Error Back Propagation Learning Algorithm Research Articles

Continuous–Discrete Time Neural Network Observer for Nonlinear Dynamic Systems Application to Vehicle Systems

This paper proposes a novel continuous-discrete (sampled data) time neural network (NSNN) observer for nonlinear systems. It can therefore be applied to systems with a high degree of non-linearity with no prior knowledge of the system dynamics. The proposed observer is a three-layer feedforward neural network that has been intensively trained using the error backpropagation learning algorithm, which includes an e-modification term to ensure robustness of the observer. A structure of the output predictor with a corrective term is added in the structure of the NN observer to overcome the problem of discrete time measurement. Simulations using MATLAB and CarSim are illustrated to demonstrate the performance of the proposed state observer strategy to reconstruct the state variables and parameters of a vehicle system.

Research on Financial Early Warning Based on Combination Forecasting Model

Since entering the 21st century, “economic globalization” has become a hot topic. Under the impact of “economic globalization”, the competition of the Chinese domestic market continues to intensify, and Chinese enterprises are facing enormous pressure for survival and development. Among them, there are many cases of poor business operation caused by financial crisis which have directly put these companies in trouble, even causing them to go bankrupt. Therefore, it is very practical to establish a scientific data model to analyze and predict the financial data of enterprises. It can not only monitor the financial status of the enterprise in real time, but also play an effective financial early warning role. This research focuses on using the combined forecasting method to establish a more comprehensive financial early warning model to solve the related financial crisis forecasting problem. Specifically, two different forecasting methods are first adopted in this study to conduct financial early warning research. The first is time series forecasting. It is a dynamic data processing statistical method, which is often used in forecasting research in the business field. The second is the BP neural network algorithm (referred to as BP), which is an error back-propagation learning algorithm, which is often used in the field of artificial intelligence. Then, the prediction error values of the two methods are compared and they are applied to the combined prediction method. Finally, a new error prediction formula is obtained. The result shows that the BP method provides the best performance over others, while the combinational forecasting method offers better performance than any single method.

Application of Reference Voltage Control Method of the Generator Using a Neural Network in Variable Speed Synchronous Generation System of DC Distribution for Ships

In the case of DC power distribution-based variable speed engine synchronous generators, if the output reference voltage is kept constant regardless of the generator engine operating speed, it may cause damage to the internal device and windings of the generator due to over-flux or over-excitation. The purpose of this study is to adjust the generator reference voltage according to the engine speed change in the DC distribution system with the variable speed engine synchronous generator. A method of controlling the generator reference voltage according to the speed was applied by adjusting the value of the variable resistance input to the external terminal of the automatic voltage regulator using a neural network controller. The learning data of the neural network was measured through an experiment, and the input pattern was set as the rotational speed of the generator engine, and the output pattern was set as the input current of the potentiometer. Using the measured input/output pattern of the neural network, the error backpropagation learning algorithm was applied to derive the optimum connection weight to be applied to the controller. For the test, the variable speed operation range of the generator engine was set to 1100–1800 rpm, and the input current value of the potentiometer according to the speed increase or decrease within the operation range and the output of the voltage output from the actual generator were checked. As a result of neural network control, it was possible to confirm the result that the input current value of the potentiometer accurately reached the target value 4–20 mA at the point where the initial speed change occurred. It was confirmed that the reference voltage was also normally output in the target range of 250–440 V.

Prediction of cutting forces in drilling AL6082-T6 by using artificial neural networks

This research investigates the comparison of the thrust force (Fz) and cutting torque (Mz) predictive models based on artificial neural networks (ANN). The models were developed based on three-level full factorial design of experiments conducted when drilling of Al6082-T6 alloy material with cutting speed, feed rate and tool diameter as the process parameters. The predictive ANN models of thrust force (Fz) and cutting torque (Mz) were developed using a multilayer feed forward neural network, trained using an error back propagation learning algorithm. The developed models are shown capable of predicting them (Fz and Mz) to a great level. The confirmation experiments provided favorable results with accuracy of 4% and 4.5%, for the thrust force (Fz) and the cutting torque (Mz) respectively. Furthermore the R coefficient for the prediction model of the thrust force is 0.99608 and 0.9946 for the cutting torque. As a result they can be considered as very accurate and appropriate for their prediction. This study promotes the better machining quality and the sustainability of the process that are achieved at the same time

Prediction of visual outcomes by an artificial neural network following intravitreal injection and laser therapy for retinopathy of prematurity

AimsTo construct a program to predict the visual acuity (VA), best corrected VA (BCVA) and spherical equivalent (SE) of patients with retinopathy of prematurity (ROP) from 3 to 12 years...

Classification of Arrhythmia using Wavelet Transform and Neural Network Model

Cardiovascular diseases are a major cause of death. Change in normal human heart beat may result in different types of cardiac arrhythmias. An Irreversible damage to the heart is possible. In this paper a method is proposed to classify different arrhythmias and normal sinus rhythm, through a combination of wavelet Transform and Artificial Neural Networks (ANN) accurately and efficiently. Adaptive filtering using Recursive Least squares (RLS) adaptive algorithm is utilized to nullify AC and DC noises from the sample ECG signal set. ECG data’s are collected from MITBIH database. As ECG signal is a non- stationary signal wavelet transform is used to decompose the signal at various resolutions. This allows accurate detection and extraction of features. In our approach, discrete wavelet transforms (DWT) coefficients set is obtained from wavelet decomposition which would contain the maximum information about the arrhythmia. RR interval, QRS duration, PR duration is extracted from the wavelet decomposition. With these parameters classification of arrhythmia is done. Multilayer feed forward ANNs employ error back propagation (EBP) learning algorithm were trained and tested using the extracted parameters are used for training and testing the error back propagation (EBP) algorithm. Multilayer feed forward ANNs are employed through this EBP learning algorithm. This classification is done for 84 patient samples. The overall accuracy of our approach is 98.8%.

Fractional neural observer design for a class of nonlinear fractional chaotic systems

In this paper, a novel observer structure for nonlinear fractional-order systems is presented to estimate the states of fractional-order nonlinear chaotic system with unknown dynamical model. A new fractional error back-propagation learning algorithm is derived to adapt weights of the artificial neural network, by taking advantage of the Lyapunov stability strategy of fractional-order systems which is called Miattag–Leffler stability. The main contribution is the extension of neural observer for fractional dynamics in a way that satisfies Miattag–Leffler conditions. Observer design procedure guarantees the convergence of observer error to the neighborhood of zero. Furthermore, the robustness of the proposed estimator against uncertainties and external disturbances are the main benefits of the proposed method. Simulation results show the effectiveness and capabilities of the proposed observer.

Artificial neural network model for evaluation the effect of surface properties amendment on slurry erosion behavior of AISI 5117 steel

Purpose The aim of this paper is to evaluate the effect of surface treatment on slurry erosion behavior of AISI 5,117 steel using artificial neural network (ANN) technique. Design/methodology/approach The slurry erosion wear behavior of electroless nickel-phosphorus (Ni-P) coated, carburized and untreated AISI 5,117 alloy steel was investigated experimentally and theoretically using ANN technique based on error back propagation learning algorithm. Findings From the obtained results, it can be concluded that the proposed AAN model can be successfully used for evaluating slurry erosion behavior of the Ni-P coated, carburized and untreated AISI 5,117 steel for wide range of operating conditions and Ni-P coating and carburizing improve the slurry erosion resistance of AISI 5,117 steel; however, the coating is more efficient. Originality/value Slurry erosion is a serious problem for the performance, reliability and service life of engineering components used in many industrial applications. To improve the performance of these components, engineering surface technologies have been attracting a great deal of attention. The extent of slurry erosion is dependent on a wide range of variables. To account all variables that effect on erosion behavior, prediction of erosion behavior by soft computational technique is one of the most important requirements. ANN has the ability to tackle the problem of complex relationships among variables that cannot be accomplished by traditional analytical methods.

Classification of healthy and mastitis Murrah buffaloes by application of neural network models using yield and milk quality parameters

This paper describes a Neural Network (NN) model to classify healthy and mastitis Murrah buffaloes using pH, electrical conductivity, dielectric constant and yield of milk as input parameters and California Mastitis Test (CMT) score as the output parameter. The purpose of this study was to develop such a cost-effective and intelligent classification model, which would serve an alternative to the prevailing Somatic Cell Count (SCC) based techniques to detect mastitis in Murrah buffaloes, because the latter techniques are sophisticated, lengthy and time consuming as well as necessary instruments for carrying out the tests are not, generally, available at the grassroots level or to the small dairy holders. Accordingly, a total of 534 milk samples were collected from 100 lactating Murrah buffaloes, which were scrutinized for mastitis using CMT. The animals were classified into three categories, i.e., healthy, subclinical and clinical mastitis buffaloes and assigned CMT scores as 1, 2, and 3, respectively. The NN models were based on error back propagation learning algorithm with Bayesian regularization mechanism and various combinations of internal parameters. The performance of NN models was compared with that of conventional Multiple Linear Regression (MLR) models also developed in this study. The classification accuracy achieved by the best NN model was 8.02 Root Mean Square percent error (%RMS) while that attained by MLR model was 26.47 %RMS. Further, for classifying healthy vs. subclinical mastitis Murrah buffaloes, sensitivity, specificity and Diagnostic Odds Ratio (DOR) with the best NN model was found to be 98%, 97.72% and 54.87, respectively, having Area under Relative Operating Characteristic (ROC) Curve (AUC) as 0.96 vis-à-vis MLR model attaining the same as 58.87%, 76.72%, and 52.26, respectively, and AUC as 0.81. In case of classifying healthy vs. clinical mastitis Murrah buffaloes, the best NN model achieved sensitivity, specificity and DOR as 99%, 97.28% and 57.92, respectively, with AUC as 0.98 while that with MLR model were determined as 69.23%, 78.20% and 55.46, respectively, and AUC as 0.87. Evidently, the NN model outperformed classical MLR model, in this study. Hence, it can be deduced that NN paradigm has potential to efficiently detect healthy and mastitis Murrah buffaloes on the basis of milk yield and milk quality parameters.

Artificial neural networks employment in the prediction of evapotranspiration of greenhouse-grown sweet pepper

ABSTRACT This study aimed to investigate the applicability of artificial neural networks (ANNs) in the prediction of evapotranspiration of sweet pepper cultivated in a greenhouse. The used data encompass the second crop cycle, from September 2013 to February 2014, constituting 135 days of daily meteorological data, referring to the following variables: temperature and relative air humidity, wind speed and solar radiation (input variables), as well as evapotranspiration (output variable), determined using data obtained by load-cell weighing lysimeter. The recorded data were divided into three sets for training, testing and validation. The ANN learning model recognized the evapotranspiration patterns with acceptable accuracy, with mean square error of 0.005, in comparison to the data recorded in the lysimeter, with coefficient of determination of 0.87, demonstrating the best approximation for the 4-21-1 network architecture, with multilayers, error back-propagation learning algorithm and learning rate of 0.01.

A Comparative Study of the RSM and ANN Models for Predicting Surface Roughness in Roller Burnishing

In this paper the comparison of the surface roughness prediction models based on response surface methodology (RSM) and artificial neural networks (ANN) is described. The models were developed based on five-level design of experiments conducted on Aluminum alloy 6061 work material with spindle speed, interference, feed, and number of tool pass as the roller burnishing process parameters. The ANN predictive models of surface roughness was developed using a multilayer feed forward neural network and trained with the help of an error back propagation learning algorithm based on the generalized delta rule. Mathematical models of second order RSM and developed ANN models were compared for surface roughness. The comparison evidently indicates that the prediction capabilities of ANN models are far better as compared to the RSM models. The minutiae of experimentation, development of model, testing, and performance comparison are presented in the paper.

Artificial neural network models for depicting mass flow rate of R22, R407C and R410A through electronic expansion valves

The utilization of electronic expansion valves (EEVs) in refrigeration and air conditioning systems is increased for energy saving and comfort environmental. However, experimental data and refrigerant mass flow models through EEVs are very limited in open literature. In this study, a new technique using artificial neural network (ANN) is applied to depict the mass flow rates of R22 and its alternatives R407C and R410A flowing through EEVs based on the error back propagation learning algorithm. Two strategies are followed; the first is to construct individual ANN models for each refrigerant, and the second is to construct a generalized ANN model for the three investigated refrigerants. The experimental results from open literature are used to construct the ANN models. The ANN models results showed a good agreement with the corresponding experimental data. For individual models, the relative deviations for R22, R407C, and R410A are within ±0.7%, ±1.1%, and ±0.006%, respectively. While for generalized model, the relative deviations are within ±2.5%. Also the generalized model was tested out of its construction range in a predictive mode and it was found to be a reliable tool to estimate the refrigerants mass flow rates.

Development of a Robust Identifier for NPPs Transients Combining ARIMA Model and EBP Algorithm

This study introduces a novel identification method for recognition of nuclear power plants (NPPs) transients by combining the autoregressive integrated moving-average (ARIMA) model and the neural network with error backpropagation (EBP) learning algorithm. The proposed method consists of three steps. First, an EBP based identifier is adopted to distinguish the plant normal states from the faulty ones. In the second step, ARIMA models use integrated (I) process to convert non-stationary data of the selected variables into stationary ones. Subsequently, ARIMA processes, including autoregressive (AR), moving-average (MA), or autoregressive moving-average (ARMA) are used to forecast time series of the selected plant variables. In the third step, for identification the type of transients, the forecasted time series are fed to the modular identifier which has been developed using the latest advances of EBP learning algorithm. Bushehr nuclear power plant (BNPP) transients are probed to analyze the ability of the proposed identifier. Recognition of transient is based on similarity of its statistical properties to the reference one, rather than the values of input patterns. More robustness against noisy data and improvement balance between memorization and generalization are salient advantages of the proposed identifier. Reduction of false identification, sole dependency of identification on the sign of each output signal, selection of the plant variables for transients training independent of each other, and extendibility for identification of more transients without unfavorable effects are other merits of the proposed identifier.

Automatic Object Recognition from Satellite Images using Artificial Neural Network

Object recognition from satellite images is a very important application for various purposes. Objects can be recognized based on certain features and then applying some algorithm to extract those objects. Basically object recognition is a classification problem. For various remote sensing applications, waterbody acts as an important object which needs to be extracted. It is wise and better if possible, to extract waterbody object automatically from satellite data without any human intervention. This can be achieved using machine learning techniques. Artificial Neural Network (ANN) is such technique which makes machine intelligent by providing learning to it. This intelligent machine can extract objects automatically. This paper presents a methodology to extract waterbody object from satellite data in an automatic manner with the help of ANN. Training and testing dataset have been created by a domain expert which then have been used to train Multi Layer Perceptron (MLP) using Error Back Propagation (EBP) learning algorithm. Confusion matrix and Kappa coefficient have been used for accuracy assessment. General Terms Object Recognition, Satellite Images, Machine Learning, Image Classification, Feature Extraction, Kappa Coefficient, Confusion Matrix, Training and Testing Datasets

Development of an Efficient Identifier for Nuclear Power Plant Transients Based on Latest Advances of Error Back-Propagation Learning Algorithm

This study aims to improve the performance of nuclear power plants (NPPs) transients training and identification using the latest advances of error back-propagation (EBP) learning algorithm. To this end, elements of EBP, including input data, initial weights, learning rate, cost function, activation function, and weights updating procedure are investigated and an efficient neural network is developed. Usefulness of modular networks is also examined and appropriate identifiers, one for each transient, are employed. Furthermore, the effect of transient type on transient identifier performance is illustrated. Subsequently, the developed transient identifier is applied to Bushehr nuclear power plant (BNPP). Seven types of the plant events are probed to analyze the ability of the proposed identifier. The results reveal that identification occurs very early with only five plant variables, whilst in the previous studies a larger number of variables (typically 15 to 20) were required. Modular networks facilitated identification due to its sole dependency on the sign of each network output signal. Fast training of input patterns, extendibility for identification of more transients and reduction of false identification are other advantageous of the proposed identifier. Finally, the balance between the correct answer to the trained transients (memorization) and reasonable response to the test transients (generalization) is improved, meeting one of the primary design criteria of identifiers.

Artificial Neural Network-Based Automated ECG Signal Classifier

The ECG signal is well known for its nonlinear dynamic behavior and a key characteristic that is utilized in this research; the nonlinear component of its dynamics changes more significantly between normal and abnormal conditions than does the linear one. As the higher-order statistics (HOS) preserve phase information, this study makes use of one-dimensional slices from the higher-order spectral domain of normal and ischemic subjects. A feedforward multilayer neural network (NN) with error back-propagation (BP) learning algorithm was used as an automated ECG classifier to investigate the possibility of recognizing ischemic heart disease from normal ECG signals. Different NN structures are tested using two data sets extracted from polyspectrum slices and polycoherence indices of the ECG signals. ECG signals from the MIT/BIH CD-ROM, the Normal Sinus Rhythm Database (NSR-DB), and European ST-T database have been utilized in this paper. The best classification rates obtained are 93% and 91.9% using EDBD learning rule with two hidden layers for the first structure and one hidden layer for the second structure, respectively. The results successfully showed that the presented NN-based classifier can be used for diagnosis of ischemic heart disease.

Modelling moisture sorption characteristics in dried acid casein using connectionist paradigm vis-à-vis classical methods

Six connectionist models are proposed to predict sorption (adsorption and desorption) characteristics at three temperatures, i.e., 25°, 35° and 45 °C over a water activity range of 0.11–0.97 in dried acid casein prepared from buffalo skim milk. Also, several conventional empirical sorption models were used for fitting the sorption data. The dataset comprised of 210 data points. The Error Back Propagation (EBP) learning algorithm with Bayesian Regularisation/Levenberg-Marquardt optimisation techniques as well as various combinations of connectionist network parameters was employed. The connectionist models predicted the adsorption characteristics with an accuracy ranging between 1.32 and 2.60 Root Mean Square percent error (RMS%) as compared to the best (among six conventional empirical models used) GAB model, which attained RMS% between 1.92 and 5.77. While for desorption characteristics, the connectionist models attained RMS% between 1.56 and 4.08 vis-a-vis the best (among six conventional empirical models used) GAB model with RMS% ranging between 1.4 and 5.01. Hence, the results revealed that the connectionist models outperformed the conventional empirical models and, generally, best described the experimental adsorption and desorption data for dried acid casein prepared from buffalo skim milk.

Comparison of Supervised and Unsupervised Learning Algorithms for Pattern Classification

This paper presents a comparative account of unsupervised and supervised learning models and their pattern classification evaluations as applied to the higher education scenario. Classification plays a vital role in machine based learning algorithms and in the present study, we found that, though the error back-propagation learning algorithm as provided by supervised learning model is very efficient for a number of non-linear real-time problems, KSOM of unsupervised learning model, offers efficient solution and classification in the present study.

Decorrelated Feature Space and Neural Nets Based Framework for Failure Modes Clustering in Electronics Subjected to Mechanical Shock

Electronic systems under extreme shock and vibration environments may sustain several failure modes simultaneously. Previous experience indicates that the dominant failure modes experienced by packages in a drop and shock framework are in the solder interconnects including cracks at the package and the board interface, pad cratering, copper trace fatigue, and bulk-failure in the solder joint. In this paper, a method has been presented for failure mode classification using a combination of Karhunen Loeve transform with parity-based stepwise supervised training of a perceptrons. New is the early classification of multiple failure modes in the pre-failure space using supervised neural networks in conjunction with a Karhunen Loeve transform. The feature space has been formed by joint time frequency analysis. Because the cumulative damage may be accrued under repetitive loading with exposure to multiple shock events, the area array assemblies have been exposed to shock and feature vectors constructed to track damage initiation and progression. The error back propagation learning algorithm has been used for stepwise parity of each particular failure mode. The classified failure modes and failure regions belonging to each particular failure mode in the feature space are also validated by simulation of the designed neural network used for parity of feature space. Statistical similarity and validation of different classified dominant failure modes is performed by multivariate analysis of variance and Hotelling's T-square. The results of different classified dominant failure modes are also correlated with the experimental cross sections of the failed test assemblies. The methodology adopted in this paper can perform real-time fault monitoring with identification of a specific dominant failure mode, and is scalable to system level reliability.

Determination of Voltage Stabilityin Distribution Network Using ANN Technique

Voltage stability condition, reliability and service quality are the important parameters of a distribution system which need to be satisfied from the customer point of view.To measure voltage stability of distribution network a voltage stability index is presented in this paper. Voltage stability improvement of network is facilated by network reconfiguration and voltage stability level is quantified by voltage stability index. Network reconfiguration is a process which alters the feeder topological structure by changing the open/close status of the sectionalizing (normally closed) and ties switches (normally open) in the system. In this work, a two stage search of switching option i.e. local search and global search is implemented to achieve desired network configuration. A multilayer ANN model with Error Back Propagation Learning (EBPL) algorithm is simulated for global search to obtained optimal set of candidate switching. The proposed scheme is tested on an 11 kV practical radial distribution system consisting of 52 buses. After reconfiguration, better voltage stable condition of the system is attained. Active power loss of the system is reduced from 0.854 MW to 0.485 MW and reactive power loss of the system is reduced from 0.355MVar to 0.199MVar after achieving optimal configuration.