Power systems
Hamid Reza Sezavar; Hamid Karimi; Navid Fahimi
Abstract
A comparative approach is pretend in this paper that evaluates different Artificial Intelligence (AI) methods for diagnosing power transformer faults using Dissolved Gas Analysis (DGA). Traditional approaches like the Rogers Ratio Method and Duval Triangle have been used for many years, but offer unreliable ...
Read More
A comparative approach is pretend in this paper that evaluates different Artificial Intelligence (AI) methods for diagnosing power transformer faults using Dissolved Gas Analysis (DGA). Traditional approaches like the Rogers Ratio Method and Duval Triangle have been used for many years, but offer unreliable results for complex cases. Although, newer AI methods present better results, but still vary in how well they work. In this paper, several AI approaches are evaluated including Support Vector Machines (SVMs), Random Forest (RF), Gradient Boosting Machines (GBMs), Deep Neural Networks (DNNs) and a new combinational model is proposed based on comparing results. A real DGA dataset is used for covering six different fault types for the proposed testing. The results show that while all AI methods do better than traditional approaches, the combinational approach performs the best with 92.3% accuracy. This is found 20.2% better than traditional methods and 4.8% better than the best single AI model. Rational explanation is provided for how each method works and practical recommendation is presented for choosing the right approach based on particular requirements and available resources in real practices.
Optimization
Hamid Karimi; Hamid Reza Sezavar
Abstract
This paper presents a multi-objective energy management strategy for a hybrid electricity-hydrogen microgrid (MG) aimed at improving economic and operational performance. The proposed framework simultaneously minimizes total operating cost and reduces peak load, which are conflicting objectives in MG ...
Read More
This paper presents a multi-objective energy management strategy for a hybrid electricity-hydrogen microgrid (MG) aimed at improving economic and operational performance. The proposed framework simultaneously minimizes total operating cost and reduces peak load, which are conflicting objectives in MG operation. To address this challenge, the problem is formulated as a multi-objective optimization model using a normalized weighted sum approach, enabling an effective trade-off between objectives while preventing the creation of new peaks in the load profile. The studied hydrogen-based MGs includes renewable energy sources such as solar and wind power, a battery energy storage system (BESS), a diesel generator, and demand response programs. Hydrogen is also utilized as an energy carrier to enhance system flexibility and support improved integration of renewable generation. Coordinated scheduling of generation units, storage systems, and flexible loads improves system efficiency. To enhance model accuracy, a long short-term memory (LSTM) forecasting method is applied to predict solar irradiance and wind speed using historical data, supporting more reliable scheduling decisions. The proposed model is validated through a general case study. Simulation results show a peak load reduction of 442 kW and an improvement in the load factor of 16.75%, confirming the effectiveness of the proposed approach
Power systems
Hamid Reza Sezavar
Abstract
This paper presents a comparative study on the application of artificial intelligence for optimizing External Lightning Protection Systems (ELPS) in photovoltaic power (PV) plants. The research addresses the critical need for advanced protection systems in solar installations, which are particularly ...
Read More
This paper presents a comparative study on the application of artificial intelligence for optimizing External Lightning Protection Systems (ELPS) in photovoltaic power (PV) plants. The research addresses the critical need for advanced protection systems in solar installations, which are particularly vulnerable to lightning strikes due to their expansive outdoor configurations. Through a detailed comparative analysis, the study evaluates multiple AI approaches, including metaheuristic algorithms and machine learning models. The investigation reveals that metaheuristic algorithms often have lower accuracy compared to modern AI techniques. All comparisons are based on a multi-level optimization framework, systematically addressing air termination design, grounding system configuration, and overall system integration. The results show superiority in sensitivity analysis in the transformer model. Compared to other models, the random forest (RF) model, along with the artificial neural network (ANN) model, has a higher speed in data analysis. However, physics-informed neural networks (PINN) achieve remarkable improvements, delivering 93% protection coverage with only 3.2% grounding error while significantly reducing design convergence times.
Power systems
Hamid Reza Sezavar; Saeed Hasanzadeh
Abstract
Insulator pollution levels are critical for ensuring the operational stability and safety of power transmission systems. Traditional methods for detecting pollution are often invasive, inaccurate, and time-consuming. To address these issues, this study investigates the application of Artificial Intelligence ...
Read More
Insulator pollution levels are critical for ensuring the operational stability and safety of power transmission systems. Traditional methods for detecting pollution are often invasive, inaccurate, and time-consuming. To address these issues, this study investigates the application of Artificial Intelligence (AI), specifically Gradient Boosting Machines (GBM), to classify insulator pollution levels based on Partial Discharge (PD) characteristics. We utilize a combination of time-domain and frequency-domain features extracted from PD signals to train a predictive model. The results indicate that the proposed model achieves a high classification accuracy, averaging between 92% and 95% across various contamination levels. Furthermore, the study analyzes the model's sensitivity to environmental factors, including humidity and Hydrophobicity Class (HC), revealing important insights that could influence classification performance. By employing this AI-driven approach, we aim to significantly enhance the efficiency of power grid maintenance, ultimately contributing to the long-term stability and reliability of transmission systems. The findings from this research underscore the potential of AI in revolutionizing pollution assessment methods and optimizing maintenance practices in power infrastructure.
