Applied Sciences

Research

Jump to: Review

28 pages, 7265 KiB

Open AccessArticle

Accurate Rotor Temperature Prediction of Permanent Magnet Synchronous Motor in Electric Vehicles Using a Hybrid RIME-XGBoost Model

by Jianzhao Shan, Zhongyuan Che and Fengbin Liu

Appl. Sci. 2025, 15(7), 3688; https://doi.org/10.3390/app15073688 - 27 Mar 2025

Viewed by 322

Abstract

With the growing global focus on environmental protection and carbon emissions, electric vehicles (EVs) are becoming increasingly popular. Permanent magnet synchronous motors (PMSMs) have emerged as a core component of the drive system due to their high-power density and compact design. The rotor [...] Read more.

With the growing global focus on environmental protection and carbon emissions, electric vehicles (EVs) are becoming increasingly popular. Permanent magnet synchronous motors (PMSMs) have emerged as a core component of the drive system due to their high-power density and compact design. The rotor temperature of PMSMs significantly affects their operating efficiency, management strategies, and lifespan. However, real-time monitoring and acquisition of rotor temperature are challenging due to cost and space limitations. Therefore, this study proposes a hybrid model named RIME-XGBoost, which integrates the RIME optimization algorithm with XGBoost, for the precise modeling and prediction of PMSM rotor temperature. RIME-XGBoost utilizes easily monitored dynamic parameters such as motor speed, torque, and currents and voltages in the d-q coordinate system as input features. It simultaneously optimizes three hyperparameters (number of trees, tree depth, and learning rate) to achieve high learning efficiency and good generalization performance. The experimental results show that, on both medium-scale datasets and small-sample datasets in high-temperature ranges, RIME-XGBoost outperforms existing methods such as SMA-RF, SO-BiGRU, and EO-SVR in terms of RMSE, MBE, R-squared, and Runtime. RIME-XGBoost effectively enhances the prediction accuracy and computational efficiency of rotor temperature. This study provides a new technical solution for temperature management in EVs and offers valuable insights for research in related fields. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

16 pages, 3813 KiB

Open AccessArticle

Power Prediction in Photovoltaic Systems with Neural Networks: A Multi-Parameter Approach

by Zeynep Bala Duranay and Hanifi Guldemir

Appl. Sci. 2025, 15(7), 3615; https://doi.org/10.3390/app15073615 - 26 Mar 2025

Viewed by 303

Abstract

In this study, a neural network-based power prediction for a photovoltaic system was conducted using a multi-parameter approach, considering radiation, temperature, wind speed, humidity, and cloud cover. Photovoltaic systems are highly popular renewable energy sources due to their robust, modular, and environmentally friendly [...] Read more.

In this study, a neural network-based power prediction for a photovoltaic system was conducted using a multi-parameter approach, considering radiation, temperature, wind speed, humidity, and cloud cover. Photovoltaic systems are highly popular renewable energy sources due to their robust, modular, and environmentally friendly characteristics. Although photovoltaic systems offer many advantages, their dependency on irradiation for energy generation and their sensitivity to meteorological parameters pose a significant disadvantage, leading to intermittent energy production. Since these parameters affect the quality of power generated at the plant, they introduce uncertainty in power systems. Therefore, it is crucial to consider these factors in energy planning and management. In this study, to mitigate uncertainty in power systems and contribute to energy planning by predicting power production, power data obtained from a power plant, along with meteorological data, were used in Single Layer Perceptron Neural Network. The predicted power values obtained from the proposed model were compared with the actual values, and the results of this comparison were presented. Furthermore, to demonstrate the model’s performance, the R and MSE values were provided as 0.98 and 0.03, respectively, indicating a strong correlation between predicted and actual values and a low prediction error. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

28 pages, 6396 KiB

Open AccessArticle

Three-Layer Framework Integrating Optimal Placement of Supervisory, Control, and Acquisition System Measurements with Clustering-Based Electric Substations Selection for State Estimation of Medium-Voltage Distribution Networks

by Vasilica Dandea, Stefania Galbau, Mihai-Alexandru Baciu and Gheorghe Grigoras

Appl. Sci. 2025, 15(4), 1942; https://doi.org/10.3390/app15041942 - 13 Feb 2025

Viewed by 451

Abstract

One of the biggest challenges, both from a technical and economic point of view, of the Distribution Network Operators refers to identifying the locations (electric distribution substations) integrated into a supervisory, control, and acquisition (SCADA) system to perform on-site measurements used in the [...] Read more.

One of the biggest challenges, both from a technical and economic point of view, of the Distribution Network Operators refers to identifying the locations (electric distribution substations) integrated into a supervisory, control, and acquisition (SCADA) system to perform on-site measurements used in the state estimation of the electric distribution networks (EDNs). In response to this challenge, a robust and resilient three-layer methodology has been proposed to solve the state estimate issue of the EDNs based on an optimal placement algorithm of the remote terminal units integrated into the SCADA system at the level of the EDSs. The first layer allows a clustering algorithm-based determination of the classes of the EDSs with similar features of the load profiles. The second layer identifies the “candidate” classes and decides the pilot EDSs with on-site SCADA measurements. The third layer allows the state estimation of the EDN based on the load values measured in the pilot EDEs. The framework was tested and validated using a medium voltage EDN of a Romanian DNO supplying an urban area. The results obtained highlighted that the accuracy had been ensured for on-site measurements in 12 of 39 EDSs (representing approximately 30% of EDSs integrated into the SCADA system), leading to a mean average percentage error of 2.6% for the load estimation and below 1% for the state variables determined by a power flow calculation at the level of the EDN. Consequently, the framework can significantly decrease the investments associated with integrating the SCADA system by the DNOs, with great benefits regarding the state estimation of the EDNs. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

10 pages, 2460 KiB

Open AccessArticle

Attention-Based Hydrogen Refueling Imputation Model for Efficient Hydrogen Refueling Stations

by Keunsoo Ko and Changgyun Kim

Appl. Sci. 2024, 14(22), 10332; https://doi.org/10.3390/app142210332 - 10 Nov 2024

Viewed by 828

Abstract

During hydrogen refueling, the data values determining the state of charge (SoC) of a vehicle can be missing due to internal and external factors. This causes inaccurate SoC estimation, resulting in oversupply or undersupply. To overcome this issue, an attention-based hydrogen refueling imputation [...] Read more.

During hydrogen refueling, the data values determining the state of charge (SoC) of a vehicle can be missing due to internal and external factors. This causes inaccurate SoC estimation, resulting in oversupply or undersupply. To overcome this issue, an attention-based hydrogen refueling imputation (AHRI) model, which restores missing values, is proposed in this paper. In particular, considering that data variables can vary depending on the environmental conditions and equipment in a hydrogen refueling station (HRS), we use the attention mechanism. It determines the primary features, which improves the predictive performance and helps adapt to new conditions. Using the observed data during hydrogen refueling, we train the proposed AHRI model and verify its efficacy. Experimental results show that the proposed AHRI model outperforms existing imputation models significantly. Here, AHRI achieves 0.95 and 0.82 in terms of

R^{2}

when 20% and 40% of the values are missing, respectively. These results indicate that the proposed model can be used to solve the data missing problems in HSRs. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

28 pages, 8909 KiB

Open AccessArticle

A Novel Data-Driven Approach with a Long Short-Term Memory Autoencoder Model with a Multihead Self-Attention Deep Learning Model for Wind Turbine Converter Fault Detection

by Joel Torres-Cabrera, Jorge Maldonado-Correa, Marcelo Valdiviezo-Condolo, Estefanía Artigao, Sergio Martín-Martínez and Emilio Gómez-Lázaro

Appl. Sci. 2024, 14(17), 7458; https://doi.org/10.3390/app14177458 - 23 Aug 2024

Cited by 2 | Viewed by 1317

Abstract

The imminent depletion of oil resources and increasing environmental pollution have driven the use of clean energy, particularly wind energy. However, wind turbines (WTs) face significant challenges, such as critical component failures, which can cause unexpected shutdowns and affect energy production. To address [...] Read more.

The imminent depletion of oil resources and increasing environmental pollution have driven the use of clean energy, particularly wind energy. However, wind turbines (WTs) face significant challenges, such as critical component failures, which can cause unexpected shutdowns and affect energy production. To address this challenge, we analyzed the Supervisory Control and Data Acquisition (SCADA) data to identify significant differences between the relationship of variables based on data reconstruction errors between actual and predicted values. This study proposes a hybrid short- and long-term memory autoencoder model with multihead self-attention (LSTM-MA-AE) for WT converter fault detection. The proposed model identifies anomalies in the data by comparing the reconstruction errors of the variables involved. However, more is needed. To address this model limitation, we developed a fault prediction system that employs an adaptive threshold with an Exponentially Weighted Moving Average (EWMA) and a fixed threshold. This system analyzes the anomalies of several variables and generates fault warnings in advance time. Thus, we propose an outlier detection method through data preprocessing and unsupervised learning, using SCADA data collected from a wind farm located in complex terrain, including real faults in the converter. The LSTM-MA-AE is shown to be able to predict the converter failure 3.3 months in advance, and with an F1 greater than 90% in the tests performed. The results provide evidence of the potential of the proposed model to improve converter fault diagnosis with SCADA data in complex environments, highlighting its ability to increase the reliability and efficiency of WTs. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

21 pages, 5233 KiB

Open AccessArticle

Three Novel Artificial Neural Network Architectures Based on Convolutional Neural Networks for the Spatio-Temporal Processing of Solar Forecasting Data

by Llinet Benavides Cesar, Miguel-Ángel Manso-Callejo and Calimanut-Ionut Cira

Appl. Sci. 2024, 14(13), 5955; https://doi.org/10.3390/app14135955 - 8 Jul 2024

Cited by 2 | Viewed by 1521

Abstract

In this work, three new convolutional neural network models—spatio-temporal convolutional neural network versions 1 and 2 (ST_CNN_v1 and ST_CNN_v2), and the spatio-temporal dilated convolutional neural network (ST_Dilated_CNN)—are proposed for solar forecasting and processing global horizontal irradiance (GHI) data enriched with meteorological and astronomical [...] Read more.

In this work, three new convolutional neural network models—spatio-temporal convolutional neural network versions 1 and 2 (ST_CNN_v1 and ST_CNN_v2), and the spatio-temporal dilated convolutional neural network (ST_Dilated_CNN)—are proposed for solar forecasting and processing global horizontal irradiance (GHI) data enriched with meteorological and astronomical variables. A comparative analysis of the proposed models with two traditional benchmark models shows that the proposed ST_Dilated_CNN model outperforms the rest in capturing long-range dependencies, achieving a mean absolute error of 31.12 W/m², a mean squared error of 54.07 W/m², and a forecast skill of 37.21%. The statistical analysis carried out on the test set suggested highly significant differences in performance (p-values lower than 0.001 for all metrics in all the considered scenarios), with the model with the lowest variability in performance being ST_CNN_v2. The statistical tests applied confirmed the robustness and reliability of the proposed models under different conditions. In addition, this work highlights the significant influence of astronomical variables on prediction performance. The study also highlights the intricate relationship between the proposed models and meteorological and astronomical input characteristics, providing important insights into the field of solar prediction and reaffirming the need for further research into variability factors that affect the performance of models. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

32 pages, 2217 KiB

Open AccessArticle

Statistical Comparison of Time Series Models for Forecasting Brazilian Monthly Energy Demand Using Economic, Industrial, and Climatic Exogenous Variables

by André Luiz Marques Serrano, Gabriel Arquelau Pimenta Rodrigues, Patricia Helena dos Santos Martins, Gabriela Mayumi Saiki, Geraldo Pereira Rocha Filho, Vinícius Pereira Gonçalves and Robson de Oliveira Albuquerque

Appl. Sci. 2024, 14(13), 5846; https://doi.org/10.3390/app14135846 - 4 Jul 2024

Cited by 7 | Viewed by 2199

Abstract

Energy demand forecasting is crucial for effective resource management within the energy sector and is aligned with the objectives of Sustainable Development Goal 7 (SDG7). This study undertakes a comparative analysis of different forecasting models to predict future energy demand trends in Brazil, [...] Read more.

Energy demand forecasting is crucial for effective resource management within the energy sector and is aligned with the objectives of Sustainable Development Goal 7 (SDG7). This study undertakes a comparative analysis of different forecasting models to predict future energy demand trends in Brazil, improve forecasting methodologies, and achieve sustainable development goals. The evaluation encompasses the following models: Seasonal Autoregressive Integrated Moving Average (SARIMA), Exogenous SARIMA (SARIMAX), Facebook Prophet (FB Prophet), Holt–Winters, Trigonometric Seasonality Box–Cox transformation, ARMA errors, Trend, and Seasonal components (TBATS), and draws attention to their respective strengths and limitations. Its findings reveal unique capabilities among the models, with SARIMA excelling in tracing seasonal patterns, FB Prophet demonstrating its potential applicability across various sectors, Holt–Winters adept at managing seasonal fluctuations, and TBATS offering flexibility albeit requiring significant data inputs. Additionally, the investigation explores the effect of external factors on energy consumption, by establishing connections through the Granger causality test and conducting correlation analyses. The accuracy of these models is assessed with and without exogenous variables, categorized as economical, industrial, and climatic. Ultimately, this investigation seeks to add to the body of knowledge on energy demand prediction, as well as to allow informed decision-making in sustainable energy planning and policymaking and, thus, make rapid progress toward SDG7 and its associated targets. This paper concludes that, although FB Prophet achieves the best accuracy, SARIMA is the most fit model, considering the residual autocorrelation, and it predicts that Brazil will demand approximately 70,000 GWh in 2033. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

Review

Jump to: Research

25 pages, 2212 KiB

Open AccessReview

A Review of Smart Photovoltaic Systems Which Are Using Remote-Control, AI, and Cybersecurity Approaches

by Andreea-Mihaela Călin (Comșiț), Daniel Tudor Cotfas and Petru Adrian Cotfas

Appl. Sci. 2024, 14(17), 7838; https://doi.org/10.3390/app14177838 - 4 Sep 2024

Cited by 4 | Viewed by 2412

Abstract

In recent years, interest in renewable energy and photovoltaic systems has increased significantly. The design and implementation of photovoltaic systems are various, and they are in continuous development due to the technologies used. Photovoltaic systems are becoming increasingly complex due to the constantly [...] Read more.

In recent years, interest in renewable energy and photovoltaic systems has increased significantly. The design and implementation of photovoltaic systems are various, and they are in continuous development due to the technologies used. Photovoltaic systems are becoming increasingly complex due to the constantly changing needs of people, who are using more and more intelligent functions such as remote control and monitoring, power/energy prediction, and detection of broken devices. Advanced remote supervision and control applications use artificial intelligence approaches and expose photovoltaic systems to cyber threats. This article presents a detailed examination of the applications of various remote-control, artificial intelligence, and cybersecurity techniques across a diverse range of solar energy sources. The discussion covers the latest technological innovations, research outcomes, and case studies in the photovoltaics field, as well as potential challenges and the possible solutions to these challenges. Full article

(This article belongs to the Special Issue Advanced Forecasting Techniques and Methods for Energy Systems)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Advanced Forecasting Techniques and Methods for Energy Systems

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (8 papers)

Research

Review

Further Information

Guidelines

MDPI Initiatives

Follow MDPI