فصلنامه روشهای هوشمند در صنعت برق
پیاپی 52 (زمستان 1401)

Due to the complexity of distribution networks, preventive maintenance is very important. Incentive regulation is also one of the factors influencing the performance of distribution companies, which in turn complicates maintenance planning. This paper addresses the issue of preventive maintenance planning to enhance reliability in the presence of reward and penalty as a motivational factor. Therefore, the profit function of the distribution company, which includes the cost of repairs and reward-penalty, is optimized. In the incentive model for measuring the performance, reliability indices are compared by feeders, and in contrast, the repair program is obtained for each feeder separately. Due to the different causes of feeder failure such as their structural properties and weather conditions, increasing the accuracy of performance comparison from companies to feeders, in addition to penalties and rewards assigning leads to maximise the impact of maintenance at the level of their reliability and is accompanied by cost savings. For this purpose, the information of a real network including 194 feeders is considered as primary data. After categorizing the feeders and calculating penalties and rewards, the profit from the provision of services are optimized by BPSO method. As a result, the preventive maintenance program is obtained separately for feeders for three general categories of frequent failures, which includes substation failure, line failure and tree branch collision in a period of 5 years. The optimization results show that the proposed method, while maximizing the profits of distribution companies, also improves their performance in terms of reliability.

Advances in power electronic applications, high reliability power distribution systems and flight control propulsion technologies, have increased the utilization of electric power in modern aircrafts. This has led to the concept of more electric aircraft. In this paper, a new structure is proposed to generate the electric power in aircraft. The aircraft power generation and consumption system is modeled as a microgrid. The proposed more-reliable power generation system is based on induction generator, rectifier and voltage source inverter. Induction generator is coupled with turbojet motor. The smart active power sharing between generators is achieved by using improved droop control method. Optimal droop coefficients are assigned by combining mathematical approaches including descending gradient method and minimizing the mean squares of the frequency deviation. The aircraft electrical network is simulated with the proposed controller in MATLAB and its performance is analyzed. The simulation results show the proper power sharing between generators. The advantages of the proposed method over the conventional structure include the elimination of the mechanical constant speed drive which increase reliability and proper power sharing.

In this paper, a novel approach for comprehensive state-space modelling of the grid connected multi-level inverters is proposed. Details of the developed method is presented using cascaded H-bridge converters, however it can be applied to other topologies of the grid connected inverters as well. In multi-level converters, due to their nonlinear characteristic, application of the nonlinear controllers is more beneficial to ensure stability of the system in a wide range of operation. Hence, the state-space model is required to design a nonlinear controller. To achieve converter model, it is divided into some sub-circuits considering different operational intervals in a switching cycle. To verify accuracy and effectiveness of the obtained state-space model, a laboratory setup of a multi-level. Converter with two H-bridges has been designed and implemented. Also, results of the developed state-space model has been compared with the simulation/experimental results of the grid-connected converter.  According to the simulation and experimental result, accuracy of the model is verified. It should be noted that all of the simulations have been performed by EMTDC/PSCAD toolbox.

Memristor is the fourth fundamental element after resistor, capacitor, and inductor. Memristor can become an essential element of SRAM and DRAM caches because of its zero power consumption in data storage and non-volatile state. It can effectively improve the efficiency, speed, and power consumption of circuits. In this paper, we propose a 4T1M memory cell reducing the cell area by maintaining the maximum properties of 6T1M. To simulate the proposed memory cell, the length of the memristors is 10 nm, and the resistance of their on and off states is selected as 1 kΩ and 200 kΩ, respectively. Also, the cell MOS transistors are simulated by the 32 nm HP CMOS PTM model. Simulations in H-Spice software, at 0.9 V power supply, have been conducted to compare the proposed cell characteristics with two conventional six-transistor (6T) and six-transistor one-memristor (6T1M) cells. The results show that using a memristor in a memory cell causes zero power consumption during data storage for a long time and reduces the occupied area by 36.7% compared to the 6T1M cell. The speed of writing “1” data on the proposed cell is only 30 ps, which shows a 3-fold improvement compared to the 6T1M cell, but no significant change is observed when writing “0” data. The static power of the proposed cell is 133 times less than that of a six-transistor cell, and its dynamic power is about the same as the 6T1M cell, but it consumes 60 times less energy than a six-transistor cell.

Alzheimer’s disease is one of the most common diseases in the 21st century. Alzheimer's patients lose their brain cells gradually and eventually die. It is often diagnosed when the symptoms appear and little work can be done for the patient. Using of learning algorithms is useful for diagnosing of Alzheimer. Previous studies used Support Vector Machine, K-Nearest Neighbor, and Linear Discriminant Analysis in order to diagnose the disease. These methods have some problems such as low accuracy, high computation complexity or high execute time. Therefore in this research, a method based on brain emotional learning and wavelet feature is used. First, the white and gray matters of the brain were separated by a threshold selection method. Second, the texture properties of the images were extracted by wavelet transform algorithm. Third, the dimensional reduction is done on the properties extracted by principal component analysis. Finally, the features were classified using Brain Emotional Learning Algorithm and Brain Emotional Learning Based Pattern Recognizer. Results showed that run time of brain emotional learning algorithm is 0.22 second and Brain Emotional Learning algorithm with 95% accuracy and Brain Emotional Learning Based Pattern Recognizer with 97% accuracy are better than Support Vector Machine with 83% accuracy.

Speech emotions recognition due to its various applications has been considered by many researchers in recent years. With the extension of deep neural network training methods and their widespread usage in various applications. In this paper, the application of convolutional and transformer networks in a new combination in the recognition of speech emotions has been investigated, which is easier to implement than existing methods and has a good performance. For this purpose, basic convolutional neural networks and transformers are introduced and then based on them a new model resulting from the combination of convolutional networks and transformers is presented in which the output of the basic convolutional network is the input of the basic transformer network. The results show that the use of transformer neural networks in recognizing some emotional categories performs better than the convolutional neural network-based method. This paper also shows that the use of simple neural networks in combination can have a better performance in recognizing emotions through speech. In this regard, recognition of speech emotions using a combination of convolutional neural networks and a transformer called convolutional-transformer (CTF) for RAVDESS dataset achieved an accuracy of %80.94; while a simple convolutional neural network achieved an accuracy of about %72.7. The combination of simple neural networks can not only increase recognition accuracy but also reduce training time and the need for labeled training samples.

Detecting and controlling stress levels in drivers is especially important to reduce the potential risks while driving. Accordingly, in this study, a detection system was presented to identify four levels of stress (low, neutral, high and very high) in drivers based on physiological signals. The proposed method used the drivedb database, which includes the recording of physiological signals from 17 healthy volunteers while driving on specific routes on city streets and highways. A set of statistical and entropy features along with morphological features that were calculated only for the ECG signals, were used. The calculated features were applied as inputs to the classification units to detect stress levels. Support vector machine (SVM), k nearest neighbors (kNN) and decision tree (DT) were evaluated as classification methods. The main purpose of this study was to improve the accuracy of stress level detectionusing the idea of classifiers fusion. To achieve this goal, the combination of individual classification units, each of which used only the features of one of the ECG, EMG and GSR signals, was performed by the Demster-Shafer method. Using genetic algorithm as feature selection method, SVM classifier and Dempster-Shafer fusion strategy, the best stress detection accuracy of 96.9% was obtained. While the highest detection accuracy among individual classifiers was 75% and obtained by a subsystem that used ECG features.The results show significant performance of the proposed method compared to previous studies that used the same dataset.

The interleaved boost converters are the power circuits that provide high-voltage, high-power with regulated output voltage for renewable energy systems which are generally suffer from low-voltage and unregulated output voltages. The soft switching methods reduce electromagnetic noises and switching losses in these converters. In this paper, a ZVT interleaved boost converter with a simple auxiliary circuit is proposed. The proposed converter has a simple structure with low size and cost. In the proposed converter, soft switching condition is provided without any extra voltage and current stress on the main switches. The auxiliary circuit comprises two diodes and one auxiliary switch. The leakage inductance of the utilized coupled inductors is used as resonant inductor. The auxiliary switches benefit from significantly reduced voltage stress without requiring floating gate driver. The proposed converter can achieve zero voltage switching operation for the main switches and zero current switching for diodes and auxiliary switches, which causes to alleviate the reverse recovery problems of all diodes.

A new design of a dual-polarized base-station antenna with a wideband, low profile and high performance is introduced in this work for the LTE700/GSM850/GSM900 applications. The proposed base-station antenna in this study is comprised of four single-polarized magneto-electric dipoles (MEDs) antenna are positioned with a square arrangement to produce ±45° slant polarization. Each antenna is involved with an electrical dipole, a Γ-shaped feed structure, a magnetical dipole, a metallic resonator, and a cylindrical-shaped reflector. Three metallic portions with a slit produce the electrical dipole. Adjusting the dimensions of these segments and resonator plate improve the antenna impedance bandwidth. Also, the cylindrical-shaped reflector increases the antenna gain and has a significant rule to stabilize the antenna radiation pattern. The measurements exhibit that this antenna achieves an expanded frequency bandwidth of 38.1% (686.2-1008.8 MHz) for |S11| < –15 dB, realized peak gain of 11.45 dBi, low cross-polarization, and half-power beamwidths (HPBWs) of approximately 60.4°, 64.7°, and 66.8° at frequencies of 700 MHz, 8500 MHZ, and 900 MHz respectively. Results approve that the above-mentioned antenna is applicable for mobile cellular networks systems.