فصلنامه هوش محاسباتی در مهندسی برق
سال دهم شماره 3 (پاییز 1398)

In this paper, a new single-stage power factor correction (PFC) converter is presented. In the input stage of this structure, a buck converter is used to correct the power factor. So, the voltage stress of bulk capacitor reduced, and one of the main problems in the single stage PFC converters is solved. Moreover, by employing a compensator circuit, the input current dead angle which is the main drawback of the existing buck type PFCs is obviated, thus the total harmonic distortion is very low. In converters with buck PFC, input current dead angle results in high THD. In the proposed converter, using an inductor coupled with the main transformer, this problem is solved. A part of input power is transmitted directly to the output that helps to improve efficiency. The proposed converter is analyzed and the validity of the proposed solution is proven in the form of theoretical analysis, simulation and practical results.

Electroencephalogram signals (EEGs) show the electrical activity of brain neurons. EEG is a non-invasive method that can be used to detect abnormal brain activities. Seizure is one of these abnormal activities and is the most common manifestation of epilepsy. Spikes are the most important characteristic of the seizure prone EEG signals. By detecting spikes, it is possible to detect epileptic seizures from EEG signals. EEG signals are non-stationary signals, so the wavelet transform that has appropriate time and frequency resolution can be a good option for extracting features of EEG signals. In this paper, after the extraction process using wavelet transform, artificial neural networks (ANNs) are used to classify healthy and epileptic signals. Particle swarm optimization (PSO) is also used as a novel approach to select weights and biases of network to improve network performance. The results of the implementation of the proposed algorithm have a 96.2% accuracy, which shows acceptable performance compared to existing methods.

Microgrid can operate in both grid-connected and islanded modes. Islanding can be scheduled or unscheduled. In the event of unscheduled islanding, the microgrid should be able to maintain its sustainability. Therefore, it is necessary to determine and supply the amount of required spinning reserve for stable islanding. In this paper, a two-stage strategy for optimal scheduling of the microgrid is proposed while uncertainties of renewable generations and demand are considered. In the first stage, co-optimization of power and reserve is carried out with the presence of distributed generations and batteries. Therefore, in the second stage, the demand is also used as reserve to provide required reserve for stable islanding. In order to determine the optimal amount of reserve of demand, a novel objective function is proposed. The objective function minimizes the sum of the standby reserve cost, expected cost of called reserve, and expected cost of emergency curtailment. One of the salient features of the proposed model is its modularity. That is, it can be added easily to any scheduling model in which the demand is not considered as reserve. Mathematical analysis and numerical simulations show the accuracy and efficiency of the proposed method.

Nowadays, determining the contribution of the individual consumers to the harmonic contamination is required for improving power quality. Due to changes occurred in lines, transformers, loads and generators over time, network operating conditions and impedance of different sectors are frequently varied resulting in the change in the contribution of harmonic sources. In this paper, an intelligent-based three-step method is developed to continuously determine the contribution of each harmonic source to the harmonic voltage of different buses without any access to the voltage phase data at different buses in the network. In this algorithm, firstly, the K-means clustering method is used for pre-processing of the measured data in order to reduce the background harmonic destructive effect on accuracy of the harmonic contribution determination. After calculating the harmonic contribution, the K-nearest neighbor method is used to generalize the results, and subsequently to create a continuous harmonic contribution matrix (HCM). Finally, the method is applied to a standard power network-based calculation example. The results demonstrate the capability of the proposed algorithm to evaluate the effects of harmonic sources in power networks.

In this paper, the fuzzy neural controller has been used to control the speed of the surface permanent magnet synchronous motor, despite its uncertainty in parameters and torque load. This method first uses a variety of local controllers (such as PIDs, LQRs, etc.) for different engine operating points and for different uncertainties and torque for precise engine control. Then the adaptive fuzzy controller learns that all of the local controllers are included and therefore, despite the indeterminacy in the parameters and torque of the motor, the reference speed with fast response and the least stable mode error are followed. Fuzzy neural network training algorithm is a mixed method, which is a combination of two methods of least squares and descending gradients with error propagation method. The least squares method is used to adjust the linear parameters of the output layer and the descending gradient algorithm uses an error propagation method for adjusting and updating the nonlinear parameters of the fuzzy layer. In the end, simulation of this controller is compared with H∞, Fuzzy and PID controller. Simulation results show the effectiveness of the proposed method in the paper.

The future power system, with the unprecedented penetration of renewable energy resources, will be faced with many uncertainties which may lead to problems in grid operation. Consequently, the uncertainty assessment of system performance is essential. Hence, the conventional power flow methods, given the constraints of the electricity and gas networks, may not be suitable for distribution networks such as the multi-carrier microgrids. This paper presents an effective method for optimizing the simultaneous utilization of different energy infrastructures in an environment with different uncertainties considering the constraints of the network. The aim is to study the effect of uncertainties on the optimal operation management of an interconnected microgrids. The fluctuation behavior of loads, renewable resources, and electricity price has been investigated in the proposed model based on probabilistic power flow technique. The results of the system are extracted as random variables, which are depicted in probabilistic and cumulative distribution forms. In this study, the multi-carrier microgrids not only exchange energy with the main grid, but also, energy interchange among the microgrids is possible. Simulations are presented, applying the procedure to an illustrative system of three interconnected MCMGs, and the results justify the effectiveness of the proposed technique.