Recent Advances in Electrical & Electronic Engineering - Online First

Early detection of Alzheimer's disease (AD) is crucial due to its rising prevalence and the economic burdens it imposes on individuals and society. This study aimed to propose a technique for the early detection of AD using MRI scans.

The methodology involved collecting data, preparing the data, creating both single and combined models, assessing with ADNI data, and confirming with additional datasets. The approach was chosen by comparing various scenarios. The top six individual ConvNet-based classifiers were combined to form the ensemble model. The evaluation showed high accuracy rates across various classification groups. Validation of additional data showed impressive accuracy, exceeding results from numerous previous studies and aligning with others.

Although ensemble methods outperformed individual models, there were no notable distinctions among different ensemble approaches. The ensemble model was constructed using the top six individual ConvNet-based classifiers in deep learning (DL), achieving high accuracy rates across various classification categories: 98.66% for Normal control | AD, 96.56% for Normal control | Early MCI, 94.41%

Early MCI/Late MCI, 99.96% for Late MCI | AD, 94.19% for four-way classification, and 94.93% for three-way classification. Validation results underscored the limited effectiveness of individual models in practical settings, contrasting with the promising outcomes of the ensemble method.

Human health is significantly compromised by air pollution, especially by local air quality. The majority of our society spends their lives in a confined geographical location, which if subjected to air pollution can expose them to long-term air contamination. It is also possible that poor air quality can pose serious health risks, especially to susceptible individuals thereby impacting their lifestyle. Air quality can be improved with appropriate plantation, but they are underutilized. Various air purification devices have been developed in response to the ever-increasing air pollution level.

However, artificial means of air purification are not very viable in terms of cost, accessibility to society, and reliable tools to purify air. This research integrates traditional solutions with modern technology to counter air purification by selectively using plant species and placing them in desired locations suitable for urban settings. The study aims to measure the constituents of various air pollutants spanning across regions to identify and accumulate pollution data using IoT-based smart devices, remit, and feed this information to cloud-based storage for further processing. In addition, advanced predictive intelligence is utilized to determine the plant species that can suffice the need for air purification through organic means in a given geographical zone resulting in enhancement of Air Quality (AQ), with minimal cost, prolonged shelf life, future proof and non-detrimental consequences.

Implementation outcome gives a promising outcome. Accurate readings of various air pollutants are aggregated. Suitable trees are identified to tackle these pollutants and their absorbing capacity is determined. Various predictive methods are employed and the random forest model recorded the best results. The sensory units of the model successfully captured the pollutant data and any major fluctuations were reported. The prediction pipeline recorded a mean precision, recall, and f-score value of about 0.95, 0.92, and 0.94 respectively while the mean accuracy of 0.965 was also noted. The observed training and validation accuracy with our model were 0.96 and 0.93 respectively.

Hence, the proposed ‘AeroGlan’ model may be locally applied as an air pollutants monitoring device and also to suggest suitable plant species required to counter air contamination in that locality.

The rapid growth of distribution grids and the increase in load demand have made distribution grids play a crucial role in urban development. However, distribution networks are prone to failures due to multiple events. These faults not only incur high maintenance costs, but also result in reduced productivity as well as huge economic losses. Therefore, accurate and fast fault localization methods are very important for the safe and stable operation of distribution systems.

Firstly, the Ghost-Asf-YOLOv8 network is employed to assess the three-phase fault voltage travelling waveforms at both ends of the line, determine the temporal range of the fault occurrence, and differentiate its line mode components. Subsequently, the ICCEMDAN algorithm is employed to decompose the line mode components, thereby yielding the IMF1 components. The key feature information is then enhanced through the application of NTEO. Finally, the Ghost-Asf-YOLOv8 network is employed to further narrow down the time range of the initial traveling wave head, thereby enabling the calculation of the fault location and the determination of the traveling wave arrival time.

Experiments are conducted based on the simulation data of the constructed hybrid line model, and the comparison experiments between the TEO algorithm and the NTEO algorithm are conducted, which show that the NTEO has good noise immunity when applied to fault localization. In addition, the proposed ICCEMDAN-NTEO method is also compared with the fault localization methods based on DWT and HHT, and the results show that the method has high accuracy. Finally, the light weighted YOLOv8 model captures the traveling wave time quickly and accurately to compensate for the shortcomings of the visualization data.

This work presents a novel fault localization method that integrates traditional and artificial intelligence techniques, offering rapid detection and minimal localization error.

Integration of cloud components in a wireless mesh network of Wireless-Optical Broadband Access Network (WOBAN) contributes to enhancing network performance.

This study aims to deploy the minimum cloudlets at the optimum location and to offload excess traffic from overloaded cloudlets to underloaded ones.

The objective of this study is to optimize cloudlet positioning and traffic offloading for cost-effective deployment, better resource utilization, reduced blocking probability, and lower delays.

The proposed methodology introduces a Cluster-Based Heuristic Approach (CBHA) for efficient cloudlet placement along with a traffic offloading mechanism using a Customized Donkey-Smuggler Optimization (CDSO) to enhance the overall network performance.

Simulations show the effectiveness of the proposed approach for resource utilization, blocking probability, delay, and cost.

The problem in position optimization of the cloudlet, along with traffic offloading, is solved using the proposed approaches to get better network performance in a cost-efficient manner.

Arc fault has become an increasingly prominent problem affecting the safe operation of power distribution networks. Research on arc fault detection can effectively reduce electrical fire accidents caused by arc faults, which is of great significance for ensuring the safe and reliable operation of power distribution networks.

In this paper, an arc fault detection method based on current convolution is proposed for single-phase AC series arc faults. Firstly, the phase of the measured phase current is acquired through the phase-locked loop. Then, the measured phase current is convoluted with the standard sinusoidal signal whose phase is the same as the measured phase current, and the DC component is obtained by low-pass filtering.

The occurrence of an arc fault is recognized by detecting the change in the DC component.

Finally, the simulation results verify that the proposed method can detect the arc fault quickly and accurately.

With the continuous development of permanent magnet synchronous motors (PMSM) and the increasing demand for the application of flywheel battery, the requirements for PMSMs are also increasing.

A multi-objective genetic algorithm is used to solve the optimal design solution.

Multi-objective genetic algorithm is fast and accurate in calculation results, and it is easy to obtain the optimal solution. The results show that the cogging torque is reduced by 23.6%, the torque ripple is reduced by 25%, and the average torque is increased by 1.2%.

A multi-objective optimization design was conducted on a surface-mounted PMSM. Firstly, the sensitivity of different optimization variables was calculated. The high-sensitivity parameters were selected as the final optimization variables. The response surface between the optimization variables and the optimization objectives was calculated. The genetic algorithm was used to solve the optimal design solution. The effectiveness of the optimization results was verified by the combination of finite element simulation and experimental tests.

In grid-connected operation control, the Brushless Doubly-Fed Reluctance Generator (BDFRG) faces issues with strong parameter coupling and weak disturbance rejection.

This paper proposes a direct power control strategy with a novel integral sliding mode controller. By analyzing the correlation between the voltages on the stator control winding side and the active/reactive power, a direct power control model is derived from the d-q rotating coordinate system, achieving decoupling control of active and reactive power. An integral sliding surface, along with a smoothing function, is introduced to improve the switching behavior as the system approaches the sliding surface. Stability ranges for the parameters Kd and Kq are determined by constructing a Lyapunov function.

Results from simulations and hardware-in-the-loop (HIL) experiments demonstrate that the direct power control strategy with a novel integral sliding mode controller reduces chattering and improves the static and dynamic performance of the system, compared to conventional sliding mode control strategy.

The proposed direct power control strategy not only addresses the chattering issues during sliding mode switching but also ensures system stability and efficiency through optimized parameter adjustment.

Diabetes mellitus, stemming from insulin deficiency or resistance, poses acute and chronic health issues driven by factors like age, obesity, genetics, and lifestyle. It significantly impacts health, leading to conditions like heart disease, vision problems, and kidney dysfunction, with a notable mortality rate reported by the WHO in 2019. The modern diet has escalated diabetes risk. Machine learning techniques play a pivotal role in disease prediction, aiding timely interventions.

The primary aim of this research work is to explore and contrast the effectiveness of various existing machine-learning models for diabetes disease classification. The goal is to identify the optimal solution that yields the highest accuracy.

In the initial phase, we implemented data pre-processing, followed by the application of a diverse range of machine learning methods to classify diabetes mellitus. Subsequently, a comprehensive analysis was conducted on machine learning algorithms, considering both the complete dataset features and those selected through Particle Swarm Optimization (PSO). The assessment covered various metrics such as accuracy score, precision, F1 score, and log loss for Support Vector Classifier (SVC), K-Nearest Neighbours (KNN), Random Forest (RF), ADA Boost, XG Boost, Extra Tree, and Decision Tree. Ultimately, the introduction of hyperparameter tuning was aimed at enhancing performance and attaining the highest level of accuracy.

The proposed model HSVC combines the Particle Swarm Optimization (PSO) feature selection strategy with optimized hyperparameters, showcasing outstanding performance and achieving an accuracy of 98.66%.

The models developed in this study can potentially be applied or recommended for the classification of other health conditions in different domains, such as Parkinson’s disease, heart disease, and many more.