Explore the vast range of titles available.

In this article, the authors propose two models for BLDC motor winding temperature estimation using machine learning methods. For the purposes of the research, measurements were made for over 160 h of motor operation, and then, they were preprocessed. The algorithms of linear regression, ElasticNet, stochastic gradient descent regressor, support vector machines, decision trees, and AdaBoost were used for predictive modeling. The ability of the models to generalize was achieved by hyperparameter tuning with the use of cross-validation. The conducted research led to promising results of the winding temperature estimation accuracy. In the case of sensorless temperature prediction (model 1), the mean absolute percentage error MAPE was below 4.5% and the coefficient of determination R2 was above 0.909. In addition, the extension of the model with the temperature measurement on the casing (model 2) allowed reducing the error value to about 1% and increasing R2 to 0.990. The results obtained for the first proposed model show that the overheating protection of the motor can be ensured without direct temperature measurement. In addition, the introduction of a simple casing temperature measurement system allows for an estimation with accuracy suitable for compensating the motor output torque changes related to temperature.

This article investigates the influence of faults in the phase current measurement channel on torque generation in permanent magnet synchronous motor (PMSM) drives. It is demonstrated that measurement errors, depending on their type and origin, significantly distort the actual electromagnetic torque as a function of the motor shaft angle. Since the current controller operates on incorrect values, indirect diagnostic methods are required. This study shows that analyzing the output signals of the current component controllers, particularly the torque regulator output and its phase relation to the electrical or mechanical angle, enables fault detection and classification. The waveform, frequency, and amplitude of these signals provide information about the fault’s nature, with offset errors having a more pronounced impact on torque across the entire load range than gain-related errors.

Keratoconus is a progressive eye disease characterized by the thinning and bulging of the cornea, leading to visual impairment. Early and accurate diagnosis is crucial for effective management and treatment. This study investigates the application of machine learning models to identify keratoconus based on corneal topography and biomechanical data. We collected a dataset comprising 144 corneal scans from adults aged 18–35, including an equal proportion of keratoconus and normal cases. Various machine learning algorithms were trained and evaluated on datasets containing different parameters obtained using the Pentacam device. The Random Forest algorithm demonstrated the highest reliability, achieving an accuracy of 98% during training and 96% on the test set, while also identifying the most diagnostically relevant measurements. Unlike prior studies, our approach enables detailed comparison between model-selected features and clinically recognized diagnostic parameters. This interpretability provides a clinically meaningful bridge between AI-driven predictions and expert-based decision-making. The results suggest that machine learning models, particularly Random Forest, can effectively aid in the early detection of keratoconus in young individuals, potentially improving patient outcomes through timely intervention.

This article addresses the issue of the elevator cabin door drive system failure diagnosis. The analyzed component is one of the most critical and the most vulnerable part of the entire elevator. Existing solutions in the literature include methods such as spectral analysis of system vibrations, motor current signature analysis, fishbone diagrams, fault trees, multi-agent systems, image recognition, and machine learning techniques. However, there is a noticeable gap in comprehensive studies that specifically address classification of the multiple types of system components failures, class imbalance in the dataset, and the need to reduce data transmitted over the elevator’s internal bus. The developed diagnostic system measures the drive system’s parameters, processes them to reduce data, and classifies 11 device failures. This was achieved by constructing a test bench with a prototype cabin door drive system, identifying the most frequent system faults, developing a data preprocessing method that aggregates every driving cycle to one sample, reducing the transmitted data by 300 times, and using machine learning for modeling. A comparative analysis of the fault detection performance of seven different machine learning algorithms was conducted. An optimal cross-validation method and hyperparameter optimization techniques were employed to fine-tune each model, achieving a recall of over 97% and an F1 score approximately 97%. Finally, the developed data preparation method was implemented in the cabin door drive controller.

This article presents an estimation method of the BLDC rotor position with asymmetrically arranged Hall sensors. Position estimation is necessary to control the motor by methods other than block commutation. A sinusoidal control method was selected for the research, which significantly reduces torque ripples and acoustic noise and is quite simple to implement. Inaccurate performance of the elements determining the position of the BLDC motor rotor causes a large error in the position estimation and has a negative impact on the operation of the drive controlled in this way. Using the developed control algorithms, it is possible to correctly determine the mechanical position of the rotor even for multi-pole motors. The proposed method is relatively easy to implement and does not require modification of control systems, being limited to changes only in the software of such devices. The tests of the actual system clearly show the usefulness of such a control method and its effectiveness.

Electromechanical energy harvesters are used to extract energy from vibrations occurring in nature, transport, or industry. The main problem with such solutions is that their output voltage is completely dependent on the frequency and amplitude of the vibrations, which can make it difficult to power a specific device or charge a battery. Therefore, it is necessary to use solutions that meet these requirements. Most harvesters contain additional, specialized mechanical gearboxes, called mechanical rectifiers or power electronic interfaces, used to match the harvester’s output voltage to the load. Design work was carried out, the construction of the proposed energy harvester was described, and the operation principle of the author’s control algorithm was presented. The results of the research confirm the possibilities of influencing the output voltage and power of the harvester system independently of the frequency and excitation amplitude.

Low-power BLDC motors are often and willingly used in many drive devices due to their functional advantages. They are also used in advanced positioning systems, where their good dynamic performance parameters are used. The control systems use shaft position sensors mounted on motors, the structure of which is based on magnetic elements and Hall sensors. The aim of this article was to investigate the influence of the BLDC motor quality on the correct operation of the control semiconductor system. The article presents the effect of BLDC motor shaft observation system’s inaccuracies on the friction and current amplitudes of individual inverter keys. Waveforms of the controller phase currents are considered and recorded on a test bench that allows precise sensor position changes. In addition, the effect of sensor misalignment on power losses in individual inverter transistors is investigated. The article shows a significant influence of the motor shaft observation system’s assembly accuracy on the current amplitudes of individual driver transistors and their power losses, which makes it necessary to consider these parameters when constructing power electronic systems.

Keratoconus is a progressive corneal disease that requires early and accurate detection to prevent severe visual impairment. This study presents a deep learning-based classification model for distinguishing between healthy and keratoconic eyes using dynamic corneal imaging data from the CORVIS system. A hybrid CNN-RNN architecture was developed, combining a fine-tuned InceptionV3 network for spatial feature extraction with a recurrent LSTM module to capture temporal patterns across image sequences. To ensure robust evaluation, a 10-fold stratified cross-validation strategy was employed, with data splits performed at the patient level to avoid data leakage. The model achieved an average accuracy, precision, recall, and F1-score of approximately 0.90 across folds, demonstrating strong generalization performance. Boxplot visualizations of metric distributions further confirmed model stability and revealed minimal performance variance. Class-wise analysis showed high effectiveness in detecting both healthy and keratoconic cases, although slightly greater variability was observed in the classification of healthy eyes. These results indicate that the proposed method is a promising tool for keratoconus screening and may complement existing diagnostic workflows. Further validation on external datasets is recommended prior to clinical deployment.

Cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) is associated with improved survival in recurrent ovarian cancer (ROC) but carries a high risk of postoperative complications. Accurate perioperative risk stratification remains an unmet need. To develop and internally validate a perioperative risk model for postoperative complications in ROC patients using information available by the end of surgery (pre- and intra-operative data), and to compare logistic regression (LR) and artificial neural networks (ANN) as possible predictive models. A retrospective analysis of 71 patients treated with CRS and HIPEC between 2011 and 2022 was performed. Clinical, surgical, and perioperative variables were analysed. Predictors were restricted to variables available at or before the end of the index operation, which prevents information leakage by using only data available at prediction time. LR and ANN models were developed and assessed with cross-validation. Performance reporting followed TRIPOD (Type b) and TRIPOD + AI, with Brier score, and calibration slope/intercept from out‑of‑fold (OOF) predictions. Thresholded metrics (accuracy, precision, recall, F1 score, and the area under the receiver operating characteristic curve, AUC) were summarised at a prespecified probability cut-off. Exploratory univariate odds ratio analyses with Holm-adjusted p-values were used to explore procedure–complication associations. Postoperative complications occurred in 45% of patients. LR identified blood loss ( p  = 0.005) and number of procedures ( p  = 0.042) as significant predictors of complications. The LR model achieved an accuracy of 66.2%, precision of 64.3%, recall of 56.2%, F1 score of 60.0%, and AUC of 0.700. The ANN model achieved an accuracy of 97.2%, precision of 94.3%, recall of 100%, F1 score of 97.1%, and AUC of 0.967. Hysterectomy with adnexa (OR = 11.67, p  = 0.035) and metastasectomy (OR = 7.42, p  = 0.042) were significantly associated with higher postoperative complication rates. ANN demonstrated superior predictive performance compared to LR in identifying postoperative complications after CRS and HIPEC, as indicated by ROC analysis. Combining traditional statistical modelling with modern machine learning may enhance ROC for perioperative risk stratification after CRS with HIPEC. A well-calibrated, interpretable LR model together with a highly discriminative ANN could enable more tailored allocation of intensive care resources and earlier identification of high-risk patients, potentially improving the safety of this demanding but beneficial treatment. However, external multicentre validation is required before clinical implementation.

Background: Recurrent ovarian cancer (ROC) significantly challenges gynecological oncology due to its poor outcomes. This study assesses the impact of cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) on ROC survival rates. Materials and Methods: Conducted at the Medical University of Lublin from April 2011 to November 2022, this retrospective observational study involved 71 patients with histologically confirmed ROC who underwent CRS and subsequent HIPEC. Results: The median overall survival (OS) was 41.1 months, with 3-year and 5-year survival rates post-treatment of 0.50 and 0.33, respectively. Patients undergoing radical surgery for primary ovarian cancer had a median OS of 61.9 months. The key survival-related factors included the Peritoneal Carcinomatosis Index (PCI) score, AGO score, platinum sensitivity, and ECOG status. Conclusions: The key factors enhancing ROC patients’ survival include radical surgery, optimal performance status, platinum sensitivity, a positive AGO score, and a lower PCI. This study highlights the predictive value of the platinum resistance and AGO score in patient outcomes, underlining their role in treatment planning. Further prospective research is needed to confirm these results and improve patient selection for this treatment approach.