Search

The current paper aims to explore the capability of Artificial Neural Network models (ANN) to calculate the optimal operating policy of a single reservoir system (Al_Qaim reservoir on the Al_Khosar River). The ANN models proposed in this research were making use of the outcomes emerged from two Stochastic Dynamic Programming (SDP) models suggested by previous study on the same reservoir system i.e. Explicit Stochastic Dynamic Programming and Implicit Stochastic Dynamic Programming. The two ANN models have been used to find pattern between inflow and initial storage of the system in one hand, and the release and the final storage of the system on other hand. It is found that the topology of the first model which adopted the attributes of the ESDP is 2-6-2, while that which was implemented the ISDP attributes has a 2-10-2 ANN topology. The final results prevail that good agreement have been exist between the output (release) of the proposed ANN models and those obtained by the two (SDP) models with coefficients of determination 0.934 and 0.803 respectively. [ABSTRACT FROM AUTHOR]

This paper investigate the effect of distance on the Gaussian Mixture Models (GMM) for text dependent speaker identification. Three stages are used for three different distances from the microphone (1m, 2m, and 3m). The set of feature extraction used here include Mel frequency cepstral coefficient (MFCC), Bark frequency cepstral coefficient (BFCC) and linear predictive cepstral coefficient (LPCC). These features are obtained from 20 speakers (10 adults and 10 children) ;all spoke five Arabic words in 5 seconds. The set of classification includes two types GMM and multilayer perceptron neural network (MLP). Total results show that MFCC has the best performance in feature extraction, and GMM has better recognition than MLP as total recognition in GMM is 93.15% and recognition in MLP is 88.06%.The results show also that the recognition rate decreases from 93.15% to 80.82% as the distance is increased from 1m to 3m. [ABSTRACT FROM AUTHOR]

Scaling factor tuning is one of the most used method to enhance the performance of a fuzzy controller. This paper presents two intelligent tuning strategies to tune this factor. In the first strategy, a supervisor fuzzy controller SFC was designed to continuously adjust, on line, the scaling factor of the basic fuzzy controller BFC based on the error and change of error signals. In the second strategy, a neural network NN is used to do this task. Performance of the tuning strategies are compared with corresponding conventional fuzzy controller in terms of several performance measures such as steady state error, settling time, rising time, and peak overshoot. Simulation results show that SFC performance is better. The system implementation and tests are carried out using LabVIEW (V 8.2). [ABSTRACT FROM AUTHOR]