A neural network controller based on genetic algorithms

This paper explores feasibility of employing the non-recurrent backpropagation training algorithm for a recurrent neural network, Simultaneous Recurrent Neural network, for static optimisation. A simplifying observation that maps the recurrent network dynamics, which is configured to operate in relaxation mode as a static optimizer, to feedforward network dynamics is leveraged to facilitate application of a non-recurrent training algorithm such as the standard backpropagation and its variants. A simulation study that aims to assess feasibility, optimizing potential, and computational efficiency of training the Simultaneous Recurrent Neural network with non-recurrent backpropagation is conducted. A comparative computational complexity analysis between the Simultaneous Recurrent Neural network trained with non-recurrent backpropagation algorithm and the same network trained with the recurrent backpropagation algorithm is performed. Simulation results demonstrate that it is feasible to apply the non-recurrent backpropagation to train the Simultaneous Recurrent Neural network. The optimality and computational complexity analysis fails to demonstrate any advantage on behalf of the non-recurrent backpropagation versus the recurrent backpropagation for the optimisation problem considered. However, considerable future potential that is yet to be explored exists given that computationally efficient versions of the backpropagation training algorithm, namely quasi-Newton and conjugate gradient descent among others, are also applicable for the neural network proposed for static optimisation in this paper.

There are numerous algorithms available for training artificial neural networks. Besides classical algorithms for supervised learning such as backpropagation, associative memory and radial basis function, this training task can be employed by evolutionary computation since most of the gradient descent related algorithms can be view as an application of optimization theory and stochastic search. In this paper, the logistic model of population growth from ecology is integrated into initialization, selection and crossover operators of genetic algorithms for neural network training. These chaotic operators are very efficient in maintaining the population diversity during the evolution process of genetic algorithms. A comparison is done on the basis of a benchmark comprising several data classification problems for neural networks. Three variants of training – Backpropagation (BP), Genetic Algorithms (GA) and Genetic Algorithms with Chaotic Operators (GACO) – are described and compared. The experimental results confirm the dynamic mobility of chaotic algorithms in GACO network training, which can overcome saturation and improve the convergence rate.KeywordsChaos theoryChaotic operatorEvolutionary algorithmsNeural networks training

This paper presented the approach of prediction of pressure coefficient using feed-forward multilayer artificial neural network (ANN) by the backpropagation (BP) training algorithm. The regular shape square model of 1:1:2 (length: breadth: height) and a setback model which have a 20% setback at half of the height are considered in this study. The datasets of network architecture considered by the computational fluid dynamics method (CFD). Maximum 1% error is noticed between the Cp comparison of CFD and ANNs method. The predicted mean pressure coefficient (Cp) and fluctuation wind pressure coefficient (Cp,rms) from the calculated CFD data and predicted ANN data are plotted in the contour form. The study suggested that the setback model could control the pressure on different surfaces compared to the regular shape square model. The study also decided the ANN is a powerful tool to predict the pressure coefficient for the intermediate tapping points on the building, which may reduce the time and cost of wind tunnel study.KeywordsArtificial neural network (ANN)Backpropagation (BP) neural networkComputational fluid dynamics (CFD)Pressure coefficientPressure contourSetback buildingSquare buildingTapping pointsWind load

In this study, three artificial neural network (ANN) models are developed with back propagation (BP) optimization algorithms to predict various lightning damage modes in carbon/epoxy laminates. The proposed ANN models use three input variables associated with lightning waveform parameters (i.e., the peak current amplitude, rising time, and decaying time) to predict fiber damage, matrix damage, and through-thickness damage in the composites. The data used for training and testing the networks was actual lightning damage data collected from peer-reviewed published literature. Various BP training algorithms and network architecture configurations (i.e., data splitting, the number of neurons in a hidden layer, and the number of hidden layers) have been tested to improve the performance of the neural networks. Among the various BP algorithms considered, the Bayesian regularization back propagation (BRBP) showed the overall best performance in lightning damage prediction. When using the BRBP algorithm, as expected, the greater the fraction of the collected data that is allocated to the training dataset, the better the network is trained. In addition, the optimal ANN architecture was found to have a single hidden layer with 20 neurons. The ANN models proposed in this work may prove useful in preliminary assessments of lightning damage and reduce the number of expensive experimental lightning tests.

A multilayer perceptron with the extended Kalman filter (EKF) training algorithm is investigated as a communication receiver when intersymbol interference and white Gaussian noise are present. Besides discussing the complexity of the EKF algorithm, it is shown that the EKF has better performance than the conventional backpropagation (BP) training algorithm in the sense that is requires less training steps and also results in proper training even when the BP did not. With 2500 training symbols, the EKF resulted in about 4-dB performance improvement over the conventional BP. >

For reinforcement, the photochromic field and the cooperation between the theoretical and experimental branches of physics, the computational, theoretical artificial neural networks (CTANNs) and the resilient back propagation (R[Formula: see text]) training algorithm were used to model optical characterizations of casting (Admantan-Fulgide) thin films with different concentrations. The simulated values of ANN are in good agreement with the experimental data. The model was also used to predict values, which were not included in the training. The high precision of the model has been constructed. Moreover, the concentration dependence of both the energy gaps and Urbach’s tail were, also tested. The capability of the technique to simulate the experimental information with best accuracy and the foretelling of some concentrations which is not involved in the experimental data recommends it to dominate the modeling technique in casting (Admantan-Fulgide) thin films.

This study was aimed to investigate the development and evaluation of artificial intelligence techniques by using multilayer neural network. Levenberg–Marquardt back propagation (LMA) training algorithm was applied for calculating drinking water quality index (WQI) for Euphrates river (IRAQ). The transfer functions in the artificial network model were tangent sigmoid and linear for hidden and output layers, respectively. Eleven neurons presented for good prediction for results of (WQI) with a coefficient of correlation >0.97 and statistically calculated WQI values, inferring that the model predictions explain 94% of the variation in the calculated WQI scores. The WQI score of the Euphrates was 142 considered as poor. The analysis of sensitivity revealed that the total dissolved solids (TDS) is the highest effective variable with the relative importance of (26.3%), followed by electrical conductivity (EC) (23.1%), pH (17.3%), calcium (Ca) (0.149), chlorides (Cl) (11.2%), Hardness (5.7%), Temperature (1.3%), respectively. It can be concluded that the model presented in this study gives a useful alternate to WQI assessment, which use sub indices formulae.

"This paper uses an artificial neural network (ANN) and resilient back-propagation (Rprop) training algorithm to determine the optical constants of As 30 Se 70 − x Sn x (0 ≤ x ≤ 3) thin films. The simulated values of the ANN are in good agreement with the experimental data. The ANN models performance was also examined to predict the simulated values for As 30 Se 67 Sn 3 which was not included in the training and was found to be in accordance with the experimental data. The high precision of the ANN models as well as a great guessing performance have been exhibited. Moreover, the energy gap E g of As 30 Se 70 − x Sn x (0 ≤ x ≤ 9) thin films were calculated theoretically."

an Artificial Neural Network classifier is a nonparametric classifier. It does not need any priori knowledge regarding the statistical distribution of the class in a giver selected data Source. While, neural network can be trained to distinguish the criteria used to classify easily in a generalized manner that allows successful classification the newly arrived inputs not used during training. Through this paper it is eastliblished that back propagation neural network works successfully for the purpose of classification. Back propagation suffers from getting stuck into Local Minima. Weight optimization in Back propagation can be optimized using the Genetic Algorithm (GA). The back propagation algorithm is improved by invoking Genetic algorithm, to improve the overall performance of the classifier. The performance of a fitness algorithm using the approach suggested by us is a Hybrid System that is being analyzed in this paper. In this paper the issue of improving the fitness (weight adjustment) of Back propagation algorithm is addressed. Some of the Advantages of Hybrid algorithms are: convergence speed will be increased and the local minima problem can be overcome. The proposed Hybrid Algorithm is to perform learning as a back propagation and optimize weights using GA for classification.

The Back Propagation (BP) neural network genetic algorithm was used to identify alphabet, and the new algorithm combine the advantages of both genetic algorithm and the BP neural network. Genetic learning algorithm was used for the global optimization and BP training algorithm to accurately optimize the neural network weights and training the neural network to learn letter recognition algorithm. Add-noise alphabet of MATLAB simulation results show that the new network error recognition rate reduced by 10% compared to BP neural network and the recognition speed is also faster than the traditional BP neural network with accuracy and fast convergence.

Particle swarm optimization (PSO) motivated by the social behavior of organisms, is a step up to existing evolutionary algorithms for optimization of continuous nonlinear functions. Backpropagation (BP) is generally used for neural network training. Choosing a proper algorithm for training a neural network is very important. In this paper, a comparative study is made on the computational requirements of the PSO and BP as training algorithms for neural networks. Results are presented for a feedforward neural network learning a nonlinear function and these results show that the feedforward neural network weights converge faster with the PSO than with the BP algorithm.

A fuzzified radial basis function network (FRBFN) is a kind of fuzzy neural network that is obtained by direct fuzzification of the well known neural model RBFN. A FRBFN contains fuzzy weights and can handle fuzzy-in fuzzy-out data. This paper shows that a FRBFN can also be interpreted as a kind of fuzzy expert system. Hence it owns the advantages of simple structure and clear physical meaning. Some metrics for fuzzy numbers have been extended to the metrics for n-dimensional fuzzy vectors, which are applicable to computations in FRBFNs. The corresponding metric spaces for n-dimensional fuzzy vectors are proved to be complete. Further, FRBFNs are proved to be able to act as universal function approximators for any continuous fuzzy function defined on a compact set. This paper applies the proposed FRBFN to nonparametric fuzzy nonlinear regression problems for multidimensional LR-type fuzzy data. Fuzzy nonlinear regression with FRBFNs can be formulated as a nonlinear mathematical programming problem. Two training algorithms are proposed to quickly solve the two types of problems under different criteria and constraint conditions, namely, the two-stage and BP (Back-Propagation) training algorithms. Simulation studies are carried out to verify the feasibility and demonstrate the advantages of the proposed approaches.

Prediction of rheological properties of Iranian bread dough from chemical composition of wheat flour by using artificial neural networks

Due to the complex nature of training neural network (NN), this problem has gained popularity in the nonlinear optimization field. In order to avoid falling into local minimum because of inappropriate initial weights, a number of global search techniques are developed. This paper applies a novel global algorithm, which is electromagnetism-like mechanism (EM) algorithm, to train NN and the EM based algorithm for neural network training is presented. The performance of the proposed algorithm is evaluated in classification problems and the comparison with BP and GA algorithms shows its effectiveness.

Artificial neural networks are one of the most popular and promising areas of artificial intelligence research. Training data containing outliers are often a problem for supervised neural networks learning algorithms that may not always come up with acceptable performance. Many robust learning algorithms have been proposed so far to improve the performance of neural networks in the presence of outliers. In this paper, we investigate the performance of four different backpropagation training algorithms, which are conjugate gradient with Fletcher - Reeves updates, conjugate gradient with Polak - Ribiere updates, resilient backpropagation, and conjugate gradient with Powell - peal restart. We compare their performance in terms of Root Mean Square Error as a merit function and the training speed in seconds. Examined neural networks trained by aforementioned backpropagation learning algorithms, which used the robust M-estimators performance functions instead of MSE one, in order to get robust learning in the presence of outliers. The study results show that Traincgf is the best algorithm in terms of mean square error, while the Traincgp is the best in terms of training speed.