cues Journal Article MEASUREMENT SCIENCE AND TECHNOLOGY, 23 (2), 2012, ISSN: 0957-0233. Links | BibTeX @article{ISI:000300057900012,
title = {Pedestrian dead reckoning employing simultaneous activity recognition
cues},
author = {Kerem Altun and Billur Barshan},
doi = {10.1088/0957-0233/23/2/025103},
issn = {0957-0233},
year = {2012},
date = {2012-02-01},
journal = {MEASUREMENT SCIENCE AND TECHNOLOGY},
volume = {23},
number = {2},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
2011
|
Gezgin, Erkin; Ozdemir, Serhan Classification of manipulators of the same origin by virtue of
compactness and complexity Journal Article MECHANISM AND MACHINE THEORY, 46 (10), pp. 1425-1433, 2011, ISSN: 0094-114X. Abstract | Links | BibTeX @article{ISI:000293062700007,
title = {Classification of manipulators of the same origin by virtue of
compactness and complexity},
author = {Erkin Gezgin and Serhan Ozdemir},
doi = {10.1016/j.mechmachtheory.2011.05.009},
issn = {0094-114X},
year = {2011},
date = {2011-10-01},
journal = {MECHANISM AND MACHINE THEORY},
volume = {46},
number = {10},
pages = {1425-1433},
abstract = {This work deals with a classification method that employs concepts such
as complexity and compactness. The idea is to classify manipulators, or
any other mechanism for that matter, of the same origin, based on the
geometry of the joints, the tasks performed by the joints, the
efficiency and the manufacturing cost to generate the specified
efficiency. It is known that successive units on a single branch create
individual uncertainties that affect the eventual quality of the
performed operation [1]. An entropic expression quantifies this
uncertainty in terms of the number of links and the unit effectiveness.
The concepts of compactness and complexity have been formulated, and
these concepts are explained through serial and parallel manipulators
with varying parameters. Eventually, a cost function is created which is
a function of complexity, uncertainty and the manufacturing cost. A worked example on M = 6 Stewart-Gough platform is given how this cost
function could be taken advantage of when deciding an initial
manipulator. A genetic algorithm is used for the optimization of the
cost function, where the results are tabulated. (C) 2011 Elsevier Ltd.
All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This work deals with a classification method that employs concepts such
as complexity and compactness. The idea is to classify manipulators, or
any other mechanism for that matter, of the same origin, based on the
geometry of the joints, the tasks performed by the joints, the
efficiency and the manufacturing cost to generate the specified
efficiency. It is known that successive units on a single branch create
individual uncertainties that affect the eventual quality of the
performed operation [1]. An entropic expression quantifies this
uncertainty in terms of the number of links and the unit effectiveness.
The concepts of compactness and complexity have been formulated, and
these concepts are explained through serial and parallel manipulators
with varying parameters. Eventually, a cost function is created which is
a function of complexity, uncertainty and the manufacturing cost. A worked example on M = 6 Stewart-Gough platform is given how this cost
function could be taken advantage of when deciding an initial
manipulator. A genetic algorithm is used for the optimization of the
cost function, where the results are tabulated. (C) 2011 Elsevier Ltd.
All rights reserved. |
Coskun, Anil; Sevil, Hakki Erhan; Ozdemir, Serhan Cost effective localization in distributed sensory networks Journal Article ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 24 (2), pp. 232-237, 2011, ISSN: 0952-1976. Abstract | Links | BibTeX @article{ISI:000287066800003,
title = {Cost effective localization in distributed sensory networks},
author = {Anil Coskun and Hakki Erhan Sevil and Serhan Ozdemir},
doi = {10.1016/j.engappai.2010.10.006},
issn = {0952-1976},
year = {2011},
date = {2011-03-01},
journal = {ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE},
volume = {24},
number = {2},
pages = {232-237},
abstract = {The most important mechanism to occur in biological distributed sensory
networks (DSNs) is called lateral inhibition, (LI). LI relies on one
simple principle. Each sensor strives to suppress its neighbors in
proportion to its own excitation. In this study, LI mechanism is
exploited to localize the unknown position of a light source that
illuminated the photosensitive sensory network containing high and low
quality sensors. Each photosensitive sensor was then calibrated to
accurately read the distance to the light source. A series of
experiments were conducted employing both quality sensors. Low quality
array was allowed to take advantage of LI, whereas the high quality one
was not. Results showed that the lateral inhibition mechanism increased
the sensitivity of inferior quality sensors, giving the ability to make
the localization as sensitive as high quality sensors do. This suggests
that the networks with multitude of sensors could be made
cost-effective, were these sensory networks equipped with LI. (C) 2010
Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The most important mechanism to occur in biological distributed sensory
networks (DSNs) is called lateral inhibition, (LI). LI relies on one
simple principle. Each sensor strives to suppress its neighbors in
proportion to its own excitation. In this study, LI mechanism is
exploited to localize the unknown position of a light source that
illuminated the photosensitive sensory network containing high and low
quality sensors. Each photosensitive sensor was then calibrated to
accurately read the distance to the light source. A series of
experiments were conducted employing both quality sensors. Low quality
array was allowed to take advantage of LI, whereas the high quality one
was not. Results showed that the lateral inhibition mechanism increased
the sensitivity of inferior quality sensors, giving the ability to make
the localization as sensitive as high quality sensors do. This suggests
that the networks with multitude of sensors could be made
cost-effective, were these sensory networks equipped with LI. (C) 2010
Elsevier Ltd. All rights reserved. |
Sevil, Hakki Erhan; Ozdemir, Serhan Prediction of microdrill breakage using rough sets Journal Article AI EDAM-ARTIFICIAL INTELLIGENCE FOR ENGINEERING DESIGN ANALYSIS AND
MANUFACTURING, 25 (1), pp. 15-23, 2011, ISSN: 0890-0604. Abstract | Links | BibTeX @article{ISI:000287388000002,
title = {Prediction of microdrill breakage using rough sets},
author = {Hakki Erhan Sevil and Serhan Ozdemir},
doi = {10.1017/S0890060410000144},
issn = {0890-0604},
year = {2011},
date = {2011-02-01},
journal = {AI EDAM-ARTIFICIAL INTELLIGENCE FOR ENGINEERING DESIGN ANALYSIS AND
MANUFACTURING},
volume = {25},
number = {1},
pages = {15-23},
abstract = {This study attempts to correlate the nonlinear invariants' with the
changing conditions of a drilling process through a series of condition
monitoring experiments on small diameter (1 mm) drill bits.
Run-to-failure tests are performed on these drill bits, and vibration
data are consecutively gathered at equal time intervals. Nonlinear
invariants, such as the Kolmogorov entropy and correlation dimension,
and statistical parameters are calculated based on the corresponding
conditions of the drill bits. By intervariations of these values between
two successive measurements, a drop rise table is created. Any variation
that is within a certain threshold (+/-20% of the measurements in this
case) is assumed to be constant. Any fluctuation above or below is
assumed to be either a rise or a drop. The reduct and conflict tables
then help eliminate incongruous and redundant data by the use of rough
sets (RSs). Inconsistent data, which by definition is the boundary
region, are classified through certainty and coverage factors. By
handling inconsistencies and redundancies, 11 rules are extracted from
39 experiments, representing the underlying rules. Then 22 new
experiments are used to check the validity of the rule space. The RS
decision frame performs best at predicting no failure cases. It is
believed that RSs are superior in dealing with real-life data over fuzzy
set logic in that actual measured data are never as consistent as here
and may dominate the monitoring of the manufacturing processes as it
becomes more widespread.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This study attempts to correlate the nonlinear invariants' with the
changing conditions of a drilling process through a series of condition
monitoring experiments on small diameter (1 mm) drill bits.
Run-to-failure tests are performed on these drill bits, and vibration
data are consecutively gathered at equal time intervals. Nonlinear
invariants, such as the Kolmogorov entropy and correlation dimension,
and statistical parameters are calculated based on the corresponding
conditions of the drill bits. By intervariations of these values between
two successive measurements, a drop rise table is created. Any variation
that is within a certain threshold (+/-20% of the measurements in this
case) is assumed to be constant. Any fluctuation above or below is
assumed to be either a rise or a drop. The reduct and conflict tables
then help eliminate incongruous and redundant data by the use of rough
sets (RSs). Inconsistent data, which by definition is the boundary
region, are classified through certainty and coverage factors. By
handling inconsistencies and redundancies, 11 rules are extracted from
39 experiments, representing the underlying rules. Then 22 new
experiments are used to check the validity of the rule space. The RS
decision frame performs best at predicting no failure cases. It is
believed that RSs are superior in dealing with real-life data over fuzzy
set logic in that actual measured data are never as consistent as here
and may dominate the monitoring of the manufacturing processes as it
becomes more widespread. |
2010
|
Altun, Kerem; Barshan, Billur; Tuncel, Orkun Comparative study on classifying human activities with miniature
inertial and magnetic sensors Journal Article PATTERN RECOGNITION, 43 (10), pp. 3605-3620, 2010, ISSN: 0031-3203. Links | BibTeX @article{ISI:000280006700036,
title = {Comparative study on classifying human activities with miniature
inertial and magnetic sensors},
author = {Kerem Altun and Billur Barshan and Orkun Tuncel},
doi = {10.1016/j.patcog.2010.04.019},
issn = {0031-3203},
year = {2010},
date = {2010-10-01},
journal = {PATTERN RECOGNITION},
volume = {43},
number = {10},
pages = {3605-3620},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
Akgun, Yenal; Gantes, Charis J; Kalochairetis, Konstantinos E; Kiper, Gokhan A novel concept of convertible roofs with high transformability
consisting of planar scissor-hinge structures Journal Article ENGINEERING STRUCTURES, 32 (9), pp. 2873-2883, 2010, ISSN: 0141-0296. Abstract | Links | BibTeX @article{ISI:000281995300032,
title = {A novel concept of convertible roofs with high transformability
consisting of planar scissor-hinge structures},
author = {Yenal Akgun and Charis J Gantes and Konstantinos E Kalochairetis and Gokhan Kiper},
doi = {10.1016/j.engstruct.2010.05.006},
issn = {0141-0296},
year = {2010},
date = {2010-09-01},
journal = {ENGINEERING STRUCTURES},
volume = {32},
number = {9},
pages = {2873-2883},
abstract = {In this paper, a new adaptive scissor-hinge structure is introduced,
which can be converted by means of actuators between a multitude of
curvilinear arch-like shapes, where it can be stabilized and carry
loads. The key point of this new structure is the proposed Modified
Scissor-Like Element (M-SLE). With the development of this element, it
becomes possible to change the geometry of the whole system without
changing the dimensions of the struts or the span. The proposed
scissor-hinge structure discussed here is planar, but it is also
possible to combine structures in groups to create spatial systems.
After outlining the differences of the proposed structure with existing
designs, the dimensional properties of the M-SLE are introduced. Then,
geometric principles and shape limitations of the whole structure are
explained. Finally, structural analysis of the structure in different
geometric configurations is performed, in order to discuss stiffness
limitations associated with the advantage of increased mobility. (C)
2010 Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In this paper, a new adaptive scissor-hinge structure is introduced,
which can be converted by means of actuators between a multitude of
curvilinear arch-like shapes, where it can be stabilized and carry
loads. The key point of this new structure is the proposed Modified
Scissor-Like Element (M-SLE). With the development of this element, it
becomes possible to change the geometry of the whole system without
changing the dimensions of the struts or the span. The proposed
scissor-hinge structure discussed here is planar, but it is also
possible to combine structures in groups to create spatial systems.
After outlining the differences of the proposed structure with existing
designs, the dimensional properties of the M-SLE are introduced. Then,
geometric principles and shape limitations of the whole structure are
explained. Finally, structural analysis of the structure in different
geometric configurations is performed, in order to discuss stiffness
limitations associated with the advantage of increased mobility. (C)
2010 Elsevier Ltd. All rights reserved. |
Altun, Kerem; Barshan, Billur Representing and evaluating ultrasonic maps using active snake contours
and Kohonen's self-organizing feature maps Journal Article AUTONOMOUS ROBOTS, 29 (2), pp. 151-168, 2010, ISSN: 0929-5593. Links | BibTeX @article{ISI:000278810500002,
title = {Representing and evaluating ultrasonic maps using active snake contours
and Kohonen's self-organizing feature maps},
author = {Kerem Altun and Billur Barshan},
doi = {10.1007/s10514-010-9181-4},
issn = {0929-5593},
year = {2010},
date = {2010-08-01},
journal = {AUTONOMOUS ROBOTS},
volume = {29},
number = {2},
pages = {151-168},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
2009
|
Tuncel, Orkun; Altun, Kerem; Barshan, Billur Classifying Human Leg Motions with Uniaxial Piezoelectric Gyroscopes Journal Article SENSORS, 9 (11), pp. 8508-8546, 2009. Links | BibTeX @article{ISI:000272189700009,
title = {Classifying Human Leg Motions with Uniaxial Piezoelectric Gyroscopes},
author = {Orkun Tuncel and Kerem Altun and Billur Barshan},
doi = {10.3390/s91108508},
year = {2009},
date = {2009-11-01},
journal = {SENSORS},
volume = {9},
number = {11},
pages = {8508-8546},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
2008
|
Kiper, Gökhan; Soylemez, Eres; Kisisel, Özgür A U A family of deployable polygons and polyhedra Journal Article MECHANISM AND MACHINE THEORY, 43 (5), pp. 627-640, 2008, ISSN: 0094-114X. Abstract | Links | BibTeX @article{ISI:000255728900007,
title = {A family of deployable polygons and polyhedra},
author = {Gökhan Kiper and Eres Soylemez and Özgür A U Kisisel},
doi = {10.1016/j.mechmachtheory.2007.04.011},
issn = {0094-114X},
year = {2008},
date = {2008-05-01},
journal = {MECHANISM AND MACHINE THEORY},
volume = {43},
number = {5},
pages = {627-640},
abstract = {A new linkage type for resizing polygonal and polyhedral shapes is
proposed. The single degree-of-freedom planar linkages considered mainly
consist of links connected by revolute joints. It is shown that the
group of mechanisms obtained realize Cardanic Motion. The polyhedral
linkages proposed are constructed by implementing the proposed planar
linkages on the faces and interconnecting them by links at the vertices
to retain the solid angles of the polyhedral shape of interest. (C) 2007
Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A new linkage type for resizing polygonal and polyhedral shapes is
proposed. The single degree-of-freedom planar linkages considered mainly
consist of links connected by revolute joints. It is shown that the
group of mechanisms obtained realize Cardanic Motion. The polyhedral
linkages proposed are constructed by implementing the proposed planar
linkages on the faces and interconnecting them by links at the vertices
to retain the solid angles of the polyhedral shape of interest. (C) 2007
Elsevier Ltd. All rights reserved. |
Ozdemir, Serhan Power transmission entropy Journal Article International Journal of Heavy Vehicle Systems, 15 (1), pp. 82–88, 2008. BibTeX @article{ozdemir2008power,
title = {Power transmission entropy},
author = {Serhan Ozdemir},
year = {2008},
date = {2008-01-01},
journal = {International Journal of Heavy Vehicle Systems},
volume = {15},
number = {1},
pages = {82--88},
publisher = {Inderscience Publishers},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
2007
|
Ozdemir, Serhan Measures of uncertainty in power split systems Journal Article MECHANISM AND MACHINE THEORY, 42 (2), pp. 159-167, 2007, ISSN: 0094-114X. Abstract | Links | BibTeX @article{ISI:000243623800003,
title = {Measures of uncertainty in power split systems},
author = {Serhan Ozdemir},
doi = {10.1016/j.mechmachtheory.2006.04.008},
issn = {0094-114X},
year = {2007},
date = {2007-02-01},
journal = {MECHANISM AND MACHINE THEORY},
volume = {42},
number = {2},
pages = {159-167},
abstract = {This paper discusses the overlooked uncertainty inherent in every
transmission. The uncertainty aspect has been often, for the sake of
clarity, ignored. Instead, mechanical transmissions have been
characterized traditionally by their transmission efficacies. It is
known that transmission localities are sources of power loss, depending
on many factors, hence sources of uncertainty. Thus each transmission of
power should not only be designated by a constant of efficiency but also
by an expression of uncertainty, reflecting the probability of
transmission. Furthermore, Shannon's and Renyi's expressions of entropy
have been used to quantify this so-called transmission uncertainty. The
entropy of a transmitting unit is given in these two forms and then
compared. Practical formulations for flow optimization are given. (C)
2006 Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper discusses the overlooked uncertainty inherent in every
transmission. The uncertainty aspect has been often, for the sake of
clarity, ignored. Instead, mechanical transmissions have been
characterized traditionally by their transmission efficacies. It is
known that transmission localities are sources of power loss, depending
on many factors, hence sources of uncertainty. Thus each transmission of
power should not only be designated by a constant of efficiency but also
by an expression of uncertainty, reflecting the probability of
transmission. Furthermore, Shannon's and Renyi's expressions of entropy
have been used to quantify this so-called transmission uncertainty. The
entropy of a transmitting unit is given in these two forms and then
compared. Practical formulations for flow optimization are given. (C)
2006 Elsevier Ltd. All rights reserved. |
2006
|
Alpay, S; Bilir, L; Ozdemir, S; Ozerdem, B Wind speed time series characterization by Hilbert transform Journal Article INTERNATIONAL JOURNAL OF ENERGY RESEARCH, 30 (5), pp. 359-364, 2006, ISSN: 0363-907X. Abstract | Links | BibTeX @article{ISI:000236746700006,
title = {Wind speed time series characterization by Hilbert transform},
author = {S Alpay and L Bilir and S Ozdemir and B Ozerdem},
doi = {10.1002/er.1156},
issn = {0363-907X},
year = {2006},
date = {2006-04-01},
journal = {INTERNATIONAL JOURNAL OF ENERGY RESEARCH},
volume = {30},
number = {5},
pages = {359-364},
abstract = {Predictions of wind energy potential in a given region are based on
on-location observations. The time series of these observations would
later be analysed and modelled either by a probability density function
(pdf) such as a Weibull curve to represent the data or recently by soft
computing techniques, such as neural networks (NNs). In this paper,
discrete Hilbert transform has been applied to characterize the wind
sample data measured on Izmir Institute of Technology campus area which
is located in Urla, Izmir, Turkey, in March 2001 and 2002. By applying
discrete Hilbert transform filter, the instantaneous amplitude, phase
and frequency are found, and characterization of wind speed is
accomplished. Authors have also tried to estimate the hourly wind data
using daily sequence by Hilbert transform technique. Results are
varying. Copyright (c) 2005 John Wiley & Sons, Ltd.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Predictions of wind energy potential in a given region are based on
on-location observations. The time series of these observations would
later be analysed and modelled either by a probability density function
(pdf) such as a Weibull curve to represent the data or recently by soft
computing techniques, such as neural networks (NNs). In this paper,
discrete Hilbert transform has been applied to characterize the wind
sample data measured on Izmir Institute of Technology campus area which
is located in Urla, Izmir, Turkey, in March 2001 and 2002. By applying
discrete Hilbert transform filter, the instantaneous amplitude, phase
and frequency are found, and characterization of wind speed is
accomplished. Authors have also tried to estimate the hourly wind data
using daily sequence by Hilbert transform technique. Results are
varying. Copyright (c) 2005 John Wiley & Sons, Ltd. |
2005
|
Atan, E; Ozdemir, S Intelligence modeling of the transient asperity temperatures in meshing
spur gears Journal Article MECHANISM AND MACHINE THEORY, 40 (1), pp. 119-127, 2005, ISSN: 0094-114X. Abstract | Links | BibTeX @article{ISI:000226442400009,
title = {Intelligence modeling of the transient asperity temperatures in meshing
spur gears},
author = {E Atan and S Ozdemir},
doi = {10.1016/j.mechmachtheory.2004.06.006},
issn = {0094-114X},
year = {2005},
date = {2005-01-01},
journal = {MECHANISM AND MACHINE THEORY},
volume = {40},
number = {1},
pages = {119-127},
abstract = {Temperature rise in the contact zone of meshing gears is a serious
problem in gear design. The temperature rise on lubricated surfaces may
result in the significant decrease on the material strength and
lubricant viscosity which reduces the film thickness, causing solid to
solid contact. The equations and the evaluations of the rise in
temperature were given in [Proc. VDI Berichte 2 (1665) (2002) 615-626]
and reiterated in this paper briefly. The data from [Proc. VDI
Berichte 2 (1665) (2002) 615-626] are used to establish an artificial
intelligence model where a multi layer feedforward neural network has
been employed. The model accepts surface roughness, gear ratio,
horsepower and the number of teeth as input variables, and outputs
calculated pinion surface asperity temperatures. The aim of the present
work is to provide a straightforward and simple way to compute the
asperity temperature rise for a given set of variables. R-square value
for the computed temperature values proves the method satisfactory. (C)
2004 Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Temperature rise in the contact zone of meshing gears is a serious
problem in gear design. The temperature rise on lubricated surfaces may
result in the significant decrease on the material strength and
lubricant viscosity which reduces the film thickness, causing solid to
solid contact. The equations and the evaluations of the rise in
temperature were given in [Proc. VDI Berichte 2 (1665) (2002) 615-626]
and reiterated in this paper briefly. The data from [Proc. VDI
Berichte 2 (1665) (2002) 615-626] are used to establish an artificial
intelligence model where a multi layer feedforward neural network has
been employed. The model accepts surface roughness, gear ratio,
horsepower and the number of teeth as input variables, and outputs
calculated pinion surface asperity temperatures. The aim of the present
work is to provide a straightforward and simple way to compute the
asperity temperature rise for a given set of variables. R-square value
for the computed temperature values proves the method satisfactory. (C)
2004 Elsevier Ltd. All rights reserved. |
Ozdemir, S Local sparse coding control of CVPSTs Journal Article INTERNATIONAL JOURNAL OF HEAVY VEHICLE SYSTEMS, 12 (2), pp. 104-120, 2005, ISSN: 1744-232X. Abstract | Links | BibTeX @article{ISI:000240128000004,
title = {Local sparse coding control of CVPSTs},
author = {S Ozdemir},
doi = {10.1504/IJHVS.2005.006378},
issn = {1744-232X},
year = {2005},
date = {2005-01-01},
journal = {INTERNATIONAL JOURNAL OF HEAVY VEHICLE SYSTEMS},
volume = {12},
number = {2},
pages = {104-120},
abstract = {This paper discusses simulations of a control scheme based on locally
sparse coded networks (CMACs) for a novel previously proposed
continuously variable transmission (CVT), a hybrid continuously variable
power split transmission (CVPST) (Osdemir and Schueller, 2002).
Automotive transmissions match the speed and the torque of the power
source to the speed and torque requirements of the load. Properly
designed CVTs have shown potential to improve efficiency and
performance. The main advantage of CMACs is fast computation because of
their simple operational principles. Simulation results have shown that
memory contents either reach a stable limit cycle or an attractor based
on the selection of network parameters and the training method. Both
online and offline training are possible.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper discusses simulations of a control scheme based on locally
sparse coded networks (CMACs) for a novel previously proposed
continuously variable transmission (CVT), a hybrid continuously variable
power split transmission (CVPST) (Osdemir and Schueller, 2002).
Automotive transmissions match the speed and the torque of the power
source to the speed and torque requirements of the load. Properly
designed CVTs have shown potential to improve efficiency and
performance. The main advantage of CMACs is fast computation because of
their simple operational principles. Simulation results have shown that
memory contents either reach a stable limit cycle or an attractor based
on the selection of network parameters and the training method. Both
online and offline training are possible. |
2004
|
Boke, H; Akkurt, S; Ozdemir, S; Gokturk, EH; Saltik, ENC Quantification of CaCO3-CaSO3 center dot 0.5H(2)O-CaSO4 center dot
2H(2)O mixtures by FTIR analysis and its ANN model Journal Article MATERIALS LETTERS, 58 (5), pp. 723-726, 2004, ISSN: 0167-577X. Abstract | Links | BibTeX @article{ISI:000188216300035,
title = {Quantification of CaCO3-CaSO3 center dot 0.5H(2)O-CaSO4 center dot
2H(2)O mixtures by FTIR analysis and its ANN model},
author = {H Boke and S Akkurt and S Ozdemir and EH Gokturk and ENC Saltik},
doi = {10.1016/j.matlet.2003.07.008},
issn = {0167-577X},
year = {2004},
date = {2004-02-01},
journal = {MATERIALS LETTERS},
volume = {58},
number = {5},
pages = {723-726},
abstract = {A new quantitative analysis method for mixtures of calcium carbonate
(CaCO3), calcium sulphite hemihydrate (CaSO3.1/2H(2)O) and gypsum
(CaSO4.2H(2)O) by FTIR spectroscopy is developed. The method involves
the FTIR analysis of powder mixtures of several compositions on KBr disc
specimens. Intensities of the resulting absorbance peaks for CaCO3,
CaSO3.1/2H(2)O and CaSO4.2H(2)O at 1453, 980, 1146 cm(-1) were used as
input data for an artificial neural network (ANN) model, the output
being the weight percent compositions of the mixtures. The training and
testing data were randomly separated from the complete original data
set. Testing of the model was done with successfully low-average error
levels. The utility of the model is in the potential ability to use FTIR
spectrum to predict the proportions of the three substances in unknown
mixtures. (C) 2003 Elsevier B.V. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A new quantitative analysis method for mixtures of calcium carbonate
(CaCO3), calcium sulphite hemihydrate (CaSO3.1/2H(2)O) and gypsum
(CaSO4.2H(2)O) by FTIR spectroscopy is developed. The method involves
the FTIR analysis of powder mixtures of several compositions on KBr disc
specimens. Intensities of the resulting absorbance peaks for CaCO3,
CaSO3.1/2H(2)O and CaSO4.2H(2)O at 1453, 980, 1146 cm(-1) were used as
input data for an artificial neural network (ANN) model, the output
being the weight percent compositions of the mixtures. The training and
testing data were randomly separated from the complete original data
set. Testing of the model was done with successfully low-average error
levels. The utility of the model is in the potential ability to use FTIR
spectrum to predict the proportions of the three substances in unknown
mixtures. (C) 2003 Elsevier B.V. All rights reserved. |
2003
|
Tayfur, G; Ozdemir, S; Singh, VP Fuzzy logic algorithm for runoff-induced sediment transport from bare
soil surfaces Journal Article ADVANCES IN WATER RESOURCES, 26 (12), pp. 1249-1256, 2003, ISSN: 0309-1708. Abstract | Links | BibTeX @article{ISI:000186662500004,
title = {Fuzzy logic algorithm for runoff-induced sediment transport from bare
soil surfaces},
author = {G Tayfur and S Ozdemir and VP Singh},
doi = {10.1016/j.advwatres.2003.08.005},
issn = {0309-1708},
year = {2003},
date = {2003-12-01},
journal = {ADVANCES IN WATER RESOURCES},
volume = {26},
number = {12},
pages = {1249-1256},
abstract = {Utilizing the rainfall intensity, and slope data, a fuzzy logic
algorithm was developed to estimate sediment loads from bare soil
surfaces. Considering slope and rainfall as input variables, the
variables were fuzzified into fuzzy subsets. The fuzzy subsets of the
variables were considered to have triangular membership functions. The
relations among rainfall intensity, slope, and sediment transport were
represented by a set of fuzzy rules. The fuzzy rules relating input
variables to the output variable of sediment discharge were laid out in
the IF-THEN format. The commonly used weighted average method was
employed for the defuzzification procedure.
The sediment load predicted by the fuzzy model was in satisfactory
agreement with the measured sediment load data. Predicting the mean
sediment loads from experimental runs, the performance of the fuzzy
model was compared with that of the artificial neural networks (ANNs)
and the physics-based models. The results of showed revealed that the
fuzzy model performed better under very high rainfall intensities over
different slopes and over very steep slopes under different rainfall
intensities. This is closely related to the selection of the shape and
frequency of the fuzzy membership functions in the fuzzy model. (C) 2003
Elsevier Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Utilizing the rainfall intensity, and slope data, a fuzzy logic
algorithm was developed to estimate sediment loads from bare soil
surfaces. Considering slope and rainfall as input variables, the
variables were fuzzified into fuzzy subsets. The fuzzy subsets of the
variables were considered to have triangular membership functions. The
relations among rainfall intensity, slope, and sediment transport were
represented by a set of fuzzy rules. The fuzzy rules relating input
variables to the output variable of sediment discharge were laid out in
the IF-THEN format. The commonly used weighted average method was
employed for the defuzzification procedure.
The sediment load predicted by the fuzzy model was in satisfactory
agreement with the measured sediment load data. Predicting the mean
sediment loads from experimental runs, the performance of the fuzzy
model was compared with that of the artificial neural networks (ANNs)
and the physics-based models. The results of showed revealed that the
fuzzy model performed better under very high rainfall intensities over
different slopes and over very steep slopes under different rainfall
intensities. This is closely related to the selection of the shape and
frequency of the fuzzy membership functions in the fuzzy model. (C) 2003
Elsevier Ltd. All rights reserved. |
Akkurt, S; Ozdemir, S; Tayfur, G; Akyol, B The use of GA-ANNs in the modelling of compressive strength of cement
mortar Journal Article CEMENT AND CONCRETE RESEARCH, 33 (7), pp. 973-979, 2003, ISSN: 0008-8846. Abstract | Links | BibTeX @article{ISI:000183190600005,
title = {The use of GA-ANNs in the modelling of compressive strength of cement
mortar},
author = {S Akkurt and S Ozdemir and G Tayfur and B Akyol},
doi = {10.1016/S0008-8846(03)00006-1},
issn = {0008-8846},
year = {2003},
date = {2003-07-01},
journal = {CEMENT AND CONCRETE RESEARCH},
volume = {33},
number = {7},
pages = {973-979},
abstract = {In this paper, results of a project aimed at modelling the compressive
strength of cement mortar under standard curing conditions are reported.
Plant data were collected for 6 months for the chemical and physical
properties of the cement that were used in model construction and
testing. The training and testing data were separated from the complete
original data set by the use of genetic algorithms (GAs). A
GA-artificial neural network (ANN) model based on the training data of
the cement strength was created. Testing of the model was also done
within low average error levels (2.24%). The model was subjected to
sensitivity analysis to predict the response of the system to different
values of the factors affecting the strength. The plots obtained after
sensitivity analysis indicated that increasing the amount Of C3S, SO3
and surface area led to increased strength within the limits of the
model. C2S decreased the strength whereas C(3)A decreased or increased
the strength depending on the SO3 level. Because of the limited data
range used for training, the prediction results were good only within
the same range. The utility of the model is in the potential ability to
control processing parameters to yield the desired strength levels and
in providing information regarding the most favourable experimental
conditions to obtain maximum compressive strength. (C) 2003 Elsevier
Science Ltd. All rights reserved.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In this paper, results of a project aimed at modelling the compressive
strength of cement mortar under standard curing conditions are reported.
Plant data were collected for 6 months for the chemical and physical
properties of the cement that were used in model construction and
testing. The training and testing data were separated from the complete
original data set by the use of genetic algorithms (GAs). A
GA-artificial neural network (ANN) model based on the training data of
the cement strength was created. Testing of the model was also done
within low average error levels (2.24%). The model was subjected to
sensitivity analysis to predict the response of the system to different
values of the factors affecting the strength. The plots obtained after
sensitivity analysis indicated that increasing the amount Of C3S, SO3
and surface area led to increased strength within the limits of the
model. C2S decreased the strength whereas C(3)A decreased or increased
the strength depending on the SO3 level. Because of the limited data
range used for training, the prediction results were good only within
the same range. The utility of the model is in the potential ability to
control processing parameters to yield the desired strength levels and
in providing information regarding the most favourable experimental
conditions to obtain maximum compressive strength. (C) 2003 Elsevier
Science Ltd. All rights reserved. |
