Adv. eng. intell. Syst. - Academia.edu (original) (raw)

Papers by Adv. eng. intell. Syst.

Advances in Engineering and Intelligence Systems, 2024

Compressive strength (CS) and slump flow (SL) are two of the most essential parameters in High-Pe... more Compressive strength (CS) and slump flow (SL) are two of the most essential parameters in High-Performance Concrete (HPC), which directly affect its structural capacity, durability, and workability. The following study represents an important step toward developing novel hybrid models for predicting CS and SL. The contribution in this paper proposes the following: the radial basis function (RBF) model will be enhanced by using two optimization algorithms, namely Horse Herd Optimization (HHO) and Wild Geese Algorithm (WGA). Accordingly, two hybrid models have been proposed, referred to as RBFH and RBWG. For model testing, comprehensive metrics include the coefficient of determination (R²), RMSE, MAE, VAF, and SI. It has the highest R² values of 98.17 for CS and 93.54 for SL predictions among all the datasets, while it also records the lowest SI values of 0.064 and 0.037 for CS and SL, respectively. These are indicative of the accuracy and reliability of the RBWG model in modelling the properties of HPC. This work's significance consists of improving the concrete mix design by giving correct predictions of HPC performance, which will lead to optimized resource utilization, minimized costs, and reduced negative environmental impacts due to construction. The results highlight hybrid machine learning models as the potential to solve complex challenges in civil engineering and provide new approaches toward sustainable and efficient infrastructure development.

Advances in Engineering and Intelligence Systems, 2024

Fly ash-containing concrete has only been the subject of a small amount of research focused on fo... more Fly ash-containing concrete has only been the subject of a small amount of research focused on forecasting the hardened concrete qualities. So little research has been done to predict the characteristics of self-compacting concrete in both its fresh and hardened states (SCC). Using support vector regression (SVR), it is planned to construct networks for estimating SCC's before and after hardening attributes. The goal of this research is to identify the SVR technique's critical parameters utilizing Henry gas solubility optimization (HGSO) and particle swarm optimization (PSO). SCC's fresh-phase characteristics include the slump flow, V-funnel test, and L-box test, whereas its hardened-phase features involve the strength of the compressive. The outcomes show tremendous promise for all assessed qualities in the assessment and development sections. In terms of development and assessment, it was clear that the presented networks have an excellent R^2 value. In other words, it signifies that the correlation between the real and anticipated characteristics of SCC from hybrid systems is satisfactory, which reflects superlative precision in the process of approximation and development. Overall, the HGSO-SVR model beats PSO-SVR, showing the capacity of the algorithm to choose the most effective parameters for the method under examination.

Advances in Engineering and Intelligence Systems, 2024

The widespread use of electronic devices in various applications has led to increased harmonic di... more The widespread use of electronic devices in various applications has led to increased harmonic distortion in power lines, resulting in a decreased power factor and reduced overall power quality. Voltage rectifiers connected to the line can be significant sources of this distortion. To mitigate these effects, it is crucial to enhance the power factor and minimize line current harmonics. This study employs a cascade-controlled boost converter for power factor correction, leveraging both voltage and current regulation. By integrating a self-tuning regulator (STR) adaptive controller with online identification within each control loop, this approach dynamically adjusts to fluctuations in current consumption over time. The proposed method effectively reduces Total Harmonic Distortion (THD) and improves power factor, contributing to more efficient and reliable power systems.

Advances in Engineering and Intelligence Systems, 2024

This paper introduces a novel methodology for predicting Unconfined Compressive Strength (UCS) in... more This paper introduces a novel methodology for predicting Unconfined Compressive Strength (UCS) in rocks by integrating Support Vector Regression (SVR) with two cutting-edge optimization algorithms: the Seahorse Optimizer (SO) and the COOT Optimization Algorithm (COOT). Unlike traditional UCS prediction methods that often struggle with slow convergence and local minima entrapment, this approach leverages the nonlinear modeling capabilities of SVR and enhances its performance through advanced optimizers. The SVSH (SVR+SO) and SVCO (SVR+COOT) models were developed and evaluated using a comprehensive dataset of rock samples with UCS measurements. Comparative analysis demonstrates that the proposed models not only achieve significantly higher prediction accuracy but also exhibit faster convergence compared to standalone SVR. These results underscore the potential of the hybrid SVR-optimizer models to set a new benchmark in UCS prediction, offering greater precision and computational efficiency. Among them, the SVSH models had the maximum accuracy with an excellent R2 value of 0.998 and a scanty RMSE value of 1.261. Therefore, the results confirm that the proposed SVSO model is a promising tool to be used by engineering and geological professionals. This ensures a very strong and reliable UCS prediction method that is vital for the improvement of civil engineering project safety and efficiency.

Advances in Engineering and Intelligence Systems, 2024

Precise anticipation of the ultimate bearing capacity (Qu)for rock-socketed piles is an important... more Precise anticipation of the ultimate bearing capacity (Qu)for rock-socketed piles is an important task in civil engineering, construction, and the design of foundations. The approach adopted here is new and solves the problem using KNN combined with two modern nature-inspired optimization frameworks, namely the Honey Badger Algorithm (HBA) and Equilibrium Slime Mould Algorithm (ESMA). The hybrid model in this paper combines the K-nearest neighbor with HBA and ESMA. The main objective was to improve the prediction performance of Qu for rock-socketed piles. This hybridization technique utilized the KNN model's strengths and two new optimizers to overcome the inherent uncertainty related to all variables that affect bearing capacity. These HBA and ESMA frameworks were proven capable of performing decent tuning of the KNN model, while their outcomes showed significantly improved predictive power. The hybrid model realized high accuracy for Qu estimation by considering various influencing parameters like soil properties, pile dimensions, and load conditions. The output of this study adds to the development in the area of geotechnical engineering by providing a sound methodology for Qu estimation in rock-socketed piles. The hybridization technique is a promising avenue for future exploration and practical applications, especially the KNHB frameworks that have obtained reliable outcomes by their very accurate R2 of 0.9945 and RMSE of 771.0886Outcomes support engineers and designers in formulating educated judgments concerning the foundation in soft soil environments. Ultimately, this exploration promotes safer and more efficient construction practices by minimizing risks associated with foundation failures.

Advances in Engineering and Intelligence Systems, 2024

The undrained shear strength (USS) of soil is essential in diverse structural engineering applica... more The undrained shear strength (USS) of soil is essential in diverse structural engineering applications, including the design of earth dams, rock fills, foundations, highways, railways, and slope stability analysis. Traditional empirical and theoretical methods for estimating USS based on field tests often rely on correlation assumptions, leading to imprecise results. These conventional strategies are also limited in terms of efficiency, both in time and cost. To address these limitations, this study introduces innovative machine learning techniques, employing the Support Vector Regression (SVR) model to accurately predict USS. To enhance model performance, three meta-heuristic optimization algorithms Differential Squirrel Search Algorithm (DSSA), Golden Section Search Optimization (GSSO), and Northern Goshawk Optimization (NGO) were utilized. The frameworks were trained using four key input metrics: plastic limit (PL), liquid limit (LL), sleeve friction (SF), and overburden weight (OBW). The performance of the proposed frameworks was evaluated using five criteria: R², RMSE, MSE, SI, and SMAPE. Among the three hybrid frameworks developed for estimating USS, the SVR optimized with the Northern Goshawk Optimization (SVNG) algorithm outperformed the others, achieving the lowest RMSE value of 39.30 in the testing step and the highest R² value of 0.9980 in the testing step. These results demonstrate the superiority of the SVNG model in providing precise and efficient predictions of USS, surpassing conventional methods.

Advances in Engineering and Intelligence Systems, 2024

In recent years, there has been a noticeable surge in population, accompanied by the rapid develo... more In recent years, there has been a noticeable surge in population, accompanied by the rapid development of industries, services, and agriculture within communities. This growth has intensified pressure on water resources, resulting in a simultaneous decline in both the quantity and quality of these vital resources. Dissolved organic carbon (DOC) stands out as a pivotal factor in determining the quality of natural freshwater bodies. It serves as a crucial indicator of water pollution, offering insights into the presence of both natural and man-made organic pollutants. This study focuses on predicting the levels of dissolved organic carbon in South Florida utilizing artificial intelligence algorithms. Specifically, Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Multilayer Perceptron (MLP), Radial Basis Function (RBF), Decision Tree (DT), and Support Vector Regression (SVR) algorithms were employed. The Sparrow Search Algorithm was utilized to fine-tune the hyperparameters of these algorithms. The effectiveness of the hybrid models generated was evaluated through correlation and error parameters, along with the number of iterations required to achieve the lowest error rate. Upon comprehensive review, the SVR-SSA hybrid model emerged as the most promising, exhibiting an R2 value of 0.997293, a Root Mean Square Error (RMSE) of 0.2906, and a Normalized Mean Square Error (NMSE) of 5×10-4 at the thousandth iteration in the train dataset. These values for the test data set are calculated as 0.999754, 0.1625, and 2×10-4, respectively. Hence, it is deemed the most optimum model for this research endeavour.

Advances in Engineering and Intelligence Systems, 2024

Voltage source inverters (VSIs) are the most common type of inverter used in power electronic sys... more Voltage source inverters (VSIs) are the most common type of inverter used in power electronic
systems for distributed generation sources. In three-phase systems, using three-phase conversion
to the synchronous reference frame, linear controllers can be easily used to control the inverter due
to the conversion of AC parameters in the synchronous frame to dc components. The design of
linear controllers in single-phase systems is often difficult because of the oscillating nature of these
systems. In some cases, single-phase conversion to synchronous frame is used, which has its
problems, such as delays in converting single-phase values to two-phase. To tackle this problem,
proportional-resonance (PR) instruments can be used at a specific resonant frequency. However,
for single-phase systems where all the harmonies are present in the inverter current, an unlimited
number of PR instruments are required. Therefore, not all harmonies can be controlled in this way.
To overcome this, a iterative control system can be used. The control method presented in this paper
uses delay intervals with positive feedback. Under these conditions, the control gain in the dynamic
model of the iterative control system is very high for all harmonies and allows tracking the reference
current and voltage values with a zero steady-state error in all harmonies. In addition, the
performance of the iterative control system in the face of single-phase systems will be very
favorable.

Advances in Engineering and Intelligence Systems, 2024

Dynamic management of virtual machines (VMs) in cloud data centers is essential for achieving fle... more Dynamic management of virtual machines (VMs) in cloud data centers is essential for achieving flexibility, efficiency, and high productivity in cloud systems. With the rapid growth of cloud technologies and increasing demand for cloud-based services, effective resource management has become a critical factor for ensuring quality service delivery while minimizing operational costs. This is particularly important for organizations seeking to optimize their resource utilization and adapt dynamically to fluctuating workloads. This paper introduces a novel model designed to address these challenges by enhancing existing algorithms and employing advanced techniques. The approach integrates genetic algorithms and refrigeration simulation into the migration replacement process, leveraging absorbing Markov chains for predictive analysis. By continuously monitoring resource status, analyzing incoming data, and forecasting critical server conditions, the model effectively reduces unnecessary VM migrations. Simulations conducted using the Clodsim environment demonstrate the model’s efficiency in reducing energy consumption across low, medium, and high-load scenarios. The proposed method achieves an average reduction in energy consumption of 17% compared to state-of-the-art methods, while also minimizing violations of service-level agreements (SLA). This research highlights the importance of combining predictive analytics with robust optimization techniques to improve cloud resource management. By achieving significant energy savings and enhancing system reliability, the proposed model offers a practical, sustainable framework for dynamic VM management, addressing the challenges posed by growing user demands and resource constraints in modern cloud data centers.

Advances in Engineering and Intelligence Systems, 2024

Designing or appraising framed concrete buildings exposed to high eccentric loads requires precis... more Designing or appraising framed concrete buildings exposed to high eccentric loads requires precise reinforced concrete (RC) torsional strength estimates. Unfortunately, semi-empirical formulae still fail to adequately predict the torsional capacity of RC beams, particularly over-reinforced and strong ones. To overcome this limitation, accurate Machine Learning (ML) models might replace more sophisticated and computationally intensive models. This work evaluates and determines the most effective tree-based machine learning algorithms to estimate the torsional capacity (T_r) of RC beams subjected to pure torsion. The objective of the present work is to provide innovative hybrid models that combine the concepts of the Adaptive neuro fuzzy inference systems (ANFIS) model with other optimization approaches, such as the Giant trevally algorithm (GTA) and Honey badger algorithm (HBA), to accurately forecast the T_r. A total of 202 RC rays were collected to form the data set. To facilitate the application of the ANFIS models, a training set and a testing set were created from the database. Out of the 202 samples in the database, 25 percent (51) were used for evaluation and 75 percent (151) were used for learning. ANF-GTA demonstrated superior performance compared to ANF-HBA, with a 50% higher performance in the learning phase and an 80% higher performance in the evaluation stage, as measured by the MedSE index values. The ANF-GTA obtained lower SMAPE index values of 5.2192 and 5.39 compared to the values of 8.0622 and 9.3783 obtained by the ANF-HBA during the learning and assessment phases, respectively.

Advances in Engineering and Intelligence Systems, 2024

The rapid evolution of healthcare services has been driven by the convergence of the Internet of ... more The rapid evolution of healthcare services has been driven by the convergence of the Internet of Things and cloud computing technologies, inherently transforming telemedicine. This study evaluates the transformative effect of IoT and cloud-based solutions through the telemedicine domain, especially in the remote delivery of healthcare services and medical consultations. It explores how IoT-enabled medical devices facilitate continuous remote patient monitoring and real-time health data collection and emphasizes the role of wearable devices comprising smartwatches and fitness trackers in elevating data accuracy and offering personalized health insights. Cloud computing's key role in telemedicine is investigated through its scalable data storage and processing capabilities, which manage the vast amounts of data generated by IoT devices. The study also discusses the benefits of cloud-based telemedicine platforms, including scalability, accessibility, cost-effectiveness, and the implementation of security measures to address data privacy concerns. The combination of IoT and cloud methods indicates various advantages, such as improving patient outcomes, expanding healthcare accessibility, and enabling data-driven decision-making for medical professionals. Real-world case studies outline successful implementations of these methods, which result in decreased hospital readmissions, improved patient engagement in health management, and centralized critical care expertise. However, significant challenges are addressed, including data security, interoperability, and reliable internet connectivity in various healthcare settings. The reports highlight the necessity of more studies on the integration of artificial intelligence in telemedicine and the adoption of edge computing solutions to overcome cloud-related issues in low-connectivity scenarios. This study provides valuable insights into redefining telemedicine in the digital era and serves as a guide for future developments in healthcare technology.

Advances in Engineering and Intelligence Systems, 2024

Solar energy is a widely embraced renewable resource, characterized by its unpredictable, fluctua... more Solar energy is a widely embraced renewable resource, characterized by its unpredictable, fluctuating, and stochastic nature. To mitigate risks and optimize asset utilization cost-effectively, precise analysis and forecasting of solar radiation can prove beneficial. This work aims to provide a hybrid model using machine learning to accurately predict solar Direct normal irradiance with the least amount of error. In this work, long short-term memory has been optimized using Particle swarm optimization, Grasshopper optimization algorithm, and Slime mold algorithm. SMA-LSTM, which has the best performance result compared to other developed models, is presented as the main method for this work. The data used is from June 1, 2022, to July 30, 2023. Many factors, such as the coefficient of determination, root mean square error, mean absolute percentage error, and mean absolute error have been used in presenting this work, and SMA-LSTM results with the lowest amount of R^2 has illustrated acceptable performance.

Advances in Engineering and Intelligence Systems, 2024

Surge arresters are essential protective devices in power systems that redirect excess voltage an... more Surge arresters are essential protective devices in power systems that redirect excess voltage and prevent damage to sensitive equipment. Extensive research has been conducted in electro-thermal modeling for surge arresters to better understand and mitigate faults and accidents. These efforts have largely focused on FEM and heat transfer modeling techniques to analyze temperature distribution, electric field effects, thermal-mechanical stress, burning point analysis, puncture risks, and thermal runaway behavior. This study also presents frequency and time series analysis of surge arresters during short circuits. By integrating these two analyses, engineers can enhance their understanding of surge arrester behavior during short circuits, refine design strategies to address issues like resonance and harmonics, and ensure reliable system protection against overvoltage occurrences.

Advances in Engineering and Intelligence Systems, 2024

Glucose oxidase is a crucial enzyme used in industries such as chemicals, pharmaceuticals, food, ... more Glucose oxidase is a crucial enzyme used in industries such as chemicals, pharmaceuticals, food, and biotechnology due to its ability to oxidize glucose into hydrogen peroxide and gluconic acid (C6H12O7). However, its limited and costly production necessitates strategies for improved efficiency. Immobilization, which involves attaching enzymes to surfaces, is a key method that enhances glucose oxidase's activity, stability, and reusability, making it more practical for repeated industrial use. This study comprehensively reviews glucose oxidase immobilization techniques, emphasizing their evolution, advanced methods, and diverse applications, particularly in biomedical fields. By immobilizing glucose oxidase, its stability, activity, and reusability are significantly enhanced, addressing the challenges of enzyme production and operational efficiency. The review discusses various immobilization methods, including physical adsorption, covalent binding, entrapment, and encapsulation, comparing their advantages and limitations. Incorporating nanomaterials and computational methods, such as molecular modeling and simulations, further optimizes the immobilization process, providing insights into enzyme-support interactions and improving performance in applications like biosensors, biofuel cells, and drug delivery systems. The study also explores future trends, including artificial intelligence and innovative support materials, to drive advancements in glucose oxidase immobilization. By integrating these techniques, the study aims to pave the way for more effective and versatile biomedical solutions, contributing to the ongoing evolution of enzyme immobilization technology.

Advances in Engineering and Intelligence Systems, 2024

Traditional fault-location methods applied to a double-circuit transmission line are usually impl... more Traditional fault-location methods applied to a double-circuit transmission line are usually implemented according to an abc-domain or a sequence network equivalent circuit that also varies with fault type. However, this dependence of those fault location algorithms from the line parameters suffers from inaccuracy injected under varying environmental conditions altering the parameters of the double-circuit transmission line. Herein, an Extreme Learning Machine-based line parameter-independent fault location method is suggested that can learn the nonlinear relationship between the voltages, currents measured, and fault locations with high accuracy. In the presented paper, the proposed method is simulated for different fault types at random distances in a power grid containing a double-circuit transmission line. The simulated data are then utilized for training the intelligent fault location system. Further, different distances and resistances fault locations are estimated to check the accuracy of the proposed approach. The obtained results are compared with the results of two other intelligent fault detection approaches, such as ANN and SVM and better accuracy and reliability are shown in ELM. The outputs of these tests show considerable improvements in the proposed technique of fault location on a double circuit transmission line under different environmental conditions.

Advances in Engineering and Intelligence Systems, 2024

The limited hydration capacity and challenges related to delayed expansion prevent fly ash (FA) a... more The limited hydration capacity and challenges related to delayed expansion prevent fly ash (FA) and MgO expansive additive (MEA) from being used significantly. Nonetheless, utilizing these two procedures in hydraulic mass concrete applications is a frequently used approach that yields favorable outcomes. To construct and assess machine learning-based algorithms to assess the volume expansion (V_e) of cement paste, which consists of FA and MEA, 170 experimental findings from published studies are employed. A novel approach called least square support vector regression (LSSVR) has been developed. The efficacy of LSSVR is significantly impacted by its hyperparameters (c and g), which were fine-tuned using the Dwarf Mongoose Optimization Algorithm (DMOA) and the Equilibrium Optimization Algorithm (EOA). Based on the results obtained, it can be inferred that there exists a significant potential for both 〖LSSVR〗_E and 〖LSSVR〗_D models to accurately predict the V_e of cement paste that incorporates fly ash and MgO expansive addition. In the training and testing phases, the Theil inequality coefficient (TIC) values for 〖LSSVR〗_E are observed to be 0.0906 and 0.01043, which are comparatively higher than the TIC values for 〖LSSVR〗_D, which are 0.0382 and 0.0044, respectively. By predicting the volume expansion accurately, engineers can adjust the proportions of FA and MEA to achieve desired expansion properties, improving the durability and stability of concrete structures. Accurate prediction models allow for better control of thermal stresses, reducing the risk of thermal cracking and extending the structure's lifespan.

Advances in Engineering and Intelligence Systems, 2024

Wireless Sensor Networks (WSNs) are vital for applications such as environmental monitoring, surv... more Wireless Sensor Networks (WSNs) are vital for applications such as environmental monitoring, surveillance, and healthcare, where comprehensive network coverage is essential for accurate data collection. However, achieving full coverage in WSNs presents significant challenges due to resource constraints, such as limited battery life, processing capabilities, and environmental factors like terrain and obstacles. To address these issues, coverage optimization techniques are employed to maximize spatial coverage while minimizing energy consumption and deployment costs. This paper provides a thorough overview of these coverage optimization techniques, categorizing them based on different deployment strategies, including static and dynamic sensor placement. It explores their respective advantages, limitations, and application scenarios, offering valuable insights for researchers and practitioners. The study is motivated by the need to better understand how to improve WSN coverage efficiency and ensure reliable data collection in diverse environments. The research aims to synthesize existing knowledge on WSN coverage optimization, identify gaps in current strategies, and guide future studies in this field. Key findings emphasize the effectiveness of various techniques in enhancing coverage, such as mobility-based approaches and energy-aware algorithms, while also addressing practical challenges like sensor redundancy and environmental unpredictability. Ultimately, this paper contributes to the ongoing efforts to develop more adaptive, scalable, and energy-efficient solutions for WSN coverage optimization.

Advances in Engineering and Intelligence Systems, 2024

Network alignment plays a pivotal role in fields such as network science, biology, and social net... more Network alignment plays a pivotal role in fields such as network science, biology, and social network analysis by identifying common structures and relationships across different networks. The process is challenging due to the diversity of network structures and the necessity to align networks from various domains or periods. To address these challenges, nature-inspired meta-heuristic optimization algorithms have emerged as powerful tools. These algorithms, inspired by natural processes such as evolution, swarm behavior, and other biological phenomena, provide effective solutions for complex optimization problems. This paper offers a thorough examination of the application of these meta-heuristic algorithms to network alignment. It explores a wide range of nature-inspired algorithms, including genetic algorithms, particle swarm optimization, ant colony optimization, and simulated annealing. The review delves into the underlying principles of each algorithm, their practical applications, and their performance in network alignment tasks. By analyzing detailed discussions and practical examples, the paper highlights the strengths and limitations of various meta-heuristic algorithms. It assesses their effectiveness in aligning networks across different scenarios, providing valuable insights for researchers and practitioners. The findings emphasize the potential of these algorithms in overcoming the complexities of network alignment, offering guidance for employing these techniques effectively. The paper also explores future research directions, suggesting ways to advance the field by leveraging nature-inspired algorithms. As a comprehensive resource, it consolidates existing knowledge and enhances understanding, supporting the development of innovative solutions and improved strategies for network alignment.

Advances in Engineering and Intelligence Systems, 2024

Self-Compacting Concrete (SCC) has been widely utilized in construction projects and academic res... more Self-Compacting Concrete (SCC) has been widely utilized in construction projects and academic research due to its environmentally friendly components, such as fly ash and superplasticizers, which reduce water requirements. SCC’s ability to self-deposit eliminates the need for vibration, resulting in cost and energy savings. However, some experts are hesitant about its broader application due to insufficient training in modern materials. Accurately assessing construction aggregates' compressive strength (CS) ensures structural safety. Soft computing methods, which offer a cost-effective and highly accurate alternative to experimental techniques, have attracted interest in modeling dependent variables. This paper presents a novel approach by combining a Support Vector Machine (SVM) with advanced optimization algorithms to estimate the CS of SCC mixtures accurately. The significance of this approach lies in the ability of the optimization algorithms to enhance the performance of the SVM, yielding more precise predictions and addressing the limitations of traditional methods. The developed models were evaluated using several performance metrics, with results showing a strong correlation between predicted and actual values, achieving an R² of 97.3%. Furthermore, the root mean square error (RMSE) was calculated at 3.81 MPa, demonstrating the effectiveness of the proposed method in predicting SCC’s compressive strength with high accuracy.

Advances in Engineering and Intelligence Systems, 2024

The growing global population has significantly increased energy demand, particularly in the resi... more The growing global population has significantly increased energy demand, particularly in the residential building sector. This surge underscores the necessity for accurate energy consumption forecasting to facilitate effective planning and future demand projections. However, traditional methods such as regression face challenges in modeling household electricity consumption due to seasonal and monthly variations. Smart grid technology now allows users to manage home energy use more efficiently and effectively. This study explores optimizing Artificial Neural Network (ANN) parameters using meta-heuristic algorithms instead of traditional gradient-based methods to predict residential electricity consumption across different seasons. The experimental data to train a Radial Basis Function (RBF) neural network were utilized. The meta-heuristic algorithms employed for fine-tuning the ANN's weight and bias parameters include the Genetic Algorithm (GA), Multi-Verse Optimizer (MVO), Moth Flame Optimization (MFO), Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Advanced Grey Wolf Optimizer (AGWO), Biogeography-Based Optimization (BBO), and Particle Swarm Optimization with Grey Wolf Optimizer (PSOGWO). These algorithms were evaluated for their effectiveness in adapting to seasonal variations in electricity consumption data. The results revealed that the PSOGWO algorithm consistently outperformed the others across all seasons, particularly in spring and summer. Statistical measures indicated superior accuracy and reliability in predicting energy usage. Specifically, the nine-neuron configuration with the PSOGWO algorithm achieved a high R² value of 0.99077, reflecting lower error metrics. Conclusively, the PSOGWO model's consistent performance across seasonal variations underscores its potential for reliable residential electricity consumption forecasting, making it a valuable tool for energy management in smart grids.

Advances in Engineering and Intelligence Systems, 2024