Papers by Keyword: Neural Network Modeling

Paper TitlePage

Authors: Mohamed N. Abouelwafa, Hassan A. El-Gamal, Yasser S. Mohamed, Wael A. Al-Tabey
Abstract: As the study of fatigue failure of composite materials needs a large number of experiments as well as long time, so there is a need for new computational technique to expand the spectrum of the results and to save time. The present work represents a new technique to predict the fatigue life of Woven Roving Glass Fiber Reinforced Epoxy (GFRE) subjected to combined completely reversed bending moments and internal hydrostatic pressure, with different pressure ratios (Pr) between the applied pressure and burst pressure equal to (0, 0.25, 0.5 and 0.75). Two fiber orientations (θ), [0o,90o]3s and [±45o]3s are considered. Two neural network structures, feed-forward (FFNN) and generalized regression (GRNN), are designed, trained and tested. The groups of data considered are the maximum stress and the Pressure ratio with the different fiber orientation. On the other hand, more accurate prediction method is obtained by using a useful expert system which is designed to aid the designer to decide whether his suggested data for the composite structure is suitable or not. In this expert system a neural network is designed to consider the design data as input and to get yes or no as output. The results show the feed-forward neural network is better results than that given by the generalized regression neural network. The designed expert system helped the designer with reliable conclusions about his decision of the combination of the proposed data.
Authors: Tatiana Simankina, Olga Popova
Abstract: The algorithm for clustering based on neural network modeling using T. Kohonen's self-organizing maps for the analysis of the housing stock is considered. This analysis of housing stock is required for the planning of complex reproduction of housing and major repairs regional programs development. The mechanism of self-organization is submitted. The representative sample clustering of the housing stock is produced. Its result is 16 groups of objects with a high level of internal similarity. The basic advantages of this approach for monitoring and analysis of the city housing stock are described.
Authors: M. Reza, Soleymani Yazdi, Hoseyn Dehghan, Hoda Amini
Abstract: The main objective of the present research is to find the influence of process parameters on the state variables (i.e., surface roughness and material removal rate) in Wire Electrical Discharge Machining (WEDM) of Titanium Diboride (TiB2) nanocomposite ceramics. This work adopted an L32 orthogonal array based on Taguchi method for design of experiments. Statistically evaluating the obtained data is carried out by using the analysis of variance, signal to noise and artificial neural network techniques. Then, the effects of process parameters on the surface roughness and material removal rate are studied. Finally, the Multilayer Perceptron (MLP) neural network is used to model the WEDM of TiB2 nanocomposite ceramic. The obtained results have demonstrated very good modeling capacity of the proposed neural network. Furthermore, analyses have appropriately presented the influence of process parameters on state variables.
Authors: Ying Dong Qu, Cheng Song Cui, San Ben Chen, Qing Chun Li
Abstract: A PID controller has been developed to improve the dimensional precision of a deposit during spray forming. Simulation has been carried out based on a dynamic neural network modeling of deposit dimension. The simulation results show that the PID controller is effective for the control of deposit dimension with short regulating time and low override. Experimental results show that the error between the actual size and the target value is very small even though a disturbance from unstable melt flow rate was introduced.
Showing 1 to 5 of 5 Paper Titles