Sharif Digital Repository

In this paper, the application of neural networks for simulation and optimization of the cogeneration systems has been presented. CGAM problem, a benchmark in cogeneration systems, is chosen as a case study. Thermodynamic model includes precise modeling of the whole plant. For simulation of the steady sate behavior, the static neural network is applied. Then using dynamic neural network, plant is optimized thermodynamically. Multi layer feed forward neural networks is chosen as static net and recurrent neural networks as dynamic net. The steady state behavior of CGAM problem is simulated by MFNN. Subsequently, it is optimized by dynamic net. Results of static net have excellence agreement... 

The adaptive capability of filters is known to be increased by incorporating a neural network into the filtering procedure. In this paper, an adaptive algorithm for tracking maneuvering targets based on neural networks is proposed. This algorithm is implemented with two filters based on the current statistical model and a multilayer feedforward neural network. The two filters track the same maneuvering target in parallel and the neural network automatically considers all the state information of the two filters and adaptively adjusts the process variance of one of them to achieve better performance in different target maneuver tracking. Simulations results show that the proposed adaptive... 

In this study, we have proposed a novel approach to investigate the features of four subbands of 2-D wavelet transform in magnetic resonance images (MRIs) for normal and abnormal brains which defected by Multiple Sclerosis (MS). Concurrently, another method extracts different kinds of features in spatial domain. Totally, 116 features have been extracted. Before applying the algorithm, we have to use a registration method because of variety in size of brain images. All extracted features have been passed over the Principal Component Analysis (PCA) and have been pushed to an Artificial Neural Network (ANN) that is a feed-forward type. According to changing in position of defected parts of... 

In the last decades, much attention has been paid to the design of multi-speaker voice conversion. In this work, a new method for voice conversion (VC) using nonlinear principal component analysis (NLPCA) is presented. The principal components are extracted and transformed by a feed-forward neural network which is trained by combination of Genetic Algorithm (GA) and Back-Propagation (BP). Common pre- and post-processing approaches are applied to increase the quality of the synthesized speech. The results indicate that the proposed method can be considered as a step towards multi-speaker Voice conversion. © 2007 IEEE  

Blur identification is one important part of image restoration process. Linear motion blur is one of the most common degradation functions that corrupts images. Since 1976, many researchers tried to estimate motion blur parameters and this problem is solved in noise free images but in noisy images improvement can be done when image SNR is low. In this paper we have proposed a method to estimate motion blur parameters such as direction and length using Radon transform and Feed-Forward back propagation neural network for noisy images. To design the desired neural network, we used Weierstrass approximation theorem and Steifel reference Sets. The experimental results showed algorithm precision... 

The vibration of a simply-supported beam with rotary springs at either ends is studied. The governing equations of motion are investigated considering the nonlinear effect of stretching. These equations are made non-dimensional and solved to the first-order approximation using the two known methods, namely, the multiple scales and the mode summation. The first five natural frequencies of the beam for different pairs of the boundary condition parameters are evaluated. A multilayer feed-forward back-propagation artificial neural network is trained using these natural frequencies. The artificial neural network used in this study shows high degree of accuracy for the natural frequency of the... 

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... 

In this study, accurate estimation of nuclear power plant (NPP) parameters is done using the new and simple technique. The proposed technique using the genetic algorithm (GA) in combination with the Bayesian regularization (BR) and Levenberg- Marquardt (LM) learning algorithms identifies the appropriate architecture for estimation of the target parameters. In the first step, the input patterns features are selected using the features selection (FS) technique. In the second step, the appropriate number of hidden neurons and hidden layers are investigated to provide a more efficient initial population of the architectures. In the third step, the estimation of the target parameter is done using... 

Up to now, different methods have been developed for estimation of buildup factor (BF). However, either expensive estimation or time-consuming estimation are major restrictions/challenges of these methods. In this study a new technique utilizing combination of Monte Carlo method and the Bayesian regularization (BR) learning algorithm of multilayer feed-forward neural network (FFNN) is employed for estimation of BFs. First, the BFs of the different elements (i.e. Al, Cu, and Fe) at different energies and different mean free paths (MFPs) are modeled by the MCNP code. The results show that the calculated BFs by MCNP code are in good agreement with the reported values of American nuclear society... 

This paper introduces the concept of mapping of Artificially Intelligent (AI) computational systems. The concept of homunculus from human neurophysiology is extended to AI systems. It is assumed that an AI system behaves similarly to a mini-column or ganglion in the natural animal brain that comprises a layer of afferent (input) neurons, a number of interconnecting processing cells, and a layer of efferent (output) neurons or organs. The objective of the present study was to identify the correlation between the stimulus to each afferent neuron and the corresponding response from each efferent organ when the intelligent system is subjected to certain stimuli. To clarify the general concept, a... 

The approximation power of general feedforward neural networks with piecewise linear activation functions is investigated. First, lower bounds on the size of a network are established in terms of the approximation error and network depth and width. These bounds improve upon state- of-the-art bounds for certain classes of functions, such as strongly convex functions. Second, an upper bound is established on the difference of two neural networks with identical weights but different activation functions. © The Author(s) 2018  

Acute lymphoblastic leukemia is one of the most common hematologic malignancies among children, caused by uncontrolled growth of leukocytes. Since the main hallmarks of the disease is not specific, a considerable number of patients have been being misdiagnosed. Early diagnosis of the disease is usually made by morphological investigation of leukocytes under microscope. In light of the facts that decrease in cytoplasm-to-nucleus ratio is one of the main indicators of cancerous cells, and an accurate segmentation phase will lead to extraction of representative features, segmentation step is acknowledged as being crucial in design of a computer aided diagnosis (CAD). Previous researches have... 

According to the fact that humans and animals show marvelous abilities in walking on irregular terrain, there is a strong need for adaptive algorithms in walking of biped robots to behave like them. Since the stance leg can easily rise from the ground and it can easily rotate about the toe or the heel, the problem of controlling the biped robots is difficult. In this paper, according to the adaptive locomotion patterns of animals, coordination and control of body links have been done with Central Pattern Generator (CPG) in spinal cord and feedback network from musculoskeletal system. A one layer feedforward neural network that its inputs are the scaled joint variables and the touch sensors... 

Sensors are one of the most vital instruments in Nuclear Power Plants (NPPs), and operators and safety systems monitor and analyze various parameters reported by them. Failure to detect sensors malfunctions or anomalies would lead to the considerable consequences. In this research, a new method based on thermal–hydraulic simulation by RELAP5 code and Feed-Forward Neural Networks (FFNN) is introduced to detect faulty sensors and estimate their correct value. For design an efficient neural net, seven feature selectors (i.e., Information gain, ReliefF, F-regression, mRMR, Plus-L Minus-R, GA, and PSO), three sigmoid activation functions (i.e., Logistic, Tanh and Elliot), and three training... 

The past few years have witnessed a growing rate of attraction in adoption of Artificial Intelligence (AI) techniques and combining them to improve forecasting accuracy in different fields. Besides, stock market forecasting has always been a subject of interest for most investors and professional analysts. Stock market forecasting is a tough problem because of the uncertainties involved in the movement of the market. This paper proposes a hybrid artificial intelligence model for stock exchange index forecasting, the model is a combination of genetic algorithms and feedforward neural networks. Actually it evolves neural network weights by using genetic algorithms. We also employ preprocessing... 

Remaining useful life prediction of rolling element bearings with offline condition monitoring data is the purpose of this paper. A data driven algorithm based on feedforward neural network is proposed for this aim. Since, usually the number of offline measurements are not much enough, the generalized Weibull failure rated function is used for producing the auxiliary points that are employed for training. Considering the physics of the bearing degradation, level of vibration in the highfrequency bandwidth of the spectrum is used as a feature and its performance in bearing prognostic problem is compared with that of using popular recommended features in the diagnostic standard. Bearing... 

The current statistical model and adaptive filtering (CSMAF) algorithm is one of the most effective methods for tracking the maneuvering targets. However, it is still worthy to investigate the characteristics of the CSMAF algorithm, which has a higher precision in tracking the maneuvering targets with larger accelerations while it has a lower precision in tracking the maneuvering targets with smaller acceleration. In this paper a novel adaptive tracking algorithm for maneuvering targets is proposed. To overcome the disadvantage of the CSMAF algorithm, a simple multi-layer feedforward neural network (NN) is used By introducing NN, two sources of information of the filter are fused while its...