Sharif Digital Repository

Artificial neural networks with multilayer feed forward topology and back propagation algorithm containing two hidden layers are implemented to predict the effect of chemical composition and tensile properties on the both impact toughness and hardness of microalloyed API X70 line pipe steels. The chemical compositions in the forms of "carbon equivalent based on the International Institute of Welding equation (CEIIW)", "carbon equivalent based on the Ito-Bessyo equation (CEPcm)", "the sum of niobium, vanadium and titanium concentrations (VTiNb)", "the sum of niobium and vanadium concentrations (NbV)" and "the sum of chromium, molybdenum, nickel and copper concentrations (CrMoNiCu)", as well... 

This study is focused on simulating and optimizing design and configuration of anaerobic baffled reactor (ABR) by means of artificial neural network (ANN). This approach is aimed to assess an efficient ABR performance in various operational conditions treating municipal wastewater. For this purpose, to analyze comprehensively on a base of experimental data, the system is operated in two pilots of 48 liters net volume made of 8 compartments. In 7 months, more than 130 sets of data are obtained to be introduced to MATLAB neural network. These include removal efficiency of chemical oxidation demand (COD) and volatile fatty acids (VFAs) parameters. The finest correlative architecture obtained... 

Pressure filters are popular in small municipal water treatment plants. One of the principles for designing and using the various units of water treatment plants is the ability of assigning and predicting the performance of those units under different and various conditions that could be verified by making pilot scale tests and could be modeled by means of available programs and software such as artificial neural network. The goals of this study that was conducted to predict pressure filter efficiency are: (1) evaluations of pressure filter efficiency for turbidity removal under different conditions such as turbidity of raw water, filtration rate and filter pressure changes; (2) statistical... 

Production of highly viscous tar sand bitumen using Steam Assisted Gravity Drainage (SAGD) with a pair of horizontal wells has advantages over conventional steam flooding. This paper explores the use of Artificial Neural Networks (ANNs) as an alternative to the traditional SAGD simulation approach. Feed forward, multi-layered neural network meta-models are trained through the Back-Error-Propagation (BEP) learning algorithm to provide a versatile SAGD forecasting and analysis framework. The constructed neural network architectures are capable of estimating the recovery factors of the SAGD production as an enhanced oil recovery method satisfactorily. Rigorous studies regarding the hybrid... 

The reinforcement of concrete columns with fabric reinforced cementitious matrix (FRCM) is one of the most challenging issues in the construction of concrete structures, as there is still an absence of a promising model to assess their performance. This is because the behavior of such elements is complex and accompanied by a high margin of uncertainty. To address this issue, this study compiles a large dataset of the performance of FRCM-reinforced concrete columns under monotonic load. The obtained dataset is then used to train an artificial neural network (ANN) as a promising method for predicting the compressive strength of concrete columns with acceptable accuracy. Afterward, using a... 

The present study investigated a counter-flow cooling tower performance by integrating the Artificial Neural Networks and Intelligent Optimisation Algorithms (ANN-IOAs). For this purpose, two scenarios were evaluated. In the first scenario, inlet air wet-bulb temperature (T aw), inlet air dry bulb temperature (T ad), water to the air mass flow rate ratio (mw /ma), and rotor speed (υ) were the input parameters for the ANNs, while the output temperature (T wo) was the ANNs output. In the second scenario, the same input parameters applied for the first scenario were used as input variables and the tower efficiency (ε) was considered as an output parameter. The well-known IOAs methods, namely,... 

Facial expression imitation with applications in the design of human robot interaction (HRI) systems is an active area of research. In this study, we propose an approach for real-time imitation of human facial expression by a humanoid social robot 'Alice'. Artificial neural network (ANN) and Kinect sensor are used for recognition and classifying of the facial expressions like happiness, sadness, fear, anger and surprise; with the Alice humanoid robot imitating the comprehended expressions. Results and experiments demonstrate the effectiveness of the approach. © 2016 IEEE  

In this study, new method for NAA purposes at 30 kW Isfahan MNSR is suggested. An algorithm based on ANN is proposed to quantitatively predict the unknown elements with no need standard sample. A three-layer feed-forward ANN with back-propagation algorithm has been used to determine concentration of selenium and fluorine in Multiple Sclerosis patients and healthy people blood samples. Predicted concentration of elements show good agreement between new method and experiment results. The correlation coefficient between the experimentally determined and predicted values are 0.99104 and 0.99364, respectively. This method is a rapid and precise approach for elemental analysis. © 2022, Akadémiai... 

This paper applies a novel adaptive approach consisting of Artificial Neural Network (ANN) and Fuzzy Linear Regression (FLR) to improve car ownership forecasting in complex, ambiguous, and uncertain environments. This integrated approach is applied to forecast car ownership in Iran from 1930 to 2007. In this study, the level of car ownership is viewed as the result of demographic, politico-social, and urban structure factors including average family size, total population density, urban population density, urbanization rate, gross national product per capita, gasoline price, and total road length. To capture the potential complexity, uncertainty, and linearity relation between the car... 

This paper models car ownership in Iran based on the data in a period of years 1980 to 2007 by artificial neural network (ANN) and Fuzzy Linear Regression (FLR). The car ownership is mainly affected by purchasing power of the customers, social and demographic factors; the car ownership model has a multi variable form. To explain the effect of these factors, ANN and FLR models are applied. The major reason for applying fuzzy concept and ANN is to overcome the inter-correlation problem associated with the independent variables. In this study, average family size; total population; urban population; urbanization rate; gross national product per capita; gasoline price; total length of road are... 

In some systems with hysteresis behavior like Shape Memory Alloy (SMA) actuators and Piezo actuators, we essentially need an accurate modeling of hysteresis either for controller design or performance evaluation. One of the most interesting Hysteresis non-linearity identification methods is Preisach model in which hysteresis is modeled by linear combination of elemental operators. Despite good ability of Preisach modeling to extract main features of system with hysteresis behavior, cause of tough numerical nature of Preisach, it is not convenient to use in real-time control applications. In this paper we present a novel method based on Artificial Neural Network. For evaluation of proposed... 

University (Educational) buildings comprise a considerable contribution of commercial buildings energy consumption. Optimizing operational measures are widely suggested for energy saving in university buildings. Hence, in this study, a bi-level energy-efficient occupancy profile optimization method using a metaheuristic algorithm, integrated with a demand-driven control strategy adjusted with dynamic set-point temperature is proposed to optimize the energy consumption within in a university departmental building. To proceed with the proposed energy saving strategies; first, building's thermal behavior and AHU system performance are identified and modeled through Artificial Neural Networks... 

This paper provides a novel bad data detection processor to identify false data injection attacks (FDIAs) on the power system state estimation. The attackers are able to alter the result of the state estimation virtually intending to change the result of the state estimation without being detected by the bad data processors. However, using a specific configuration of an artificial neural network (ANN), named nonlinear autoregressive exogenous (NARX), can help to identify the injected bad data in state estimation. Considering the high correlation between power system measurements as well as state variables, the proposed neural network-based approach is feasible to detect any potential FDIAs.... 

In recent years, pozzolana are being increasingly used in different types of concretes and mortars. This investigation aims to evaluate the effect of replacing 7, 14, 21, and 28% cement (by weight) with silica fume (SF), granulated blast furnace slag (GBFS), zeolite, and fly ash (FA) under elevated temperatures. Forty mix designs were built with various water-to-cement ratios and plasticizers. Three hundred and six specimens were prepared, and the flexural, uniaxial compression and tensile tests were performed on specimens after exposure to elevated temperatures between 25°C and 900°C. The X-ray diffraction (XRD) test was then performed on the two series of specimens. Accordingly, the... 

The present study investigated a counter-flow cooling tower performance by integrating the Artificial Neural Networks and Intelligent Optimisation Algorithms (ANN-IOAs). For this purpose, two scenarios were evaluated. In the first scenario, inlet air wet-bulb temperature (T aw), inlet air dry bulb temperature (T ad), water to the air mass flow rate ratio (mw /ma), and rotor speed (υ) were the input parameters for the ANNs, while the output temperature (T wo) was the ANNs output. In the second scenario, the same input parameters applied for the first scenario were used as input variables and the tower efficiency (ε) was considered as an output parameter. The well-known IOAs methods, namely,... 

In this study, we have proposed a novel method investigates MR-Images for normal and abnormal brains which effected by Alzheimer's Disease (AD) to extract 336 number of different features based on texture analysis. Before applying this algorithm, we have to use a registration method because of variety in size of normal and abnormal images. Consequently, the output of Texture Analysis System (TAS) is a vector containing 336 elements that are features extracted from texture. This vector is considered as the input of the Artificial Neural Network (ANN) which is feed-forward one. The features extracted from the Gray-level Co-occurrence Matrix (GLCM) have been interpreted and compared with normal...