فصلنامه هوش محاسباتی در مهندسی برق
سال چهاردهم شماره 3 (پاییز 1402)

The phenomenon of Fault-Induced Delayed Voltage Recovery (FIDVR) appears in networks with high penetration of induction motor loads because the increase in requested reactive powers of motor loads after clearing the fault prevents the rapid return of the bus voltage to the pre-fault level. Load shedding is one of the effective ways to deal with the FIDVR phenomenon, which causes the amount of demand to approach the production of reactive power. In this paper, a wide-area load-shedding method is presented, which performs based on network conditions and loads. Since the introduced indicators for determining the locations and amounts of loads to be shed are based on the values of bus voltages, loads currents, and network impedance matrix; therefore, the proposed method can effectively shed the loads and deal effectively with FIDVR. The voltage estimation process is an important tool to predict the voltages at future moments and is defined based on the modified Gauss-Seidel load flow and the three-order model of the induction motor. This tool enables the proposed method to understand the effect of applying load shedding on voltage recovery and prevents the application of unnecessary ones.

Islanding is one of the important challenges in power networks in the presence of distributed generation which is considered an undesirable incident due to the possibility of damage to operators, network equipment and consumers. Therefore, it is necessary to quickly detect islanding and make a decision regarding the connection status of local distributed generation units in the network. In this paper, a passive islanding detection method is proposed using adaptive threshold-based instantaneous frequency droop characteristic. The proposed threshold limit is dynamically changed depending on under studied conditions in order to discriminate the island from load changes, capacitor bank switching, motor start-up and short circuit types. Two medium voltage networks of Cigre and 34-bus IEEE are used to evaluate the proposed method. The simulations are performed in Digsilent software and implementation of the proposed method carried out in MATLAB software. The simulation results indicate that the island is correctly detected with appropriate accuracy in a short period of time from other disturbances in the presence of wind, solar and diesel types of distributed generations.

Traditionally, photovoltaic power plants use Maximum Power Point Tracking (MPPT) algorithms to produce maximum electrical energy. However, with the increase of grid-installed photovoltaic power plants, power system operators have faced new challenges such as overload, overvoltage, and proper performance during grid voltage disturbances. As a result, network standards and codes have been updated in the direction of photovoltaic systems that are more controllable and compatible with the network. To provide various grid support functions, such as frequency response and voltage support, the Constant Power Generation (CPG) control algorithm is established by grid codes. For this purpose, MPPT algorithms have been replaced by flexible power point tracking (FPPT) algorithms. This paper proposes an algorithm to control photovoltaic systems based on a multi-level cascaded full-bridge (CHB) inverter to achieve constant power generation control by the FPPT method. The required power reference is distributed among the sub-modules of the CHB converter according to the available power of each sub-module in such a way that, as much as possible, each sub-module is loaded equally. Then, the FPPT algorithm with an adaptive voltage step is used to set the power of each sub-module in its reference value. The voltage step is calculated adaptively based on the observed condition (transient or steady state) so that the tracking performance has fast dynamics. The effectiveness of the proposed algorithm is evaluated by simulating a 500-kW photovoltaic system directly connected to a 4.7 kV grid.

In this paper, a decentralized energy management system is presented for intelligent microgrids with the presence of distributed resources using reinforcement learning. Due to the unpredictable nature of renewable energy resources, the variability of load consumption, and the nonlinear model of batteries, the design of a microgrid energy management system is associated with many challenges. In addition, centralized control structures in large-scale systems increase computational volume and complexity in control algorithms. In this paper, a fully decentralized multi-agent structure for a microgrid energy management system is proposed and the Markov decision process is used to model the stochastic behavior of agents in the microgrid. Electrical and thermal distributed resources, batteries, and consumers are considered intelligent and independent agents. They have the learning ability to explore and exploit the environment in a fully decentralized manner and achieve their optimal policies. The proposed method for hourly microgrid management is model-independent and based on learning. The method maximizes the profits of all manufacturers, minimizes consumer costs, and reduces the dependence of the microgrid on the maingrid. Finally, using real data from renewable energy sources and consumers, the accuracy of the proposed method in the Iranian electricity market is simulated and verified.

Magnetic Induction Tomography (MIT) is a promising modality for noninvasive imaging due to its contactless and nonionizing technology. In this imaging method, a primary magnetic field is applied by excitation coils to induce eddy currents in the material to be studied, and a secondary magnetic field is detected from these eddy currents using sensing coils. The image (spatial distribution of electrical conductivity) is then reconstructed using measurement data, the initial estimation of electrical conductivity, and the iterative solution of forward and inverse problems. The inverse problem can be solved using one-step linear, iterative nonlinear, and special methods. In general, the MIT inverse problem can be solved by Gauss-Newton iterative method with acceptable accuracy. In this paper, this algorithm is extended and the zoning technique is employed for the reduction of unknown coefficients. The simulation results obtained by the proposed method are compared with the real conductivity coefficients and the mean relative error rate is reduced to 24.22%. On the other hand, Gauss-Newton iterative method is extended for solving the inverse problem of the MIT, and sensitivity measurement matrices are extracted in different experimental and normalization conditions.

The presence of electric vehicle technology in distribution networks as controllable resources provides advantages such as voltage regulation, power peaking, and loss reduction. This technology has a positive effect on the performance of the distribution network, but the simultaneous presence of electric vehicles and distributed generation sources requires optimal planning because the lack of access to an application reduces the life of these technologies and can cause blackouts in the power network. Therefore, in this study, the dynamic reconfiguration of the distribution network in the simultaneous presence of distributed generation sources and electric vehicles is proposed. Also, the time-of-use mechanism has been proposed as one of the demand response applications to improve the power consumption of subscribers. The objective functions in this study include the reduction of energy loss, operational cost and energy not supplied as the objective function of reliability. In general, the optimization problem of distribution network reconfiguration is complex and non-convex. Also, considering the effect of distributed generation units and electric vehicles makes the problem more complicated than before. Therefore, finding a practical method to solve the optimization problem is one of the main challenges of this paper. For this purpose, a novel hybrid algorithm based on a combination of an improved particle swarm optimization-artificial bee colony is presented to overcome the complexities of the optimization problem. The proposed method has

In this paper, a three-level framework is proposed to determine the optimal scheduling of a Multi-Area Active Distribution Network in the presence of inter-area Soft Open Points (SOPs). In this framework, the uncertainty of renewable generation and forecasted demand is modeled using information gap decision theory in a risk-averse manner. Coordinated scheduling of Controllable Distributed Generators (CDGs) and SOPs, inter-area energy exchanges and energy trading with upstream network considering the uncertainties are the contribution of the presented method. To improve the computational efficiency and to achieve the optimal solution, the scheduling problem is modeled as a second-order conic programming in which the operational and security constraints of the network, CDG limitations, and operational constraints of SOPs are accurately modeled and the problem is solved by executing CPLEX solver in MATLAB environment. A case study on IEEE 33-bus test system showcases the superiority of the proposed model compared to meta-heuristic algorithms such as particle swarm optimization, genetic algorithm, and gravitational search algorithm.

Low-frequency noise modeling and estimation is a critical issue in the design of analog and digital circuits, especially for short-channel CMOS technologies. The conventional noise models do not fit the experimental results adequately at frequencies close to DC for amplifiers and close to the carrier frequency for oscillators. Furthermore, measuring this noise necessitates the use of specialized and expensive equipment. In this paper, the bound of carrier trapping and releasing are extracted while examining the characteristics of low-frequency noise using the EMD method, and an innovative model based on carrier number fluctuations is proposed. The method introduces a low-cost measurement system for low-frequency noise. The performance accuracy of this measurement system and the proposed model is evaluated and compared with conventional noise modeling methods. The evaluation results demonstrate the success of the proposed measurement setup and model in estimating near-DC noise.

The commutation torque ripple adversely affects the performance of the six-phase inverter of the BLDC motor with trapezoidal back EMF and creates vibration and noise for industrial applications. In this paper, the motor model is obtained in non-commutation times and during the commutation period, and according to that, a suitable method to reduce the torque ripple, by equalizing the slope of the current disconnected from the motor and the slope of the current connected to the motor during commutation, is presented. At low speeds, torque ripple is reduced using predictive pulse width modulation technique. With this method, the duty cycle of the switch involved in the commutation is predicted and applied to the switch during the commutation intervals. At high speeds, this reduction is done using quasi z-source converter and selector circuit. The quasi z-source converter and the selector circuit increase the input voltage of the inverter during commutation intervals and increase its value to four times the back EMF voltage of the motor, thus reducing the torque ripple at high speeds. The theoretical and analytical results are verified using the simulations performed in the PLECS software.

Recently, studies of squirrel-cage induction machines validate the idea that these machines follow the double-cage induction model; however, the manufacturers categorize these machines as single-cage machines. There are many references that use the squirrel-model for squirrel-cage machines, then machine parameters increase from 5 to 7. In the laboratory, parameter determination is performed to estimate 5 parameters; however, 7 parameters have to be determined. Also, the DC test is not an exact test and is used for approximate estimation. Therefore, locked rotor test equations are changed and an overload test is added. After measurements in the laboratory according to the equations of this paper, a MATLAB program is prepared and 7 parameters are determined in a loop until reaching the desired error. This paper demonstrates that laboratory squirrel-cage machines have to be corrected using the 7-parameter model. Furthermore, this paper shows the necessity of new laboratory test equations for squirrel-cage machines. Then a numerical method is used to determine the machine parameters.