Abstract: The author previously proposed an extension of differential evolution. The proposed method extends the processes of DE to handle interval numbers as genotype values so that DE can be applied to interval-valued optimization problems. The interval DE can employ either of two interval models, the lower and upper model or the center and width model, for specifying genotype values. Ability of the interval DE in searching for solutions may depend on the model. In this paper, the author compares the two models to investigate which model contributes better for the interval DE to find better solutions. Application of the interval DE is evolutionary training of interval-valued neural networks. A result of preliminary study indicates that the CW model is better than the LU model: the interval DE with the CW model could evolve better neural networks.
Abstract: By using the estimation of the Cauchy matrix of linear impulsive differential equations and Banach fixed point theorem as well as Gronwall-Bellman’s inequality, some sufficient conditions are obtained for the existence and exponential stability of almost periodic
solution for an impulsive neural networks with distributed delays. An example is presented to illustrate the feasibility and effectiveness of the results.
Abstract: In this paper, the exponential stability of periodic solutions in inertial neural networks with unbounded delay are investigated. First, using variable substitution the system is transformed to first order differential equation. Second, by the fixed-point theorem and constructing suitable Lyapunov function, some sufficient conditions guaranteeing the existence and exponential stability of periodic solutions of the system are obtained. Finally, two examples are given to illustrate the effectiveness of the results.
Abstract: Learning the gradient of neuron's activity function
like the weight of links causes a new specification which is
flexibility. In flexible neural networks because of supervising and
controlling the operation of neurons, all the burden of the learning is
not dedicated to the weight of links, therefore in each period of
learning of each neuron, in fact the gradient of their activity function,
cooperate in order to achieve the goal of learning thus the number of
learning will be decreased considerably.
Furthermore, learning neurons parameters immunes them against
changing in their inputs and factors which cause such changing.
Likewise initial selecting of weights, type of activity function,
selecting the initial gradient of activity function and selecting a fixed
amount which is multiplied by gradient of error to calculate the
weight changes and gradient of activity function, has a direct affect
in convergence of network for learning.
Abstract: Short-Term Load Forecasting (STLF) plays an important role for the economic and secure operation of power systems. In this paper, Continuous Genetic Algorithm (CGA) is employed to evolve the optimum large neural networks structure and connecting weights for one-day ahead electric load forecasting problem. This study describes the process of developing three layer feed-forward large neural networks for load forecasting and then presents a heuristic search algorithm for performing an important task of this process, i.e. optimal networks structure design. The proposed method is applied to STLF of the local utility. Data are clustered due to the differences in their characteristics. Special days are extracted from the normal training sets and handled separately. In this way, a solution is provided for all load types, including working days and weekends and special days. We find good performance for the large neural networks. The proposed methodology gives lower percent errors all the time. Thus, it can be applied to automatically design an optimal load forecaster based on historical data.
Abstract: An adaptive dynamic cerebellar model articulation
controller (DCMAC) neural network used for solving the prediction
and identification problem is proposed in this paper. The proposed
DCMAC has superior capability to the conventional cerebellar model
articulation controller (CMAC) neural network in efficient learning
mechanism, guaranteed system stability and dynamic response. The
recurrent network is embedded in the DCMAC by adding feedback
connections in the association memory space so that the DCMAC
captures the dynamic response, where the feedback units act as
memory elements. The dynamic gradient descent method is adopted to
adjust DCMAC parameters on-line. Moreover, the analytical method
based on a Lyapunov function is proposed to determine the
learning-rates of DCMAC so that the variable optimal learning-rates
are derived to achieve most rapid convergence of identifying error.
Finally, the adaptive DCMAC is applied in two computer simulations.
Simulation results show that accurate identifying response and
superior dynamic performance can be obtained because of the
powerful on-line learning capability of the proposed DCMAC.
Abstract: The use of neural networks for recognition application is generally constrained by their inherent parameters inflexibility after the training phase. This means no adaptation is accommodated for input variations that have any influence on the network parameters. Attempts were made in this work to design a neural network that includes an additional mechanism that adjusts the threshold values according to the input pattern variations. The new approach is based on splitting the whole network into two subnets; main traditional net and a supportive net. The first deals with the required output of trained patterns with predefined settings, while the second tolerates output generation dynamically with tuning capability for any newly applied input. This tuning comes in the form of an adjustment to the threshold values. Two levels of supportive net were studied; one implements an extended additional layer with adjustable neuronal threshold setting mechanism, while the second implements an auxiliary net with traditional architecture performs dynamic adjustment to the threshold value of the main net that is constructed in dual-layer architecture. Experiment results and analysis of the proposed designs have given quite satisfactory conducts. The supportive layer approach achieved over 90% recognition rate, while the multiple network technique shows more effective and acceptable level of recognition. However, this is achieved at the price of network complexity and computation time. Recognition generalization may be also improved by accommodating capabilities involving all the innate structures in conjugation with Intelligence abilities with the needs of further advanced learning phases.
Abstract: Artificial neural networks (ANN) have the ability to model input-output relationships from processing raw data. This characteristic makes them invaluable in industry domains where such knowledge is scarce at best. In the recent decades, in order to overcome the black-box characteristic of ANNs, researchers have attempted to extract the knowledge embedded within ANNs in the form of rules that can be used in inference systems. This paper presents a new technique that is able to extract a small set of rules from a two-layer ANN. The extracted rules yield high classification accuracy when implemented within a fuzzy inference system. The technique targets industry domains that possess less complex problems for which no expert knowledge exists and for which a simpler solution is preferred to a complex one. The proposed technique is more efficient, simple, and applicable than most of the previously proposed techniques.
Abstract: A novel application of neural network approach to
fault classification and fault location of Medium voltage cables is
demonstrated in this paper. Different faults on a protected cable
should be classified and located correctly. This paper presents the use
of neural networks as a pattern classifier algorithm to perform these
tasks. The proposed scheme is insensitive to variation of different
parameters such as fault type, fault resistance, and fault inception
angle. Studies show that the proposed technique is able to offer high
accuracy in both of the fault classification and fault location tasks.
Abstract: This paper presents an algorithm based on the
wavelet decomposition, for feature extraction from the ECG signal
and recognition of three types of Ventricular Arrhythmias using
neural networks. A set of Discrete Wavelet Transform (DWT)
coefficients, which contain the maximum information about the
arrhythmias, is selected from the wavelet decomposition. After that a
novel clustering algorithm based on nature inspired algorithm (Ant
Colony Optimization) is developed for classifying arrhythmia types.
The algorithm is applied on the ECG registrations from the MIT-BIH
arrhythmia and malignant ventricular arrhythmia databases. We
applied Daubechies 4 wavelet in our algorithm. The wavelet
decomposition enabled us to perform the task efficiently and
produced reliable results.
Abstract: This paper proposed classification models that would
be used as a proxy for hard disk drive (HDD) functional test equitant
which required approximately more than two weeks to perform the
HDD status classification in either “Pass" or “Fail". These models
were constructed by using committee network which consisted of a
number of single neural networks. This paper also included the
method to solve the problem of sparseness data in failed part, which
was called “enforce learning method". Our results reveal that the
constructed classification models with the proposed method could
perform well in the sparse data conditions and thus the models,
which used a few seconds for HDD classification, could be used to
substitute the HDD functional tests.
Abstract: HIV-1 genome is highly heterogeneous. Due to this
variation, features of HIV-I genome is in a wide range. For this
reason, the ability to infection of the virus changes depending on
different chemokine receptors. From this point of view, R5 HIV
viruses use CCR5 coreceptor while X4 viruses use CXCR5 and
R5X4 viruses can utilize both coreceptors. Recently, in
Bioinformatics, R5X4 viruses have been studied to classify by using
the experiments on HIV-1 genome.
In this study, R5X4 type of HIV viruses were classified using
Auto Regressive (AR) model through Artificial Neural Networks
(ANNs). The statistical data of R5X4, R5 and X4 viruses was
analyzed by using signal processing methods and ANNs. Accessible
residues of these virus sequences were obtained and modeled by AR
model since the dimension of residues is large and different from
each other. Finally the pre-processed data was used to evolve various
ANN structures for determining R5X4 viruses. Furthermore ROC
analysis was applied to ANNs to show their real performances. The
results indicate that R5X4 viruses successfully classified with high
sensitivity and specificity values training and testing ROC analysis
for RBF, which gives the best performance among ANN structures.
Abstract: this paper presents an auto-regressive network called the Auto-Regressive Multi-Context Recurrent Neural Network (ARMCRN), which forecasts the daily peak load for two large power plant systems. The auto-regressive network is a combination of both recurrent and non-recurrent networks. Weather component variables are the key elements in forecasting because any change in these variables affects the demand of energy load. So the AR-MCRN is used to learn the relationship between past, previous, and future exogenous and endogenous variables. Experimental results show that using the change in weather components and the change that occurred in past load as inputs to the AR-MCRN, rather than the basic weather parameters and past load itself as inputs to the same network, produce higher accuracy of predicted load. Experimental results also show that using exogenous and endogenous variables as inputs is better than using only the exogenous variables as inputs to the network.
Abstract: In this paper, a class of impulsive BAM fuzzy cellular neural networks with time delays in the leakage terms is formulated and investigated. By establishing a delay differential inequality and M-matrix theory, some sufficient conditions ensuring the existence, uniqueness and global exponential stability of equilibrium point for impulsive BAM fuzzy cellular neural networks with time delays in the leakage terms are obtained. In particular, a precise estimate of the exponential convergence rate is also provided, which depends on system parameters and impulsive perturbation intention. It is believed that these results are significant and useful for the design and applications of BAM fuzzy cellular neural networks. An example is given to show the effectiveness of the results obtained here.
Abstract: Water level forecasting using records of past time series is of importance in water resources engineering and management. For example, water level affects groundwater tables in low-lying coastal areas, as well as hydrological regimes of some coastal rivers. Then, a reliable prediction of sea-level variations is required in coastal engineering and hydrologic studies. During the past two decades, the approaches based on the Genetic Programming (GP) and Artificial Neural Networks (ANN) were developed. In the present study, the GP is used to forecast daily water level variations for a set of time intervals using observed water levels. The measurements from a single tide gauge at Urmia Lake, Northwest Iran, were used to train and validate the GP approach for the period from January 1997 to July 2008. Statistics, the root mean square error and correlation coefficient, are used to verify model by comparing with a corresponding outputs from Artificial Neural Network model. The results show that both these artificial intelligence methodologies are satisfactory and can be considered as alternatives to the conventional harmonic analysis.
Abstract: In neural networks, when new patterns are learned by a network, the new information radically interferes with previously stored patterns. This drawback is called catastrophic forgetting or catastrophic interference. In this paper, we propose a biologically inspired neural network model which overcomes this problem. The proposed model consists of two distinct networks: one is a Hopfield type of chaotic associative memory and the other is a multilayer neural network. We consider that these networks correspond to the hippocampus and the neocortex of the brain, respectively. Information given is firstly stored in the hippocampal network with fast learning algorithm. Then the stored information is recalled by chaotic behavior of each neuron in the hippocampal network. Finally, it is consolidated in the neocortical network by using pseudopatterns. Computer simulation results show that the proposed model has much better ability to avoid catastrophic forgetting in comparison with conventional models.
Abstract: One of the major parts of a jet engine is air intake,
which provides proper and required amount of air for the engine to
operate. There are several aerodynamic parameters which should be
considered in design, such as distortion, pressure recovery, etc. In
this research, the effects of lip ice accretion on pitot intake
performance are investigated. For ice accretion phenomenon, two
supervised multilayer neural networks (ANN) are designed, one for
ice shape prediction and another one for ice roughness estimation
based on experimental data. The Fourier coefficients of transformed
ice shape and parameters include velocity, liquid water content
(LWC), median volumetric diameter (MVD), spray time and
temperature are used in neural network training. Then, the subsonic
intake flow field is simulated numerically using 2D Navier-Stokes
equations and Finite Volume approach with Hybrid mesh includes
structured and unstructured meshes. The results are obtained in
different angles of attack and the variations of intake aerodynamic
parameters due to icing phenomenon are discussed. The results show
noticeable effects of ice accretion phenomenon on intake behavior.
Abstract: This paper describes the use of artificial neural
networks (ANN) for predicting non-linear layer moduli of flexible
airfield pavements subjected to new generation aircraft (NGA)
loading, based on the deflection profiles obtained from Heavy
Weight Deflectometer (HWD) test data. The HWD test is one of the
most widely used tests for routinely assessing the structural integrity
of airport pavements in a non-destructive manner. The elastic moduli
of the individual pavement layers backcalculated from the HWD
deflection profiles are effective indicators of layer condition and are
used for estimating the pavement remaining life. HWD tests were
periodically conducted at the Federal Aviation Administration-s
(FAA-s) National Airport Pavement Test Facility (NAPTF) to
monitor the effect of Boeing 777 (B777) and Beoing 747 (B747) test
gear trafficking on the structural condition of flexible pavement
sections. In this study, a multi-layer, feed-forward network which
uses an error-backpropagation algorithm was trained to approximate
the HWD backcalculation function. The synthetic database generated
using an advanced non-linear pavement finite-element program was
used to train the ANN to overcome the limitations associated with
conventional pavement moduli backcalculation. The changes in
ANN-based backcalculated pavement moduli with trafficking were
used to compare the relative severity effects of the aircraft landing
gears on the NAPTF test pavements.
Abstract: An artificial neural network (ANN) approach was used to model the energy consumption of wheat production. This study was conducted over 35,300 hectares of irrigated and dry land wheat fields in Canterbury in the 2007-2008 harvest year.1 In this study several direct and indirect factors have been used to create an artificial neural networks model to predict energy use in wheat production. The final model can predict energy consumption by using farm condition (size of wheat area and number paddocks), farmers- social properties (education), and energy inputs (N and P use, fungicide consumption, seed consumption, and irrigation frequency), it can also predict energy use in Canterbury wheat farms with error margin of ±7% (± 1600 MJ/ha).
Abstract: .Hardware realization of a Neural Network (NN), to a large extent depends on the efficient implementation of a single neuron. FPGA-based reconfigurable computing architectures are suitable for hardware implementation of neural networks. FPGA realization of ANNs with a large number of neurons is still a challenging task. This paper discusses the issues involved in implementation of a multi-input neuron with linear/nonlinear excitation functions using FPGA. Implementation method with resource/speed tradeoff is proposed to handle signed decimal numbers. The VHDL coding developed is tested using Xilinx XC V50hq240 Chip. To improve the speed of operation a lookup table method is used. The problems involved in using a lookup table (LUT) for a nonlinear function is discussed. The percentage saving in resource and the improvement in speed with an LUT for a neuron is reported. An attempt is also made to derive a generalized formula for a multi-input neuron that facilitates to estimate approximately the total resource requirement and speed achievable for a given multilayer neural network. This facilitates the designer to choose the FPGA capacity for a given application. Using the proposed method of implementation a neural network based application, namely, a Space vector modulator for a vector-controlled drive is presented