فصلنامه مدل سازی در مهندسی
پیاپی 80 (بهار 1404)

Today, due to the concern of emission and greenhouse gases, generation of a clean energy and using renewable energies in the best way (with high efficiency) is a very important issue. Although economic goals have always been more important than environmental goals. In this paper, more considerations have been made in order to consider the environmental issue in the proposed optimal scheduling of the system in the micro energy grid. This optimal system studies the energy hub system, which is the main part of the micro-energy grid, in the form of a CCHP-based network combined with renewable energies. This system uses three energy hubs, energy storage, and converter devices. Therefore, in this paper, a multi-stage planning framework is proposed for the energy hub system and to optimize its performance, including reducing emission and operational cost. In this model, upper and lower limits are considered for the power produced by renewable energy sources to indicate the possibility of power interruption due to their fluctuations. Also, by considering multiple objective functions, optimal decision conditions can be guaranteed for the decision operator. To solve the multi-objective problem, the augmented epsilon-constraint method is used. In addition, two optimal decision-making methods have been proposed and compared too. The results obtained after the implementation of the proposed model show the efficiency of the model in reducing the cost and environmental emission.

Domestic refrigerators have a noticeable contribution to energy consumption worldwide. Thus, it is beneficial to integrate control strategies for reducing their energy usage and improving their performance. The traditional hysteresis controller cannot perform effectively in the presence of uncertainties such as daytime or refrigerator door status. Additionally, most of the designed fuzzy controllers based on temperature error, do not address the door-opening behavior of the user. Hence, this paper employs fuzzy logic systems to improve the performance of domestic refrigerators. Based on the data collected from domestic users' door-opening events, the daytime is split into four time intervals and a fuzzy system is then designed to adjust the desired temperature of the refrigerator. Also, a fuzzy controller is designed to control the compressor speed according to the temperature error and the time of the day. A MATLAB refrigeration model is used to evaluate the proposed controller’s performance. The results demonstrate lower energy consumption and better temperature set-point tracking for the proposed controller than a fuzzy controller without addressing user behavior in the fuzzy rules design.

Differential protection is one of the most important and basic protections for power transformers. In normal conditions, the differential current is almost zero, but it increases during internal faults. Since the increase in the differential current can have other reasons than the occurrence of an internal fault, the existence of a differential current is not necessarily a good sign of the internal fault occurrence. Some phenomena, such as the inrush current, saturation of current transformers, and over-excitation, also result in the appearance of the differential current. In the presented protection scheme, the differential current signal is analyzed by the discrete Fourier transform. In the first stage, based on the point that the rate of changes in the magnitude of the second harmonic of the differential current in the internal fault with current transformer saturation is much greater than the inrush current, these two events can be distinguished. In the second stage, based on the fact that the ratio of the second harmonic to the fundamental component of the differential current drops rapidly after the occurrence of internal faults without current transformer saturation and reaches almost zero, it can be easily discriminated from other conditions. Various simulations have been performed for proper threshold settings for the above-mentioned criteria to achieve maximum separability. The results obtained from simulation studies on a real 230/63 kV, 160 MVA power transformer reveal that the proposed algorithm remains stable for external faults and transformer energization conditions even during current transformer saturation.

The use of mobile phones with Android operating system is expanding day by day. Android itself does not have a powerful malware detection tool. Therefore, attackers easily enter people's privacy through their mobile phones and put them at serious risk. So far, a lot of research has been done on malware detection. One of the main problems of these solutions is the low accuracy in multi-class detection on the dataset or the failure to achieve the desired result in both types of binary and multi-class detection. In this paper, by using Convolutional Neural Network (CNN) and changing the number of different layers, we have tried to extract the maximum number of important features from the dataset. In the data classification phase, we use the Deep Learning-based algorithm named Long Short-Term Memory (LSTM) to classify the data with the maximum possible accuracy by testing it on the selected features. The test results on the new MalMemAnalysis-2022 dataset show that the use of these two algorithms and the change in the number of layers can lead to 99.99% and 99.71% accuracies in binary and multi-class classification in malware detection, respectively, which is superior to existing methods.

One of the most powerful methods for solving nonlinear equations is the Newton-Raphson algorithm. Although this algorithm is highly efficient, it faces two challenges: sensitivity to the starting point and the possibility of getting stuck in a loop for the solution sequence. In this paper, by adding a small disruptive term to the Newton recursive relation, we practically generate a chaotic pseudo-random sequence, thus addressing these two challenges. The effectiveness of the proposed version has been numerically demonstrated on several nonlinear equations. The results show that the improved version is not sensitive to the initial conditions and escapes from the loop due to the generation of a random sequence, while maintaining an acceptable execution time due to the use of a deterministic system in generating the random sequence.

In this paper, a novel DC-DC buck-boost converter with a wide conversion ratio is introduced. With the advantages of the ZETA converter, such as continuous input current and positive polarity of the output voltage, the proposed buck-boost converter is well suited to industrial and renewable energy applications. With a voltage gain higher than the ZETA converter, the proposed converter also reduces input current stress due to its continuity. Two operating states are available for this converter in continuous conduction mode (CCM). This converter has two switches that operate simultaneously, and it can be controlled easily. As a result of the converter's continuous output current, output voltage ripple and output capacitor current stress are reduced. Computational analysis, small signal modeling and efficiency of the proposed converter considering the influence of parasitic elements are presented in this paper. This converter has also been compared with other similar and recently presented topologies. Finally, a prototype is implemented to verify the accuracy of the computational analysis, and simulation and experimental results are presented.

This article presents the open switch fault detection and diagnosis of neutral point clamped inverters using neural networks. The neutral point clamped inverter consist of four power semiconductor switches with two clamp diodes in each phase. As the number of power semiconductor switches in the inverter increases, the probability of fault occurrence significantly increases. Therefore, for ensuring uninterrupted operation and preventing additional damage during faults, rapid fault diagnosis is essential for the fault tolerant systems. In this study, a neural network based system is developed for the detection and diagnosis of open switch faults in one or two power semiconductor switches in neutral point clamped inverter. The proposed method relies on measuring the inverter output currents and calculating their average values, which are then used for training the neural network. The proposed system covers open switch fault diagnosis for one or two switches. The mentioned system was simulated in the MATLAB/SIMULINK environment, and the result demonstrate accuracy and rapid fault diagnosis for semiconductor power switches within a fundamental cycle.

In this paper, a new reduced-order average model (RAVM) for a DAB converter is presented. The model can be applied to all modulation techniques, such as Single-Phase-Shift (SPS) modulation, Extended-Phase-Shift (EPS) modulation, Dual-Phase-Shift (DPS) modulation and Triple-Phase-Shift (TPS) modulation. In the proposed model, conduction losses, inductor losses, transformer losses, switching losses with input/output filters are considered to improve the accuracy of modeling. This article employs a virtual resistance for switching power losses to model the large and small signals of the dual active bridge converter with triple-phase shift modulation. Using this virtual resistance enhances the accuracy of modeling and better estimating converter efficiency and losses, especially at high switching frequencies. The introduced model is successfully implemented in PLECS software and then large and small signal analyses are curried out. Comparison of simulation results under different operating conditions show a suitable accuracy of the proposed model in tracking the behavior of the DAB converter compared to other methods, without considering all the significant elements affecting the converter's performance.

The purpose of this research is to develop an integrated control algorithm system to track the vehicle path in the double lane change scenario with critical and emergency conditions. The proposed control system includes two layers. The first layer is responsible for controlling the motion of the vehicle and benefits from the implementation of two different types of controllers in the proposed algorithm system. In the first layer, the sliding mode control algorithm technique is used to control the longitudinal movement of the vehicle, and the task of controlling the lateral movement of the vehicle is the responsibility of the controller based on model prediction control. The second layer includes an optimal distribution function for allocating rotational torque to the four vehicle tires. In order to consider the real behavior of the vehicle, the nonlinear dynamics of the tire is considered. The proposed control algorithm is analyzed and investigated in different scenarios with different working conditions and critical and emergency conditions. The results of the performed simulations show the optimal and effectiveness performance of the proposed control algorithm. Also, MATLAB/Carsim software is used to validate the performed simulations.

The available melanoma skin cancer dermoscopy databases have low and unbalanced images with non-uniform illumination that make melanoma detection methods challenging. To address these problems, in this paper, we propose a new method to generate new data including melanoma. In fact, our new proposed method combines generative adversarial network and local binary pattern. On the other hand, first, the images existing in the dataset are fed to the generative adversarial network. Then, many new images are generated and finally, the local binary pattern is applied to them. Therefore, the number of the new generated data is large and balanced and the generated data does not have illumination changes. Also, these data show useful and meaningful features that increase the difference between melanoma and nevi. The experiments have shown that the proposed method has a good effect on increasing the accuracy of melanoma diagnosis. According to the results, the proposed method has increased the convolutional neural network's efficiency by 7%.

A liquid marble is a droplet that is surrounded by hydrophobic micro or nano-sized nanoparticles. By using these marbles, the challenges and problems related to droplets are solved. These liquid marbles are very responsive to external magnetic force when using a magnetic coating around the inner droplet or using magnetic contents. In this research, first in the experimental part, the transfer of a magnetic liquid marble is carried out under the influence of the magnetic field of a permanent rectangular magnet. In the next step, in the simulation section, by modeling the magnetic liquid marble by a droplet with an elastic shell, the effect of gravity in the movement of the marble downwards on a curved ramp is done with the finite element method. This liquid marble is used on a ramp as a liquid marble-based accelerometer. In the following, the profiles of total displacement and stress, speed and acceleration of this accelerometer are extracted.

The present study is designed with the aim of using a new and accurate method to model and evaluate possible scenarios of fire and explosion in the biogas power plant of Mashhad. In this research, the potential scenarios of leakage and rupture of the equipment were identified using the HAZID method and modeling of the possible consequences was done by considering all the influencing parameters using the PHAST software. In this study, three scenarios with leakage sizes of 5, 25 and 130 mm (complete rupture of pipes) were determined to model the consequences of the power plant. The results of modeling the consequences of the 5 mm leakage scenario showed that the radiation intensity of the explosive fire caused by the release of materials was insignificant (maximum 0.03 kW/m2) and the maximum will continue up to a distance of approximately 2 m. The amount of radiation of this fire in the 25 mm leakage scenario is nearly 2 kW/m2, and its radiation will continue up to a distance of 9 m from the accident site. While the pipeline rupture scenario (130 mm leak) leads to a significant fire eruption. The radiation of this fire in this scenario is close to 26 kW/m2 and its radiation will continue up to a distance of 30 m from the accident site. This amount of radiation is enough to cause serious damage to the people who are in this area.

In this article, a new structure of the wireless MEMS condenser microphone has been designed, analyzed and modeled. In the proposed structure, a microphone with  low pull-in voltage is used, which is in series with a Mems inductor. The internal inductor is in mutual induction mode with an external inductor and provides the required voltage for the microphone. The information received by the microphone is also transmitted through mutual induction. The sensor is designed for use in hearing aids for the deaf. The dimensions of the microphone are 250×250μm2, which is located in the middle of the Mems inductor with dimensions of 2×2mm2. The microphone and the inductor have been analyzed separately and then modeled using compact electronic elements and the formula required for their design has been calculated. The simulation of Mems microphone and Mems inductor has been done in Intellisuite software and their circuit simulation has been done in ADS software. The sensor is very low-noise and low-power, and its signal-to-noise and power consumption per input voltage of 100 mv is equal to 79.6 db and 204 mW. Due to the absence of batteries, the dimensions of the proposed sensor are small and suitable for medical and security applications.

The subject of this article is the design and modeling of an electrodynamic sensor with improved circuit specifications to measure the charge of moving particles. This sensor can be used to observe magnetic field changes, non-intrusively, to measure velocity, density, and distribution parameters of moving particles in industrial pipelines and conveyors. The two main components of the sensor are the transducer and the signal conditioning circuit. The transducer is made of a coin-shaped metal electrode and a transimpedance amplifier. The electrode design process involves obtaining the magnetic sensitivity model of the electrode and calculating the induced current. In addition, the signal conditioning circuit components, with upgraded electronics, include sequential voltage amplifiers and a noise reduction filter that are based on transducer modeling results. The results of the experiments and simulations of the final design showed that the sensor can detect a wide range of magnetic field amplitudes produced by moving particles. Also, the sensor has high sensitivity and good circuit stability and can be a good choice for use in measuring instruments.

Slug flow is a part of an intermittent flow that is avoided in industrial applications because of its irregularity and high-pressure fluctuations. The present investigation aims to comprehend the behavior of intermittent slug flow with the gate valve and its associated pressure drop, slug liquid holdup, slug frequency, and averaged slug body length using CFD simulation studies. To achieve this objective, numerical investigations of slug flow were carried out on a horizontal pipe 10 m in length and 20 mm in diameter with three distinct area contraction ratios. First, the numerical results (slug frequency) were validated by comparison with other studies, and a reasonable agreement with an error of less than 3% was achieved. Second, the numerical results were validated by comparison with experimental data in terms of time and space for slug flow encountering the gate valve, and a sensible agreement with an error of less than 15% was achieved. Eventually, pressure gradient, slug frequency, slug length, and the existence of a gate valve have been investigated using numerical methods. The results showed that the slug frequency increases by 4.3% and this phenomenon is due to the small waves that occur due to the high pressure drop at the valve location. The length of the liquid slug increases by 20.89 percent by reducing the opening of the gate valve and decreases by 30.4 percent by increasing the surface velocity of the gas. By increasing the distance from the gate valve, the pressure drops decrease by 15.65%.

The increasing trend in the use of multilevel converters for high-power applications has emphasized the significance of reliability and fault tolerance in these systems. In this research, Model Predictive Control (MPC) has been developed to generate the switching commands for the cross-switched inverter. The performance of the inverter is investigated under conditions where an open-circuit fault occurs in any of its switches. Additionally, an approach has been proposed to enhance the utilization of the capacity of the inverter in post-fault operation. For this purpose, first, a fault detection method is presented to identify the open-switch fault by measuring the deviation of measured voltage from the reference voltage. Then, by defining fault indices and monitoring their conditions during one cycle after fault detection, the faulty switch is also identified. This method uses the minimum number of voltage and current sensors. Also, to achieve the maximum possible output voltage under post-fault conditions, it is proposed to embed a cross-switched cell with capacitive links in the converter structure as an auxiliary cell. Finally, the accuracy and effectiveness of the proposed methods are verified through the simulation of a nine-level cross-switched inverter in the MATLAB/SIMULINK environment.

Energy storage systems (ESSs) can be installed in microgrids and used for reserving and feeding loads. These systems provide a wide range of applications to the power grid, such as reducing the problems of fluctuating and outages of renewable energy sources, load compliance, voltage and frequency stability, peak load management and improving power quality. Also, as production resources in daily planning, a lot of profit is obtained from the energy exchange of the microgrid with the main grid. Considering the high investment costs of ESS, in this article, to justify the economy and prevent its under- or over-utilization, a precise model is presented to determine the optimal size of the storage device. Moreover, to consider the uncertainties of the photovoltaic system, wind turbine, and electric loads, Monte Carlo simulation has been used to generate scenarios and the K-means algorithm to reduce them. However, in order to find a solution to reduce the grid vulnerability and improve its technical and economic performance, it is crucial to pay attention to reliability and resilience. ESSs lead to better energy management during peak hours and when disturbances occur. The model presented in this article examines the role of ESS in energy systems to reduce operating costs, and improve network resilience and reliability. Resilience measure that is used to reduce the effects of severe incidents on the network is considered as a term of the objective function. The system reliability index, which is to ensure the reliable operation of the network against small errors and short-term failures, is proposed as a constraint in the model. An accurate and practical ESS model improves the performance of the system in terms of economy and security, and the simulation results show the efficiency of the presented model.

In power systems, power switches of all kinds are known to be a source of transient overvoltage. Meanwhile, compared to other switch types, vacuum circuit breakers (VCBs) cause more severe overvoltage. Customers' electrical equipment may be damaged by overvoltages that arises from the power switch switching in the medium-voltage distribution network and then reaches the low-voltage distribution network through distribution transformers. This article examines how different factors affect the degree of overvoltage brought on by switching in medium-voltage distribution networks and how those overvoltages enter the low-voltage distribution network. This article explores the factors that influence the severity of overvoltages caused by switching, including the dielectric recovery power of the switch, the amount of current chopping, reignition, and prestrike, and the transformer's location in the medium-voltage distribution network. This study also examines the efficacy of installing a surge protection device (SPD) at the distribution substation to suppress transient switching overvoltages that penetrate to the low-voltage distribution level. The EMTP-RV program was used to conduct these investigations. The research findings indicate that installing SPD at the secondary side of the distribution transformer can suppress transient overvoltages brought on by the 20 kV switch's switching, which enters the low-voltage distribution network.

This paper introduces a new adaptive load frequency control method based on model predictive control using an improved unscented Kalman filter. This research aims to design a robust controller for changes in the frequency characteristic of power systems. An adaptive model predictive controller with online updating of the predictive model is designed to reproduce the fundamental frequency characteristics of the target system. A simplified predictive model is identified in the preliminary investigations of the closed-loop performance of the power system. Then, the parameters of this model are estimated according to real-time measurements. In addition, the control sequence is approximated using Laguerre functions, which helps to reduce the computational load further. A three-zone test system in the presence of generation rate constraint and dead-band is considered for simulation to prove the effectiveness of the proposed scheme. The simulation results show that the proposed adaptive control scheme can successfully stabilize frequency deviations caused by load disturbances and is robust against parameter uncertainties.

Text clustering is a method for separating specific information from textual data and can even classify text according to topic and sentiment, which has drawn much interest in recent years. Deep clustering methods are especially important among clustering techniques because of their high accuracy. These methods include two main components: dimensionality reduction and clustering. Many earlier efforts have employed autoencoder for dimension reduction; however, they are unable to lower dimensions based on manifold structures, and samples that are like one another are not necessarily placed next to one another in the low dimensional. In the paper, we develop a Deep Text Clustering method based on a local Manifold in the Autoencoder layer (DCTMA) that employs multiple similarity matrices to obtain manifold information, such that this final similarity matrix is obtained from the average of these matrices. The obtained matrix is added to the bottleneck representation layer in the autoencoder. The DCTMA's main goal is to generate similar representations for samples belonging to the same cluster; after dimensionality reduction is achieved with high accuracy, clusters are detected using an end-to-end deep clustering. Experimental results demonstrate that the suggested method performs surprisingly well in comparison to current state-of-the-art methods in text datasets.