The rainfall-runoff relationship is one of the most complex hydrological phenomena. In recent years, hydrologists have successfully applied backpropagation NEURAL NETWORK as a tool to model various nonlinear hydrological processes because of its ability to generalize patterns in imprecise or noisy and ambiguous input and output data sets. However, the backpropagation NEURAL NETWORK convergence rate is relatively slow and solutions can be trapped at local minima. Hence, in this study, a new evolutionary algorithm, namely, particle swarm optimization is proposed to train the FEEDFORWARD NEURAL NETWORK. This particle swarm optimization FEEDFORWARD NEURAL NETWORK is applied to model the daily rainfall-runoff relationship in Sungai Bedup Basin, Sarawak, Malaysia. The model performance is measured using the coefficient of correlation and the Nash-Sutcliffe coefficient. The input data to the model are current rainfall, antecedent rainfall and antecedent runoff, while the output is current runoff. Particle swarm optimization FEEDFORWARD NEURAL NETWORK simulated the current runoff accurately with R = 0.872 and E2 = 0.775 for the training data set and R = 0.900 and E2 = 0.807 for testing data set. Thus, it can be concluded that the particle swarm optimization FEEDFORWARD NEURAL NETWORK method can be successfully used to model the rainfall-runoff relationship in Bedup Basin and it could be to be applied to other basins.

Sediment transport as a complicated and important phenomenon has attracted a lot of researchers during the last century; however there are some formulae to evaluate sediment loads in aquatic systems. Most of them still face two major problems: firstly, lack of accuracy and secondly, involvement of many parameters which makes them more challenging.Artificial NEURAL NETWORKs are known as model-free universal function approximators well suited to deal with real life engineering problems including time series predictions and parameter estimation. In this paper, sediment loads are predicted using two different types of multilayer FEEDFORWARD NEURAL NETWORKs, namely Multi-Layer perception (MLP) and Radial Basis Function (RBF). The input variables for both structures are considered to be flow discharge, mean flow depth and width, mean bed material's diameter and water surface slope and the output is sediment discharge. Some different cases have been studied. The results are promising. It has been also observed that mean square prediction errors for the developed MLP is equal to 0.0063 while the devised RBF NETWORKs produces much larger mean square errors, namely 0.01260. This indicates that the MLP-load-predictor outperforms the RBF-predictor.

This paper concerns the design of a NEURAL state observer for nonlinear dynamic systems with noisy measurement channels and in the presence of small model errors. The proposed observer consists of three FEEDFORWARD NEURAL parts, two of which are MLP universal approximators, which are being trained off-line and the last one being a Linearly Parameterized NEURAL NETWORK (LPNN), which is being updated on-line. The off-line trained parts are able to generate state estimations instantly and almost accurately, if there are not catastrophic errors in the mathematical model used. The contribution of the on-line adapting part is to compensate the remainder estimation error due to uncertain parameters and/or unmodeled dynamics. A time delay term is also added to compensate the arising differential effects in the observer. The proposed observer can learn the noise cancellation property by using noise corrupted data sets in the MLPs off-line training. Simulation results in two case studies show the high effectiveness of the proposed state observing method.

In this paper, a novel hybrid method based on NEURAL NETWORK is proposed to solve quadratic differential equation. Here NEURAL NETWORK is considered as a part of large field called NEURAL computing or soft computing. The model finds the approximated solution of Riccati differential equation inside its domain for the close enough neighborhood of the initial point. This method, in comparison with existing numerical methods, shows that the use of NEURAL NETWORKs provides solutions with good generalization and high accuracy.

Control of robotic systems is an interesting subject due to their wide spectrum applications in medicine, aerospace and other industries. This paper proposes a novel continuous control mechanism for tracking problem of a 5-DOF upper-limb exoskeleton robot. The proposed method is a combination of a recently developed robust integral of the sign of the error (RISE) feedback and NEURAL NETWORK (NN) feed-forward terms. The feed-forward NN learns nonlinear dynamics of the system and compensates for uncertainties while the NN approximation error and nonlinear bounded disturbances are overcome by the RISE term. Typical NN-based controllers generally result in uniformly ultimately bounded (UUB) stability due to the NN reconstruction error. In this paper to eliminate this error and achieve asymptotic tracking, the RISE feedback term is integrated into the NN compensator. Finally, a comparative study on the system performance is conducted between the proposed control strategy and two other conventional control methods. Simulation results illustrate the effectiveness of the proposed method.

Background: One of the most common cardiovascular diseases (CVDs) in the world is myocardial infarction (MI). By analyzing electrocardiogram and vectorcardiography (VCG) signals, it is possible to identify and characterize heart diseases such as MI. One of the new methods of detection is the use of spatio-temporal parameters of VCG signals. This study aimed to correctly distinguish healthy signals from patients, achieve acceptable accuracy, and show the benefits of VCG and its application as a method to cover the shortcoming of electrocardiography. Methods: In this study, in addition to applying electrocardiogram signals in the time domain, spatio-temporal patterns of VCG signals were used to identify 80 patients with MI, and differentiate them from 80 healthy individuals. Findings: When combining the 12-lead electrocardiography (ECG) and the 3-lead VCG features applied to the FEEDFORWARD NEURAL NETWORK classifier input, an accuracy of 91. 2%, specificity of 92. 6%, and specificity of 90% were obtained. The results were in higher values than when applied separately. Conclusion: The observations indicate that combined ECG and VCG methods can be effective in distinguishing MI cases from healthy cases. It is hoped that this method may be useful in the clinical evaluation and heart failure diagnosis.

The optimum design of solar energy systems strongly depends on the accuracy of solar radiation data. However, the availability of accurate solar radiation data is undermined by the high cost of measuring equipment or non-functional ones. This study developed a feed-forward backpropagation artificial NEURAL NETWORK model for prediction of global solar radiation in Makurdi, Nigeria (7. 7322° N long. 8. 5391° E) using MATLAB 2010a NEURAL NETWORK toolbox. The training and testing data were obtained from the Nigeria metrological station (NIMET), Makurdi. Five meteorological input parameters including maximum and temperature, mean relative humidity, wind speed, and sunshine hour were used, while global solar radiation was used as the output of the NETWORK. During training, the root mean square error, correlation coefficient and mean absolute percentage error (%) were 0. 80442, 0. 9797, and 3. 9588, respectively; for testing, a root mean square value, correlation coefficient, and mean absolute percentage error (%) were 0. 98831, 0. 9784, and 5. 561, respectively. These parameters suggest high reliability of the model for the prediction of solar radiation in locations where solar radiation data are not available.

Meteorological phenomena are complex systems with different parts that are in contact with each other as well as their surroundings. The purpose of this research is to demonstrate the efficiency of NEURAL NETWORKs in predicting meteorological variables. For this purpose, the prediction of horizontal visibility that is widely used in meteorology and aviation especially at airports has been selected for analysis. The data of this study are a compilation of Metar and Synop reports of Bandar Abbas synoptic station in the period from 1 to 30 March 2014. To implement this NETWORK, at first, the whole data were randomly divided into three categories with proportions of 75, 15 and 15 percent for learning, testing and validation of NETWORK and saved in other files. The seven variables for inputs (temperature, dew point temperature, atmospheric pressure, sky cloud coverage, wind speed and wind direction) of the NETWORK with 28 various composites tested with a FEEDFORWARD NETWORK and their correlation with the output and amount of root mean square (RMS) error of NETWORK have been studied. The results show, the compositions that containing the present air phenomena are most correlated with the horizontal visibility. Besides, the dew point temperature, atmospheric pressure and the amount of cloud cover are variables that alone do not have an affect on the horizontal visibility. In this research, a NETWORK which works with training NEURAL NETWORKs by resilient backpropagation algorithm is used. This is a learning heuristic for supervised learning in FEEDFORWARD artificial NEURAL NETWORKs, which only the sign of the partial derivative is used to determine the direction of the bias and weight updates and the magnitude of their derivative has no effect on their updates. Of course, the size of their change (increment and reduce rates) is determined by a separate update value. This NETWORK with eight neurons and sigmoid transfer function in the hidden layer and the linear transfer function in the output layer is used for predicting of horizontal visibility. This NETWORK was performed with two standardization data sets between intervals 0. 0-1. 0 and 0. 1-0. 9; also, different learning rates, incremental and reduced rates for weights and biases. The results show that the normalization is not appropriate between zero and one. The appropriate amounts of learning rate, incremental and reduced rates for this NETWORK are 0. 0001, 1. 2 and 0. 35, respectively. The values of the coefficient of determination for training, test and validation data for a running NETWORK with all variables were 0. 9972, 0. 9866 and 0. 9839, respectively. These values show that nearly 99 percent of the measured horizontal visibility is affected by these independent variables and the rest of its variations are dependent on other factors.