profile - Razi University
Faculty Member of Razi University
Razi University
GHolamreza Karmi
Professor / Engineering / Dept. of Electrical and Electronic Engineeingِِِ
Current courses
| Course Name | unit | term |
|---|---|---|
| Communication Circuits | 3 | first semester Academic year 2026-2027 |
| 1 | first semester Academic year 2026-2027 | |
| 3 | 3 | first semester Academic year 2026-2027 |
Master Theses
-
Designing an Analog STDP Circuit Using a Memristor
Hadi Musavi 2026چكيده در اين كار، طراحي يك مدار آنالوگ مبتني بر ممريستور براي پياده سازي قانون يادگيري (STDP (Plasticity Dependent-Timing-Spike ارائه شده است. در اين ساختار، ممريستور به عنوان عنصر سيناپسي با وزن قابلتنظيم به كار ميرود و تغيير مقاومت آن تابع اختالف زماني بين اسپايكهاي نورون پيش سيناپسي و پسسيناپسي است. بر اساس قانون STDP ،اگر اسپايك پيشسيناپسي قبل از اسپايك پسسيناپسي رخ دهد، فرآيند تقويت بلندمدت )LTP ) اتفاق افتاده و مقاومت ممريستور افزايش مييابد؛ در مقابل، اگر اسپايك پس سيناپسي زودتر رخ دهد، تضعيف بلندمدت )LTD ) ايجاد شده و مقاومت كاهش مييابد. اين رفتار با رابطه نمايي وابسته به اختالف زمان ?t مدلسازي ميشود.براي تحقق اين مكانيزم در حوزه آنالوگ، از مدارهاي RC به منظور توليد پاسخهاي نمايي زماني و شكلدهي پنجره يادگيري استفاده شده است. ولتاژهاي حاصل از مسيرهاي پيش و پس سيناپسي بهگونهاي بر دو سر ممريستور اعمال ميشوند كه پالريته و دامنه ولتاژ مؤثر، نوع و ميزان تغيير وزن را تعيين كند. مدل رفتاري ممريستور نيز با استفاده از معادالت وابسته به حالت داخلي شبيهسازي شده و امكان پيادهسازي در محيط HSPICE فراهم گرديده است. نتايج تحليلي و شبيهسازي نشان مي دهد كه مدار پيشنهادي قادر است منحني استاندارد STDP با ناحيههاي LTP و LTD نمايي را بازتوليد كند. اين ساختار به دليل سادگي، مقياسپذيري، توان مصرفي پايين و سازگاري با فناوري CMOS ،گزينهاي مناسب براي پيادهسازي شبكههاي عصبي نورومورفيك سختافزاري و سامانههاي يادگيري تطبيقي محسوب ميشود.
-
Sequence-Based Prediction of Enzyme Function Using Artificial Intelligence Methods
Setareh Vakili sharif 2026آنزيمها يكي از مهمترين دستههاي پروتئيني در سامانههاي زيستي به شمار ميروند و شناسايي درست عملكرد آنها اهميت زيادي در حوزههايي نظير زيستفناوري، مهندسي آنزيم، كشف دارو و زيستشناسي محاسباتي دارد. روشهاي آزمايشگاهي سنتي براي تعيين عملكرد پروتئينها زمانبر و پرهزينهاند و توان پاسخگويي به حجم عظيم دادههاي پروتئيني كه اين روزها توليد ميشود را ندارند. از همين رو، روشهاي محاسباتي و مبتني بر هوش مصنوعي به گزينهاي جذاب براي پيشبيني سريع و در مقياس بزرگ عملكرد پروتئينها تبديل شدهاند. در اين پژوهش تلاش شده است مدلي هوشمند طراحي و توسعه داده شود كه بتواند صرفاً بر اساس توالي اسيدهاي آمينه، كلاس اصلي عملكردي هر آنزيم را پيشبيني كند.
-
Numerical analysis of bimetallic tube sinking process through conical dies
ALI HASAN MIRZA 2026The study of the deformation behavior of metals has various applications in human life. In general, metal forming processes are divided into two categories: bulk forming and sheet metal forming, which are explained in detail below. One of the most important types of bulk metal forming is drawing. The bimetallic tube material was considered to be 6063 aluminum alloy and pure commercial copper. The term bimetallic structures are used when two metals are connected to each other mainly in the form of sheets or tubes. Copper-aluminum tube or aluminum-copper tube combines the advantages of copper and aluminum and is used instead of pure copper tubes or pure aluminum tubes
-
All optical modulation in a THz slow light modulator and switch designed employing FEM analysis and optimization by using 800nm LASER
MOHAMMED JAMEEL KAMIL 2026 -
تشخيص صفحات جعلي درگاه هاي پرداخت اينترنتي با استفاده از روش CNN
AFSANEH GERABAJI 2026Research Aim: Phishing is a major security threat in online payments, where users are tricked into disclosing sensitive financial information through fake payment gateway pages. This study aims to design and implement an accurate and reliable model for detecting fraudulent payment gateway pages using Convolutional Neural Networks (CNN). Research method: In this research, a dataset of 80,000 samples was collected, including 50,000 legitimate pages and 30,000 phishing pages. The data were preprocessed by rendering HTML pages into images, resizing, normalization, and data augmentation before being input into the network. The CNN model, with a multi-layer architecture consisting of convolutional, pooling, and fully connected layers, extracted spatial features of the pages and >Findings: Quantitative results showed that the proposed model achieved an accuracy of 97%, a recall of 97.8%, and an area under the ROC curve (AUC) of 0.991 in distinguishing phishing pages from legitimate ones. Qualitative analysis indicated that the model effectively detected highly similar pages, with false negatives occurring only when static real screenshots were used. Comparisons with text-based features and >Conclusion: Considering its high accuracy, fast prediction speed, and deployability in practical environments, the proposed model can be applied in banking systems, browsers, and online payment applications. Future research suggestions include incorporating multimodal features, lightweight models for mobile applications, and automatic model retraining to further enhance online payment security. Keyword: Online Security, Fraud Detection, Deep Learning, Image Analysis, Financial Transaction Safety
-
امكان سنجي پياده سازي مداري تحريك عمقي مغز براي درمان بيماري پاركينسون
Shadi Ezatizadeh 2025 -
Approximate adder design considering energy and delay
Tayyebeh Karimi 2025Approximate computing is a promising approach for high-performance, and low-energy computation in inherently error-tolerant applications. This study proposes an approximate adder comprising a constant-truncation block in the least significant part and several non-overlapping summation blocks in the more significant parts of the adder. The carry-in of each block is supplied using the most significant bit of one of the input operands from the earlier block. In the most significant block, two more-precise approaches are used to generate candidate values for the carry-in. The final value of the carry-in for this block is selected based on the values of the input operands. In fact, the proposed approximate adder is input-aware, and dynamically adjusts its operation in one or two cycles to improve accuracy while limiting the average delay. The experimental results indicate that the proposed adder has a better quality-effort tradeoff than state-of-the-art approximate adders. Different configurations of the proposed adder improve delay, energy, and the energy-delay product (EDP) by 78%, 72% and 87% respectively, when compared to state-of-the-art approximate adders, all without any loss in accuracy. Additionally, the efficiency of the proposed adder is confirmed in both image dithering and stock price prediction through regression.
-
Design, simulation and fabrication of a compact Wilkinson power divider with wide bandwidth based on hybrid structure.
Mohammad Amin Nasseri 2025In this thesis, a Wilkinson power divider is designed using a hybrid structure composed of three resonators—rectangular, square, and U-shaped—instead of the conventional quarter-wavelength transmission lines used in standard Wilkinson power dividers. The main objective of this thesis is to design a compact power divider with a wide bandwidth and improved S-parameters. This thesis consists of five chapters. Chapter one provides the fundamental concepts necessary for understanding the operation of power dividers. Chapter two reviews the concept of power division and examines common types of power divider structures along with their analysis. Chapter three presents a review of several designs that have been proposed in recent years. In chapter four, the proposed power divider is introduced and thoroughly analyzed from initial design to the final structure, including equivalent LC circuit modeling, even- and odd-mode analysis, and current density distribution. Finally, chapter five summarizes the key findings and conclusions of the research. Furthermore, this thesis introduces an innovative method to achieve multiple different center frequencies without altering the main structure. In this method, by placing three rectangular-shaped structures between the input and output ports and only changing their widths, different center frequencies can be obtained. After the design process, the characteristics and parameters of the power divider are as follows: The divider operates at a center frequency of 2.14 GHz and is capable of suppressing the second to eleventh harmonics with a suppression level of -21 dB. The physical dimensions of the divider are 10.743 mm × 10.243 mm, equivalent to 0.0095?g2 , representing an 81% size reduction compared to the conventional structure at the center frequency. The return loss is better than -51 dB, insertion loss is -3.087 dB, return loss at the output ports is better than -46 dB, and isolation between the output ports is better than -45 dB.
-
Improving the performance and efficiency of the HEMT transistor based on the use of diode characteristics.
Fatemeh Zeini 2025 -
Brain tumor detection from MRI images using artificial neural network
Faezeh Parvizi 2025Brain tumors are caused by abnormal cell growth in the brain. Magnetic resonance imaging (MRI) is the most widely used method for diagnosing brain tumors. Through these MRIs, doctors analyze and identify abnormal tissue growth and can confirm whether the brain is affected by a tumor or not. Today, with the advent of artificial intelligence techniques, the diagnosis of brain tumors is performed using machine learning techniques and algorithms and artificial neural networks. The advantages of using these algorithms include rapid prediction of brain tumors, fewer errors, and greater accuracy, which helps in decision-making and choosing the most appropriate treatment for patients. In this study, an artificial neural network will be used to detect the presence of a brain tumor and its performance will be analyzed. The main goal of this research is to design an artificial neural network-based system for automatic detection of brain tumors from MRI images and classify MRI images into two categories: "brain tumor" and "normal" and ultimately achieve high diagnostic accuracy in MRI images. Keywords: Brain tumor detection, Artificial neural networks (ANN), MRI images, Convolutional neural network (CNN).
-
Estimating the Risk of Death in COVID-19 Patients Using an Optimized Transformer Model Based on CDC Data
Milad Gholami 2025 -
test pattern generation for combinational digital circuits using parallel pattern critical path tracing
Zeinab Moradi 2024Abstract Today, with the growing complexity of digital circuits and the increasing compactness of manufacturing technologies, the likelihood of failure during both the manufacturing process and the operation of digital circuits has risen. As a result, these products require testing both during production and operation to ensure proper performance. Therefore, producing a high-quality test with a minimal number of test vectors and in the shortest time is crucial. In this thesis, a simulation-based test generation method for combinational circuits is proposed. This method utilizes an approximate criterion called approximate critical path tracing with parallel patterns to evaluate the effectiveness of candidate test vectors. This criterion is based on the traditional critical path tracing method, but introduces an approximate approach for backtracking in fan-outs. By allowing some inaccuracies in the results, this method reduces the complexity of fault simulation compared to traditional methods. The parallel pattern critical path tracing method leverages both fault-level and pattern-level parallelism, resulting in significantly higher simulation speed compared to traditional methods for determining fault coverage. Applying the parallel patterns approximate critical path tracing algorithm to ISCAS85, ISCAS89, and ITC99 benchmark circuits shows that, in over 98% of the circuits, the results strongly correlate with the exact fault coverage index. Additionally, compared to the parallel pattern fault simulation method, this approach is more than 500 times faster. The proposed test generation method, which utilizes the criterion derived from the parallel pattern approximate critical path tracing algorithm to identify effective test vectors, is capable of producing high-quality test vectors quickly. Evaluations on benchmark circuits demonstrate that this method is over 15 times faster than the one using parallel pattern fault simulation, with only a 1% average increase in the number of test vectors. Key words: Digital circuits testing, Fault coverage, Simulation-based test-pattern generation, Approximate fault coverage index, Critical path tracing.
-
طراحي مبدل آنالوگ به ديجيتال دو بيتي اپتيكي بر مبناي كريستال هاي فوتوني دو بعدي
Pooria Ghafoori 2024A novel design for realizing optical 2-bit analog to digital converter based on two-dimensional photonic crystals will be proposed in this thesis. The proposed structure consists of two main parts; a nonlinear 3-channel demultiplexer, followed by an optical coder. The nonlinear demultiplexer quantizes the input analogue signal based on its optical intensity at the central wavelength of ?=1550 nm and consists of two ring resonators using nonlinear rods inside. For appropriate values of the input optical intensity, one of the ring resonators can drop the optical signal at its corresponding output port using Kerr effect. The optical coder also contains two similar ring resonators which can converts the quantized levels to 2-bit binary code. The base structure is a square lattice of silicon dielectric rods with refractive index of 3.46 implemented in an air background. The nonlinear rods used in the ring resonators are made of doped glass with refractive index and nonlinear Kerr coefficient of 1.4 and 10-14 m2W respectively, and the total footprint of the structure is about 555 ?m2. Maximum sampling rate is up to 240 GS/s .
-
Design, simulation and fabrication of Wilkinson power divider using high impedance meandered lines and new structure to increase operating bandwidth
Payman Fallahisepehr 2024 -
Designing and simulation of logic gates and circuits based on photonic crystal structures
Mehdi Bagheri 2024 -
Simulation, design and construction of Wilkinson power divider using low pass filter, PSO based optimization and analysis
Milad Mohammadi 2023 -
Design, simulation and construction of Wilkinson power divider based on the optimization of neural networks and the use of a trapezoidal resonator.
Hesam Qytasy sarabi 2023 -
design of a low-noise amplifier with high gain and low noise for (WLAN) application
2023The current research is to present a new model for low noise amplifier design in wireless local area networks. For this purpose, we consider a low noise amplifier with cascode technique and then modify the cascade. The results of this study show that the modified cascade stage has high gain and optimal noise figure. The LNA will be designed to be stable in the 3.1 to 10.6 GHz range. The results of this study can be widely used in medical devices. The results of this research showed that the amount of noise in the designed circuit has been reduced by 45.57%. Also, the amount of gain in the circuit designed by the author has also increased by 20.5%. Therefore, it can be said that in the current research, satisfactory results have been obtained regarding reducing noise and increasing gain.
-
Design and simulation of comact microstrip lowpass filter with wide stopband using cone shaped resonator based on an analytical model
Negar Moradi 2023 -
Investigation of SOI MESFET transistors to improve performance prameters by reducing noise effects in the channel
Amir Karami mzraneh 2023Today, SOI-MESFET transistors have many applications in the electronic world and due to their advantages such as high switching speed and working at high voltage and frequencies and reducing power consumption compared to BULK silicon body technology, but with this advantage There are also some limitations, such as the effect of self-heating and the effect of body buoyancy. In this research, we introduce a new silicon-on-insulator structure, which compared to the conventional structure, has advantages such as higher breakdown voltage, higher drain current, and improvement in RF parameters. In this structure, we have used an oxide region in the channel region. The oxide region is located between the gate and the drain, which causes an increase in the breakdown voltage, and the reason for this is that the breakdown tolerance of the oxide is higher than that of the semiconductor. A metal region is buried inside the oxide region. The metal region improves RF parameters and prevents electric field congestion. In this structure, the breakdown voltage is around 22 volts, compared to the basic structure, which is about 19 volts, we can see an increase of 3 volts, and also the drain current has increased compared to the basic structure, and the RF parameters have all improved, and as a result This structure is ordered to work in high power applications. This research is about a completely new structure that has excellent efficiency for working at very high powers. In this structure, by using an oxide region in the channel, it improves the breakdown voltage from 19 volts in the basic structure to 22 volts. We are new in the structure, and on the other hand, by using nickel and SI3N4 areas, we have improved the ac parameters and increased the maximum transmission power from 0.9 W/mm in the basic structure to 0.998 W/mm in the new structure, so we can safely say that It is an excellent structure for working at high powers
-
Design Of A Low Noise Amplifier With Desired Linearity Based On Volterra Series
Reza Mohammadi 2022 -
Detection of covid-19 disease using optimum convolutional neural networks
Ali Mohammadi 2022Nowadays, people all around the world are suffering from many problems and diseases, one of the most important of that in recent years is covid-19, which causes many social, economy and other harms to human society. Therefore, treatment of this disease is one of the priorities of human beings today to return to normal living condition. One of the most significant steps and backgrounds for treatment of this disease and other similar diseases is nothing but diagnose this disease in accurate and fast methods that are useful for providing the ground for treatment, increasing knowledge about of that and preventing from spreading new strains. There are many different kinds of ways to diagnose this disease, which one of the most accurate and best is imaging and scan the lung and analyze it. On the other hand, artificial neural networks have many applications such as, processing and analyzing various data based on existing data sets from numbers to images, etc. one of these types of networks, is called convolutional neural networks, which are specifically used for analyzing, classification and processing various images. In this research, our goal is processing lung scan images by convolutional neural networks and receive the results in the output. Our results appear in the output according to predefined classifications based on the primary data sets, which include conid-19, normal and pneumonia. We can benefit from these results to diagnose and combat this disease. Because of the attractiveness and practicality of this field, before the present research, various researches have been done in this field. One of most important goals of this research is improve and optimize the convolutional networks by utilizing meta- heuristic algorithms in order to reduce the layers and network connections for easier and more cost-effective implementation while maintaining network speed and accuracy.
-
The effect of pressure on the soil adjacent to the Kermanshah monorail tunnel due to its peripheral construction
Mohammad Sokhansanj 2021Abstract In this study, the simulation of the effect of different surface buildings on the Kermanshah monorail tunnel has been studied. Numerical modeling of the present study has been performed in the form of two-dimensional finite element method and in plane-strain or two-dimensional strain logic. The software selected for modeling this study was the advanced commercial-research finite element code PLAXIS 2D V8.6. At least 16 different numerical models including the cross section of Kermanshah monorail tunnel and surcharge of buildings with different number of floors have been modeled. The soil model is assumed to be a Mohr-Coulomb elastic-perfect paste model. The parameters of the numerical model have been selected based on the results of the borehole logs studied by the esteemed project consultant. The cross section of the Kermanshah monorail tunnel is in the shape of a horseshoe and is based on the new Austrian tunneling method or NATM. Building surcharge load is simulated based on its equivalent stress at the soil surface and the physics of the buildings are not modeled. The 2 to 100 stories buildings have been applied as surcharge load to the numerical model of Kermanshah monorail tunnel. Also, a free field reference model without the effect of building surcharge load is included in the 2D modeling of the research. Logical relationships are obtained between the increase in surface building surcharge and the amount of deformation. The values ??of the settlements are about five times higher than the horizontal deformations. That is, the subsidence of the soil layer above the structure of the monorail tunnel is about 5 times more affected by the lateral deformation than the increase of the upper surcharge of the building. By simultaneously comparing the maximum shear strain of different parametric numerical models with different construction surcharges, the quantitative range of shear strains is up to about 4% of the maximum shear strain. Up to an surcharge of about 15 floors, which is equivalent to the overhead of existing buildings in Kermanshah, the shear stress of different cases is almost equal (about 300 kPa). This shows that increasing the maximum 100 floors of the building above ground level has increased the amount of axial force inside the tunnel cover by at least 3.5 times. Keywords: Soil pressure, Kermanshah monorail, tunnel, peripheral structures, soil excavation, numerical modeling.
-
Denoising of the electrocardiogram signal using wavelet transform
Yosef Felekari 2021 -
بررسي قابليت خود تعميري در يك شبكه عصبي اسپايكي
Rezvan Rostami Tabar 2021چكيده مغز انسان پيچيده ترين سيستم هستي است كه از ويژگي هاي منحصر به فردي برخوردار است. طي چند دهه اخير ، پژوهشگران سعي داشته اند تا با الهام گرفتن از مغز و نحوه عملكرد آن در پردازش اطلاعات ، سيستم هايي مشابه آن ايجاد كنند . يكي از ويژگي هاي مهم مغز توانايي خود تعميري و قابليت تحمل خطاست ؛ به طوري كه در صورت آسيب ديدن چند سيناپس ، به دليل وجود سلول هايي به نام آستروسيت و از طريق سيناپس هاي سالم باقي مانده ، سيستم عملكرد خود را در حد قابل قبولي حفظ مي كند . از آن جا كه سلول هاي عصبي در محيطي تصادفي فعاليت مي كنند ، وجود نويز يك مسئله غير قابل اجتناب است . به همين دليل در اين پروژه براي اولين بار به بررسي تاثير نويز در شبكه آستروسيت – نورون خود تعمير پرداخته شده است . با وجود اعمال نويز به جريان نورون ها ، سيستم هم چنان از توانايي خود تعميري برخوردار است و از طريق فيدبك غير مستقيم از سلول آستروسيت اثرات مخرب سيناپس هاي آسيب ديده را تا حدودي جبران مي كند . نتايج نشان مي دهند كه با وجود حضور نويز در جريان نورون ها و اعمال 80 درصد خطا به شبكه ، سلول آستروسيت هم چنان سعي در جبران خطا دارد و تا حد امكان فركانس هدف را حفظ مي نمايد .
-
Investigation of the noise influence on performance of an analog circuit for biological neuron model
Leila Tasbihchi 2021Abstract: The study of neural systems has been of great interest to scientists around the world in recent years, as scientists in various sciences, including biology, neurophysiology and engineering, have tried to study, understand and characterize neural networks. In this regard, neuromorphic scientists and engineers, along with neurobiologists, have simulated and implemented different parts of the nervous system, such as neurons, through analog and digital circuits. Understanding and imitating the natural state of neuronal communication can lead to further advances in the treatment of neurological diseases and injuries. A neuromorphic circuit that implements neurons electronically can reproduce the behaviors observed in real neurons using electronic circuits. The more the operation of these circuits is similar to the operation of a real neuron, the more efficient the circuit. In most of these designs, the effect of noise, which has always existed in electronic circuits, is ignored. Noise is one of the sources of changes in the function of the nervous system that affects all aspects of the functioning of the nervous system. Therefore, it is very important to know the types of noise and its sources in the nervous system and to study its effect on the structure and function of the brain and neurons. In this dissertation, our aim is to investigate the effects of noise on the output of the analog circuit of the Izhikovich neuron model. Using MATLAB software, the mathematical equations of the used neural model are implemented and the output of the model is simulated in this software. We will also simulate the orbital part of this research using Hspice software and examine the effects of noise on the circuit output. Finally, the output of MATLAB software and the output of the simulated analog circuit in Hspice software are compared. Keywords: Neuromorphic, Neuron, Spike, Noise, Izhikevich neuron model.
-
investigation of anisotropic gate dielectric effect on short channel effects in conventional MOSFET
Aria Fathi 2020 -
طراحي يك مدار VLSI نوروموفيك براي پياده سازي پلاستيسته سيناپسي وابسته به زمان اسپايك
Fateme Rahimi 2020 -
Investigation of the Effect of Noise on An Analog Neuron Perfermance
Azin Amjadian 2020 -
Implementation Deep Convolution Neural Network on FPGA
Jonaid Shiani far 2020 -
بررسي اثر نويز بر عملكرد شبكه هاي نوروني بيولوژيكي
Niloofar Amjadian 2020 -
Sensitivity analysis of effective factors on energy consumption and thermal comfort of an office building and optimizing them using multi-objective genetic algorithm: case study
Mohammad amin Ghaderian 2019Abstract:Worldwide, the building sector is responsible for about 30% of the green gas emissions and about 40% of the energy consumption. So, improving building performance can play a crucial role to cope with climate changes and resources depletion. Optimization methods can be more effective in finding the optimal building design after eliminating the less important variables using global sensitivity analysis and thereby reducing the search space. Firstly, in this thesis to identify those input variables that have a large impact on annual energy consumption of an existing office building and thermal comfort of its occupants, to assist building energy engineers and policymakers to decide on the best strategies in retrofitting proce two different sensitivity analysis methods, namely one-factor-at-a-time (OFAT) and analysis of variance (ANOVA), have been applied to outcomes of a validated model of the studied office building. These output variables were the annual electricity and gas consumption and average absolute PMV. The variables chosen as inputs that could be changed easily for the building, they were: heating and cooling set points, air infiltration and ventilation rates, supply water temperature for heating and cooling, and overall heat transfer coefficient of external walls. Then by building simulation based optimization, has been tried to reduce building's energy need, while maintaining thermal comfort in acceptable range. Because this task involves direct coupling of the optimization algorithm to a simulation model, it is computationally intensive. To overcome this issue, this thesis has described an optimization methodology based on a combination of regression models and a Multi-objective Evolutionary Algorithm (NSGA-II) that has applied for the case study. The optimization results shows that the applied method, while maintaining occupants' thermal comfort in acceptable zone, can decrease building's annual gas and electricity consumption 88.1% and 39.2%, respectively. Keywords: building energy consumption; occupant thermal comfort; sensitivity analysis; one factor at a time (OFAT); analysis of variance (ANOVA); surrogate regression model; multi-objective optimization (NSGA-II)
-
Design and simulation analog circuit Li-Rinzel calcium model with CMOS technology
Ali Khosravi 2019 -
بررسي عددي تاثير موقعيت دريچه هاي ورودي هواي تهويه بر روي شرايط آسايش حرارتي با رويكرد كاهش مصرف انرژي
Allahbakhsh Barati 2019 -
پياده سازي سيستم كنترل فتوولنايك با استفاده از FPGA
2019 -
Design and implement a system of estimating the distance of objects by using image processing
Siavash Moslem 2019 -
Thermal buckling analysis of trapezoidal composite plates with embedded shape memory wires
Fatemeh Basati 2018 -
Design and impelementation of hardware in loop for switch reluctance machine with speed control capability
Ehsan Hajebi 2018SRM motors have attracted considerable attention due to its low cost and robust structure, high efficiency, and the ability to track at variable and high speed and high ambient temperature. SRM engines are one of the oldest types of electric motors that were left out of the lack because of proper control systems, but today with modern semiconductor technology, SRM engines can be made cheap and even easier than induction motors, and will Compete with all other electric motors soon. One of the problems with this machine is the speed control complexity that is influenced by the effect of the voltage thresholds of each phase and the detection pulses which applied on phases but nowadays engineers are working to overcome the obstacles that cause these problems over time.
-
sensorless drive and control for switched reluctance motor by using ARM microprocessor and pulse injection technique
Pourya Espari 2018SRM motors have attracted considerable attention due to its low cost and robust structure, high efficiency, and the ability to track at variable and high speed and high ambient temperature. SRM engines are one of the oldest types of electric motors that were left out of the lack because of proper control systems, but today with modern semiconductor technology, SRM engines can be made cheap and even easier than induction motors, and will Compete with all other electric motors soon. using the sensor to detect the position of the rotor increases the price and complexity of the engine structure, and in addition, in the event of burning or dusting and the mass of each sensor and also noise on sensor cables which given the fact that it is used in industrial environments or electrical appliances, It is likely that this engine will not be welcomed, but in a sensorless method that uses feedback from different parts of the engine and using a microprocessor and writing complex codes to detect the position of the rotor and start the engine and take a long time It takes itself and needs much more science, but all of these happen just one time, and the engine's surroundings make it easier to control other engine parameters.
-
FPGA-based Implementation of a Real-Time Photovoltaic System Emulator
Mohammad Amirian 2018 -
design and fabrication of real-time power transformer emulator based on field programable gate array (FPGA)
Reza Molaei 2018In this thesis design and fabrication of real time powersystem emulator for power transformer study was presented. Real time powersystem emulator historical studying and review on this emulator advantages inpower system analysis and scenario of hardware in the loop examining clearlyshow the subject importance. Many efforts have been devoted in this field andtodays trends in Field Programmable Gate Array(FPGA) utilization for itsendless advantages was increasingly reported in IEEE literatures and industrialapplications. A. Benigni and A. Monti in a paper named A Parallel Approach to Real Time Simulation of PowerElectronics System (2015)
-
Design Of Ultra Wideband Low Noise Amplifier (LNA) With Current Reuse And Noise Cancelling Technique
Ali Hatami 2018Design Of Ultra Wideband Low Noise Amplifier (LNA) With Current Reuse And Noise Cancelling Technique
-
Design of Ultra Wideband Low Noise Amplifier Using Noise Canceling And Feedback Resistance Technique
Surur Heidary 2018 -
بررسي منحني ظرفيت سازه هاي خمشي داراي ستون CFT
2018بررسي منحني ظرفيت سازه هاي داراي ستون بتن پر(CFD)
-
طراحي و شبيه سازي يك مدار آنالوگ جهت بررسي عملكرد آستروسيت در ناهمزماني يك جمعيت نوروني
Zahra Najafnia 2018Over the past several years, different control methods have been presented to avoid hyper synchronization of neurons in various papers, Because this abnormal synchrony disturbs the function of the brain and causes the diseases Like epilepsy and Parkinson. The primary treatment of disease is drug therapy. if drug treatment is unsuccessful, other therapies are there such as surgery and stimulation of the damaged area. in the surgical procedure, the damaged area of the brain is destroyed , which will have many side effects. In the direct stimulation technique, due to permanent stimulation of the damaged area, surgical problems can be reversed. Since stimulating methods of feedback and non- feedback are mathematical equations and have no biological background, tried to use astrocytic stimulant as a biological stimulant. Several studies have shown that the effect of astrocytes on the absorption or release of glutamate in the synaptic space, the storage of potassium ion, and the regulation of extracellular pH is necessary to maintain the correct level of irritability in the CNS. The impairment of this function can lead to an increase in the stimulation of the neurons and their synchronization with each other. The disorder of this function can lead to an increase in the stimulation of the neurons and their synchronization with another. Therefore, astrocytes can apply appropriate control feedback to regulate neuronal activity. In this study, we tried to investigate the biological stimulus ability induced by the behavioral dynamics of astrocytes using analog circuit design. Because astrocytes play a major role in neurological interactions in the neurons synchronization and desynchronization , the above stimulant advantage can be considered as the biological basis based on the interaction between neurons and astrocytes. To simulate the behavior of neurons in diseases caused by super-synchronization of neurons, the biological network of the neuronal population is considered, and the suggested stimulatory ability is investigate in their desynchronization. Considering the many applications that digital systems have, the design of analog circuits continues to apply where digital equivalents can not perform their tasks well. For example, in applications that require very low power, such as medicine, implantable elements, and portable and lightweight devices, digital circuits may not be appropriate. Also, all the signals in the real world are analog, and these analog signals will never be substitutable. For this reasons, the proposed system is designed with analog circuits and simulated in the HSpice software.Key words: synchronization, stimulation, astrocyte, neuronal population, analog circuit design.
-
numerical study of dynamic response of shallow foundations under machine loading a case study west oil and gas exploitation company
Farzaneh Rahimi pour 2018 -
طراحي و شبيه سازي يك مدار نورمورفيك آنالوگ با قابليت خود تعميري
Negin Veisi 2018Self-repairing is among the principal foundations of biological neural networks, upon which, adjacent synapses undertake functions of defected synapses to avoid disturbances in the network function, ending up compensating the incurred error. In the present research, a self-repairing analogue circuit is designed using an astrocyte-neuron relationship. The designed circuit is based on a software model of an astrocyte-neuron network with proven ability to detect errors and undertake self-repairing. The results obtained of the proposed circuit show that, when an error occurs in the synapses associated with a neuron, the currents of healthy synapses of the same neuron increases. This increase is made by receiving a feedback from adjacent astrocytes. This process maintains the network function and compensating the incurred error in the network. To the best of our knowledge, this is the first report to present a neural network-based analogue circuit with self-repairing capability considering the effect of astrocytes. In this research, the proposed circuit was designed and simulated in HSPICE software using standard 0.35 ?m CMOS technology.
-
Optimized Implementation Of IIR Filter On FPGA.
Raza Heidari 2018Recently, solving under determined systems is interested problem in the researchers of several science , especially that is too hard solving problem in general for reaching sparsete solution. In special term , a linear equation system can be converted to an equal convex optimization problem that has just one solution. This improvement is start for many searches and caused sparse solutions.In this thesis, we Propose a novel algorithm for sparse infiniti impulse response (IIR) filter designe. If a filter be sparse the addres and moltiplieres of zero coefficient will be deleted at last cost and complexity of implementation will redused.In recently years a lot of algorithms for sparsing signal introduced. This algorithms 0in 0in 10pt" >In this thesis we propose a novel algorithm for sparse infinitive impulse response (IIR) filter designs. The objective of the sparse digital filter design problem is to reduce the number of nonzero-valued filter coefficient.The proposed design method is i ired by particle swarm optimization(PSO) algorithm.This algorithm is included into two stages ,a preliminary optimization stage is introduced to determine the positions of filter coefficients to be zero then is reduced the error with determined the zero in first step. At last designed Butterworth IIR filter be sparse with six, zero coefficients with pass band edge .3791 , cut of frequency .531 and attenuation -45dB the implemented on virtex6 Field Programmable Gate Array(FPGA).
-
A survey on heat transfer enhancement using magnetic smart material
Mansour Borzouei 2017In this study the effect of ferrite particle in magnetic field on force heat transfer enhancement is investigated. For this purpose an experimental rig has been designed. The effect of different external magnetic field is studied. The effect of reciprocate movement of magnetic field on heat transfer enhancement is studied. The results show that using this system an enhancement of between 16.35% - 36.24% are obtained. Finally, the results are expressed in terms of dimensionless numbers.
-
Forecast e-commerce transactions in social networks
AHMED HASAN OUDAH 2017 -
طراحي و شبيه سازي مدار يادگيري بر پايه اسپايك با استفاده از مدل نورون I&F تعميم يافته و مدار آستروسيت
Milad Almasi 2017As the number of VLSI implementations of spike-based neural networks is steadily increasing, it is important to design spike-based learning algorithms and circuits which are compatible with these systems. The circuits must have adaptation and complex patterns classification capability. Spike-based learning circuits typically are used in conjunction with linear integrate and fire neurons. As a new class of current-mode conductance-based silicon neurons has beenrecently developed, it is important to evaluate how the spike-based learning circuits perform. In this research we propose VLSI implementation of a spike-based learning circuit which consists of a post-synaptic weight control module and a pre-synaptic weight update module. We propose a new circuit for the post-synaptic weight control module which consists of a generalized I&F neuron circuit, Diff-Pair integrator as synapse, voltage comparator, WTA circuit as current comparator and astrocyte circuit. Simulation results also show that the astrocyte performance as a part of current comparator is quiet acceptable.
-
Design and optimization a low voltage, low power LNA using linearization techniques
Mahsa Heidari 2017طراحي يك تقويت كننده كم نويز و بهينه سازي تابع نويز توسط الگوريتم هاي بهينه سازي در متلب
-
Design , Simulation and Fabrication of Wilkinson Power Divider Using Lowpass Filter for Application in Microwave System
Samira Menbari 2017 -
Design of high linear active mixer based on Volterra series analysis
ALI ATAEI SIAH BIDI 2017 -
Weight least square design of sparse FIR digital filter using optimization algorithm
Narjes Dastaran 2017 -
Design and simulation of analog low power DPI (differential pair integrator) synapse circuit
Khatere Teymuri 2016synapses are the essential part for neural network that transmit the information between neurons. The biological neuron transmits electrical or chemical signal between neurons. When the pre-synapse receives a spike, it releases the neurotransmitters and cause to active receptors in the post-synapse. After receive enough neurotransmitters from pre-synapses, the post-neuron reaches threshold voltage, the action potential is occurred in post-neuron. So, with this way the information transmit between neurons The synapse electrical circuit is the most important parts in neuromorphic that transmit the information between neurons. In this paper we propose a modified Differential Pair Integrator (DPI) synapse circuit that can remove problems of previous synapse circuits by producing standard first-order differential equation. The proposed circuit works in sub-threshold region with low power consumption and has biological time constant in low supply voltage (1Volt). This circuit can tune the gain and time constant with three separated controlling voltages and can play a crucial role for learning neural codes and encoding spatiotemporal spike patterns linearly in biological frequency range. This circuit has good performance and advantages in comparison with different synapse circuits, so it is suitable for applications in reconfigurable neuromorphic devices which implement biologically brain circuits. The proposed circuit is designed at 0.35µm standard CMOS technology and simulated by Hspice software. Keywords: neuron; synapse;DPI synapse circuit; sub-threshold; low power;
-
Design and Optimization of CMOS Folded Cascode Op Amp Based on Analytical Method Using MOPSO
Mitra Shahbazi 2016 -
Design & optimization of a ku-band low-power distributed LNA based on An analytical method
Mahsa Omidi 2016 -
Design and Optimization of Mixer Circuits Using Quantum Neural Network Algorithm
Tahere Haghani 2016 -
design of wideband low power low noise amplifier with high linearity
2016 -
Design and simulation of a CS-LNA using quantum Neural Network model
Parastoo Aeini 2016 -
possibility digital implementation of population biologically neurons izhikevich model on FPGA board
Mohamadjavad Hayati 2015 -
A Microstrip power divider 1.4 with different phases designation. simulation and fabrication
HAMED MORADI 2015 -
Design of microstrip bandpass filter for application GSM
Arsalan Khaledian 2015 -
Design of low noise amplifire with desirable linearity
Behnaz Baymani 2015 -
analysis and simulation of diffrent coupling states in a chipless RFID structure to encoding data in wireless systems
2015 -
design of low noise amplifier with desirable linearity
ZAHRA HASHEMI 2015 -
numerical and experimental study of heat transfer using nanofluid in shell and tube heat exchangers
2015 -
A new analytical method for noise optimization in CMOS integrated circuit designs using evolutionary algorithms
Ashkan Haghparast 2015 -
analog simulation of a decoder for brain - machine interface
Mohsen Naseri 2014 -
A neuromorphic VLSI design for the Rulkov neuron model
Afsaneh Shahimaeen 2014 -
pattern recognition by integrate and fire spiking neuronal network
2014 -
Hardware Implemention of artificial cochlea
2014 -
hardware implemention of the tripartite synapse in a neuronal population model
2014 -
survey and simulating of an evolutionary algorithm for placemen optimization in FPGA circuits
Hosna Akbarpoor 2013 -
hardware implementation of a bio-inspire deep brain stimulation in a population of hopf oscillators
Ali Yousefzade 2013 -
design simulation and fabrication of microstrip lowpass filter using fractal structure for communication systems
HESAM SIAH KAMARI 2013 -
hardware implementation of a deep barain stimulation in neural population model
2013 -
Evaluation of environmental of information by moving awarm robots
Pejman Shahhosseini 2013 -
Modeling & simulation of low power/ redundancy trade-off in transistor & logic levels
2013 -
Buffer insertion Optimization TECHNIQUES IN NANOSCALE VLSI circuits to reduce delay in interconnections
Mohammad Bostakchi 2013 -
Behanioral modeling of microwave devices using neural network
2012 -
behavioral modeling of WCDMA receiver using VHDL-AMS
Farzaneh Mohamadi 2011 -
an investigation on single event effects in phase locked loop design
Ehsan Askari 2011 -
design and modeling of low noise amplifier using neural network for blutooth applications
Sedighe Babaei sedaghat 2011 -
modeling and simulation of high frequency transistor in 0.13m SiGe:C HBTs integrated in BiCMOS TECHNOLOGY BY nEURAL nETWORK
Roza Dehkordi 2011 -
modeling and simulation techniques of substrate coupling in mixed-signal integrated circuits
2010 -
modeling and simulation of low voltage RF CMOS VLSI amplifier
2009 -
Design and Simulation Techniques in low power CMOS VLSI circuits
ADEL ALIMORADI 2009
