فصلنامه مدل سازی در مهندسی
پیاپی 63 (زمستان 1399)

Recent advances in information technology and the need for more security have led to the rapid development of intelligent biometric identification systems. Recent studies have proven that handwriting of each person is unique and can be used as one of the authentication methods. There are many researches in the literature for writer identification on a specific language. Unfortunately, there are no necessary data sets for this purpose. In this paper, for the first time, a handwritten data set of 300 persons in both Persian and English languages was collected. The main goal of this paper is to provide a model to identify the writer independent of the language written in Persian and English. After pre-processing stage, each person's handwriting is converted into blocks of a certain size called a texture. Then, using these textures, the desired features are extracted. In order to extract these features, first a two-dimensional wavelet transform is applied to each image and then, using the new algorithm for calculating the fractal dimension, which is used for the first time in this field, the feature vector is obtained. Finally, MLP neural networks are utilized for classification step. The performance of the proposed method is evaluated in different scenarios.

Reinforced concrete shear wall structures are an important part of the construction due to their ability to withstand in large earthquakes force and deformation demands. In some of these buildings, shear walls are implemented in successive bays to create more lateral resistance, to reduce drift or to satisfy architectural requirements, and adhered walls are appeared. In Open Sees, the MVLEM and SFI_MVLEM methods do not have the ability to model the common boundary element between two shear walls adhering to each other. So, modeling of these walls in Open Sees software is difficult. In this paper, we propose a method for modelling these walls, Multiple Vertical Columns (fiber) method, and modeling results are compared with experimental results and the results of other models. The ability of this method in the modeling of the adhered shear wall is measured via the actual building modelling. The results of static posh-over analysis and nonlinear dynamic analysis show that the proposed model has a good performance for the modeling of adhered walls. The main advantages of using Multiple Vertical Columns (fiber) method are the use of multiple fiber columns to model the geometry continuity of the wall in a horizontal direction (similar to the MVLEM) and the possibility of folding the vertical elements along their longitudinal axis and using horizontal joints in the mid-points to model The geometric continuity of the wall is vertical direction.

Steel sheets are used in various industries such as automotive, aerospace, fuel transport and storage tanks, metal bridges. The issue of impact and penetration in steel sheets has been the main focus of many studies. In this study, laboratory studies of flat and curved steel sheets, as well as reinforced by hardening perpendicular to sheet plate. Under the influence of impact caused by the free fall of the weights are examined. In this study, three radii of infinite curvature (flat sheet), 300 mm and 110 mm were used. This study was conducted for three levels of impact energy (three free fall crashes). Steel sheets used are steel st.12 with dimensions of 220 * 200 mm and 1 mm thicken, fastened in one direction by the fixture and is free on the other side. The material and thickness of the pendentive used are similar to the sheet and have a width of 2 cm. The results also show that the use of hardening causes a slight increase in the acceleration of the sheet and a significant reduction in its permanent deformation. The amount of energy absorbed by the reinforced sheets is slightly less than that of plain sheets. In this research, the sheet with a curvature radius of 110 mm with a pendentive has a better performance than the rest of the sheets.

In this paper, the design of penternary circuits based on graphene nanoribbon FET (GNRFET) is presented. The employed logic of the penternary corresponds to the Galois logic. The HSPICE-compatible model and 15-nanometer technology have been used to simulate the graphene nanoribbon transistor. Accordingly, the proposed NAND and NOR penternary circuits are first, designed and simulated. The results show that these proposed circuits have a significant improvement in terms of speed and power consumption compared to their CNTFET counterparts. Then, the adder circuit as the main part of digital processors in integrated circuit design is proposed with penternary logic. The transient responses of the proposed circuits are accurate. Parameters such as power consumption,, delay and power-delay product are calculated. Evaluation of the results shows that the proposed adder circuit has the power-delay product (PDP) of 179.39 fJ at the supply voltage of 0.8 V and the operating frequency of 100 MHz.

As time passes, some elements of the structures are affected by local damage, though insignificantly. With the development and expansion of the damage , these structures may be completely destroyed and impose high socioeconomic costs. This study investigated the damage identification problem in the column under the axial load using modal data (frequencies and mode shapes) and wavelet transform analysis. The obtained results showed that by increasing the axial load proportional to the critical buckling load, the frequency value decreases in all modes in healthy and faulty modes. Additionally, under the effect of the same axial load, the frequency of the faulty sample is less than that of the healthy one, and as the damage severity increases, the rate of frequency reduction increases. The wavelet transform input signal was defined on the basis of the angle between the healthy and faulty mode shapes. The output signal details in the damage locations indicated perturbation so that in all the studied modes in different ratios of critical load, the damage locations were identified with high accuracy. Moreover, the results from the research showed that the perturbations in different damage locations are independent of each other and are only affected by the severity of the location damage and the axial load has no effect on the damage detection sensitivity of the wavelet transform algorithm. Furthermore, the sum of the wavelet coefficients of damaged locations in several damage modes is equal to the wavelet coefficients of damaged locations of the sum of damage modes.

Due to the significant share of cooling loads in the peak of electricity consumption and its increasing trend in recent years (up to about 40% of the peak of electricity demand), the draft of the new edition of the National Building Regulations (Article 19) enforces several requirements for managing these loads. One of these requirements is the use of ice storage systems in buildings with high cooling loads. In order to evaluate techno-economic aspects of the implementation of these requirements, the cooling load of a pilot building with educational use in Ahvaz was simulated by Carrier HAP software, and based on two different storage strategies (i.e. full and partial storage), technical operation and necessary economic investment were analyzed. The payback period of the surplus capital required for the implementation of storage systems was determined with respect to the saving potential in electricity costs. It was determined that the economic burden of these requirements is a function of the storage strategy, building function and climate. Accordingly, in the considered building, while the use of partial storage is quite economical, the return on investment for a full storage system will be more than 50 years. The results of examining different scenarios and possible base electricity tariffs showed that a fivefold increase in the tariff will reduce the payback period to about 10 years.

Sustainable diets are important because of their less environmental impact, as well as having the micro-nutrients needed for healthy living in the present and future generations. Athletes always need special food programs to maintain their level of activity. Therefore, in this research, a mathematical model with three objective functions for the athlete's diet plan, consumed to taking into account the principles of sustainability, includes the objectives of 1- minimizing the cost of purchasing food, the cost of water consumption and the cost of producing carbon dioxide for food, 2- Maximizing the amount of protein intake and 3- Minimizing the amount of sugar and fat. For model validation, real information about nutrients and supplements required by the athletes' body has been used. The model was formulated in The GAMS computational software and solved by Cplex solver. The outputs of the model represent the performance of the model in the actual example.

Permanent magnet synchronous motors due to its high efficiency and power density, reliable performance and simple construction, industrially used. One of the problems with these Motor need accurate information to control its speed and position. Recently, because of the difficulties of the speed sensors, speed estimation is used instead of measuring it. In this paper, the state dependent model reference adaptive controller based on pseudo linearization is used to control the permanent magnet synchronous motor that its parameters are determined based on Lyapunov theory. This controller show good results by production control law, despite changing circumstances and maintain system stability despite external disturbances. Also due to uncertainty in the estimation of the position and speed of motor parameters, high impact, online identification of these parameters is necessary. In this paper, adaptive augmented observer is used to estimate speed and identify motor parameters online. The main advantage of this estimator is that speed and load torque estimation and identification of parameters are simultaneously thus volume and time calculations are reduced. The simulation results show convenient tracking of desired speed with the load torque, changing the motor parameters and the speed reference.

In this paper, A measurement-based model is proposed to assess the vulnerability of the transmission line under overloaded cascading blackout analysis in the power system online operation environment. The proposed Measurement-based model is constructed by the Artificial Neural Network due to its ability in nonlinear mapping between input and output vectors that its ability causes a suitable prediction. The Artificial Neural Network Training data set is provided by using analytical vulnerability assessment model in different operational condition to rank and obtain vulnerability status of each transmission line. Then, Arterial Neural Network links between operating conditions as the input vector and the vulnerability value of each transmission line as the output vector. The efficiency of the proposed measurement-based model in terms of speed and accuracy is investigated in the IEEE 39-, and IEEE 118-bus test case systems by comparing it to an analytical vulnerability assessment model. Finally, the security and reliability of the transmission network are enhanced by increasing the online situational awareness of the operator about the effects of each transmission line in propagating the cascading failure by using the proposed model.

In this study, using a sample volumetric element, the damage behavior of composite materials has been investigated using representative volume elements (RVEs). The two damage modes of fiber-matrix debonding and matrix cracks are modeled and analyzed as dominant modes. A new algorithm is developed to generate a random distribution of fibers in the RVE, and it is possible to create a fiber distribution with a normal distribution of fiber radii. The fiber material is considered linear elastic and the elastic-plastic matrix behavior is modeled with Drucker–Prager's criterion. The fiber-matrix debonding is modeled by cohesive elements which lead to matrix cracking. Initially, the random distribution of the fiber in RVE was created based on the assumed geometric dimensions and the fiber volume percentage with the code developed in MATLAB software. then, loading, periodic boundary condition, and homogenization process have applied by using the Abaqus software. The results showed that different mesh size is needed for RVEs with various fiber volume fractions. Moreover, the effects of different cohesive zone parameters are described in detail.

Gastric cancer is the first rank of cancer in Iran. If the disease is detected in early stages, treatment is probability to be furthered and treatment costs will be reduced. Due to the complexity of pathologic images and the fundamental challenges in these images, such as the poor contrast between the cells, cell overlapping and the contradiction in tissue coloring, the process of diagnosing this type of disease is difficult and therefore needs a proper method to eliminate these Problems. In this research, a smart model is proposed to solve these problems. Then a fuzzy-based system, a discrete wavelet transform, The region's growth and the voting mechanism are used to identify the cells. Then, an advanced morphological method was presented for separating overlapping cells. Then the cell's feature were extracted and based on it, the cells are classified using the support vector machine algorithm (SVM) with the RBF kernel. The proposed algorithm was applied to a dataset of patients including 96 Microscopy Images from Baghiyatallah Hospital in Tehran. The proposed model was evaluated using the ROC curve analysis. The results were approved by expert pathologists and reveal accuracy of 92.12% in detection of normal and cancerous cells and 94.14% in detection of benign and malignant cells which are promising for early diagnosis of this type of cancer.

In this study, a simulation is developed for dissimilar metal sheets followed by a numerical model for laser welding. The continuous disk laser welding process for dissimilar joining between 304 stainless steel and copper is simulated. For this purpose, double conical heat source distribution model was implemented into Abaqus/Standard solver using additional DFLUX subroutine written in the FORTRAN programming language in order to estimating the temperature distribution, prediction the geometry and dimensions of weld cross section including the fusion zone (FZ) and heat affected zone (HAZ). The weld joint output parameters such as temperature, strain distribution, residual stresses and weld geometry were examined. The validation of the simulation output results were confirmed by comparison with the experimental results by other researchers at three level with three input parameters. The results showed that copper, due to the higher thermal conductivity during cooling, passed through more heat at the same period of time, so these results lead to lower temperature in the center of FZ and a smaller HAZ in this dissimilar welding in comparison to similar welding. After cooling time, longitudinal residual stresses in the FZ for both materials were tensile also due to higher thermal expansion coefficient of steels, the quantity of stresses are 23% more than copper. The computed results are well agreed with experimentally measured values and show the robustness of the present numerical model used for dissimilar laser welding.