Electrical Engineering最新文献

Inertia prediction for power systems with a high proportion of renewable energy units can help coordinate inertia support methods, guide power system planning, and lower grid operational risk. Existing inertia prediction methods rarely use machine learning to predict the equivalent inertia of the power system, and there is also little consideration of the virtual inertia of the renewable energy units; some of the prediction methods rely on massive volumes of system data and suffer from issues such as data redundancy and complex pre-processing procedures. A method for predicting the equivalent inertia for power systems based on particle swarm optimization support vector machines (PSO-SVM) is proposed for this purpose. The method initially creates a database of system-equivalent inertia, which regards the power change and system frequency rate of change as feature inputs and the system-equivalent inertia as an output. Then, the optimal prediction model is matched using the feature difference matrix, and the PSO-SVM prediction method is utilized to predict the power system's equivalent inertia. The method proposed in this paper is validated by an improved three-machine nine-node power system, and the prediction accuracy is better than that of GA-BP neural network and SVM algorithms, and then the applicability in complex scenarios is validated by a ten-machine, thirty-nine-node power system as well as a site-specific power system under real-time wind speeds. The PSO-SVM prediction method reduces the maximum error by 23.64% compared to the GA-BP neural network and 68.27% compared to the SVM algorithm and the results show that the method proposed in this paper can more accurately predict inertial changes and inertial information of the system when a loading accident occurs.

This paper introduces a hybrid optimization approach, the Hybrid of African Vulture Optimizer with Genetic Operators (HAVOGO), designed to address the intricate challenges of optimal design in large distribution systems. The HAVOGO algorithm combines the robustness of the African vulture optimizer with the adaptability of genetic operators, resulting in superior optimization performance. The algorithm focuses on the simultaneous sizing and locating of distributed generation and distribution static compensator, alongside network reconfiguration, to efficiently incorporate electric vehicle charging stations into existing power distribution networks. A multi-objective optimization framework is utilized to allocate power compensating devices and optimize network reconfiguration, considering both technical and economic factors. The effectiveness of the HAVOGO algorithm is demonstrated through its application to 118-bus and 415-bus large distribution networks. Additionally, the results obtained from the HAVOGO algorithm are compared with those from other optimization algorithms and existing research in the field. Numerical results show significant improvements in performance metrics for both network sizes: for the 118-bus system, there is a reduction in active power loss by 84.72%, a decrease in voltage deviation by 76.22%, and an increase in voltage stability margin by 62.99%. Similarly, for the 415-bus system, the algorithm achieves a reduction in active power loss by 75.78%, a decrease in voltage deviation by 65.54%, and an increase in voltage stability margin by 26.06%.

String insulators play a critical role in electrical grids by isolating high voltage and preventing energy dispersion through the tower structure. Maintaining the cleanliness of these insulators is essential to ensure optimum performance and avoid malfunctions. Traditionally, human visual inspection has been used to assess cleaning needs, which can be error prone and pose a safety risk to personnel working near electrical equipment. Accurate detection of insulator condition is essential to prevent equipment failure. In this study, we used a comprehensive dataset of insulator images generated in Brazil using computer-aided design software and a game engine. The dataset consists of 14,424 images, categorized into those affected by salt, soot, and other contaminants, and clean insulators. We extracted key features from these images using VggNet and GoogleNet and classified them using a random forest algorithm, achieving a classification accuracy of 98.99%. This represents a 0.99% improvement over previous studies using the same dataset. Our research makes a significant contribution to the field by providing a more effective method for isolator management. By using advanced artificial intelligence models for accurate classification and real-time analysis, our approach improves the efficiency and reliability of insulator condition monitoring. This advance not only improves the detection of various insulator conditions but also reduces the reliance on manual inspections, which are often inaccurate and inefficient.

This paper presents a novel hybrid forecasting procedure for wind power using meteorological and historical data. The introduced method consists of three parts: effective feature selection, time series decomposition, and forecasting each decomposed time series. The minimum redundancy and maximum relevance (mRMR) algorithm is first utilized to choose the most effective features. In this stage, those selected historical features whose values are needed at the prediction time will be decomposed by the variational mode decomposition (VMD) technique and then forecasted by the long short-term memory (LSTM) networks. Then, the multivariate variational mode decomposition (MVMD) algorithm is exploited to simultaneously decompose selected features to address frequency mismatches between different series and capture the correlation among them. Given that various series and variables are involved in wind power forecasting, considering the correlation among them significantly affects prediction results. Afterward, LSTM neural networks are utilized to forecast each decomposed time series. Finally, two cases and several evaluation criteria are elaborated to assess the performance of the presented method. Experimental results confirm that the developed hybrid model, compared to the VMD-LSTM model, results in a decrease of 9.97, 4.33, and 3.32% in root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE), respectively. The mean values of these criteria are, respectively, 4.6, 3.5, and 20.8 for the proposed model.

Power quality (PQ) issues have intensified due to the rapid integration of renewable sources into the utility grid. An effective control strategy is imperative to address these problems. This paper proposes a novel approach by replacing conventional 2-level inverters with a simplified 5-level multilevel inverter (SMLI) as a shunt active power filter (SAPF) unit. The SMLI exhibits superior performance, including low total harmonic distortion in voltage, reduced electromagnetic interference, and enhanced system flexibility. Further, a multilayer Gamma filter is utilized to control SAPF and extract fundamental component from nonlinear load. The developed control not only improves the power factor at the supply end but also resolves other PQ issues. The novelty of the paper is the innovative utilization of the Gamma filter and closed loop implementation of SMLI system which can operate effectively in two modes, viz. during day and night. Furthermore, the system is also tested under partial shading conditions. The solar PV array integrated at the DC link of the SMLI supplies the load during the daytime. However, at night, when the solar PV array’s power is unavailable, the load’s demand is met solely by the grid. The SAPF functioning ensures an improvement in power quality. The proposed and developed Gamma filter control action performs quick adaptive estimation under diverse load profiles. Furthermore, reduced memory requirement during operation guarantees optimal performance under dynamic circumstances involving fluctuating solar irradiation and load parameters. The OPAL-RT real-time simulator has been used to both simulate and experimentally validate the results. The application of this research is extensively useful in grid-connected PV system for enhancing PQ issues in distribution systems with integrated renewable energy sources.

Smart grid intrusion is now increasing due to increased cyberattacks on intelligent devices. Cyberthreats like false data injection attack (FDIA) can bypass conventional security mechanisms. To defend against smart grid intrusion, in this research work, recurrent neural network with a Kalman filter is proposed to detect smart grid fault, normal, and FDIA events for a multi-sourced smart grid system. By using the stacking method, a novel parallel reinforcement learning with adaptive feature boosting is utilized to extract deterministic features. In the proposed feature extraction process, firstly Kalman filters are used to reduce feature dimension. Secondly, the resilient defence was constructed to improve the stable operation of the smart grid. The performance of the proposed Kalman filter reinforced neural network (KFRNN) is demonstrated by the presence of deterministic critical features under FDIA and without FDIA on a smart grid multi-sources data. The proposed KFRNN is evaluated by standard WUSTIL-2021 and real-time hardware-in-loop (HIL) test bed case study with FDIA. The obtained result shows that the proposed KFRNN provides resilient operation for smart grid by achieving a high classification accuracy of 97.3%, increased F1-score, increased receiver operating characteristic, and high detection probability than conventional schemes. Finally, a comprehensive simulation is performed on the IEEE 118 bus New England System to validate the effectiveness of the proposed KFRNN. From the obtained performance indexes, it is observed that the proposed intrusion detection scheme has high accuracy with enhanced resilient operation.

In the digital era, power systems are continuously implementing positive modifications on both the source and load sides. Further, power electronics interfaces are used to integrate dispersed generators, unconventional/nonlinear loads, charging stations, and so on. Consequently, frequent power quality disturbances appear in the system that are to be mitigated at the earliest to sustain the performance. Hence, this research proposes a novel intelligent power quality detection technique to identify and categorize PQ events, as mitigation requires detection. The proposed hybrid beetle formica optimized light GBM (HBFO-light GBM) offers a versatile solution by maintaining voltage control in power systems during critical operational scenarios to maintain power quality. The research at its core seeks to develop an advanced solar PV system model with a smart STATCOM, focusing on the effective preservation of energy within battery storage systems. The integration of ant colony and beetle swarm algorithms serves as a novel hybrid beetle formica optimization (HBFO) for system optimization, specifically focusing on stabilizing the output power of the shunt voltage converter within the PV system. This optimization enhances the classifier’s ability to effectively stabilize the output power, addressing potential fluctuations and biases in the system. The recorded values for various parameters in the system are as follows: the attained PV voltage, Q grid, Q inv, Q load, Vpcc, Pgrid, Pinv, Pload, PV current, and PV power are 561.49 V, 418.59 VAR, 418.59 VAR, 418.59 VAR, 176.34 V, 82.7042 W, 166.95 W, 82.70 W, 404.48 A and 193.012 KW, respectively.

The power generated from a solar panel installation needs to be controlled and increased using a large voltage gain DC–DC converter. This study delves into an innovative high gain, non-isolated DC–DC converter, referred to as the active switched inductor impedance source converter (ASIZSC). The converter includes several essential features that enhance its functionality such as improved gain, constant input current, low duty ratio, and reduced voltage stress on circuit elements. Three switches are present in the proposed converter. The duty ratio and switching frequency used to operate all three switches in the converter are similar. Simulation in MATLAB is used to confirm the functioning of the suggested converter. The simulation is carried out for a source voltage, (V_i) of 10 V, a load power of 100 W, duty ratio, (delta ) of 0.4, and switching frequency, (f_s) of 50 kHz. Hardware results as well as simulation data are given to support the effectiveness of the recommended converter. The load voltage is 120 V for a 10 V source voltage. The gain of the recommended converter is 12. To improve the dynamics of the converter, a closed loop system is developed. The designed closed loop system is simulated to verify its functionality. The viability of the MPPT operation of the ASIZSC in the photovoltaic application is confirmed through the MATLAB simulation.

The increasing load demand has caused issues in distribution systems, such as higher line losses, lower power factors, and voltage fluctuations. Addressing these challenges is crucial for power utilities to ensure the reliability and efficiency of the system. This study explores the efficient allocation of multi-type distributed generations (DGs) in radial distribution systems using an optimization approach to reduce power losses, improve voltage profiles, and maximize the total annual savings of the system. The paper introduces a novel utilization of the Puma Optimizer (PO) technique to address the optimal DG placement problem, incorporating different load models such as constant power, constant current, constant impedance, and composite load models to create a comprehensive framework for DG planning. The efficacy of the adopted PO to allocate different types of DG units is evaluated on 85-bus, 141-bus, and 415-bus systems. Additionally, the results obtained from the PO algorithm are compared with other well-known optimization algorithms and existing research in the field. Simulation results indicate that combining DG units operating at a zero-power factor with those operating at an optimal power factor significantly enhances system performance compared to DG units operating solely at a zero-power factor, a unity power factor, or a unity power factor combined with a zero-power factor. Numerical results demonstrate significant performance improvements across all network sizes. Specifically, active power losses are reduced by 96.99%, 92.33%, and 79.48%, while reactive power losses are reduced by 97.80%, 91.89%, and 78.40% for the 85-bus, 141-bus, and 415-bus systems, respectively. Additionally, the findings indicate that the PO algorithm is more robust than other selected algorithms in determining the optimal size and placement of DG units.

This study presents a novel approach to detecting and estimating partial discharge (PD) signals using a pattern recognition-based Mycielski algorithm. An experimental setup is first built in the high-voltage laboratory of Afyon Kocatepe University to test the proposed approach’s performance on PD detection and estimation in medium voltage XLPE cables. PD signals used in this study are measured from this experimental setup. In addition, a low-cost phase-resolved partial discharge analysis is realized, and PD measurement results are strengthened with a portable device with HFCT. Three different PD types are classified using the Mycielski assumption during the detection process, achieving an accuracy of 94.44%. The Mycielski algorithm is adopted to predict the PD signal’s future data in the estimation part, with the failure localization achieving an accuracy of 87.78%. The proposed method is feasible and may be applied in this field since it gives successful results for detecting and estimating PD signals. On the other hand, the accuracy of detection and estimation is open for development.