This paper presents a novel method for QRS detection in electrocardiograms (ECG). It is based on the S-Transform, a new time frequency representation (TFR). The S-Transform provides frequency-dependent resolution while maintaining a direct relationship with the Fourier spectrum. We exploit the advantages of the S-Transform to isolate the QRS complexes in the time-frequency domain. Shannon energy of each obtained local spectrum is then computed in order to localize the R waves in the time domain. Significant performance enhancement is confirmed when the proposed approach is tested with the MIT-BIH arrhythmia database (MITDB). The obtained results show a sensitivity of 99.84%, a positive predictivity of 99.91% and an error rate of 0.25%. Furthermore, to be more convincing, the authors illustrated the detection parameters in the case of certain ECG segments with complicated patterns.

Nowadays, machine learning projects have become more and more relevant to various real-world use cases. The success of complex Neural Network models depends upon many factors, as the requirement for structured and machine learning-centric project development management arises. Due to the multitude of tools available for different operational phases, responsibilities and requirements become more and more unclear. In this work, Machine Learning Operations (MLOps) technologies and tools for every part of the overall project pipeline, as well as involved roles, are examined and clearly defined. With the focus on the inter-connectivity of specific tools and comparison by well-selected requirements of MLOps, model performance, input data, and system quality metrics are briefly discussed. By identifying aspects of machine learning, which can be reused from project to project, open-source tools which help in specific parts of the pipeline, and possible combinations, an overview of support in MLOps is given. Deep learning has revolutionized the field of Image processing, and building an automated machine learning workflow for object detection is of great interest for many organizations. For this, a simple MLOps workflow for object detection with images is portrayed.

This paper proposes a new AC voltage sensorless control scheme for three-phase pulse-width modulation rectifier. A new startup process to ensure a smooth starting of the system is also proposed. The sensorless control scheme uses an adaptive neural (AN) estimator inserted in voltage-oriented control to eliminate the grid voltage sensors. The developed AN estimator combines an adaptive neural network in series with an adaptive neural filter. The AN estimator structure leads to simple, accurate and fast grid voltages estimation, and makes it ideal for low cost digital signal processor implementation. Lyapunov based stability and parameters tuning of the AN estimator are performed. Simulation and experimental tests are carried out to verify the feasibility and effectiveness of the AN estimator. Obtained results show that; the proposed AN estimator presented faster convergence and better accuracy than the second order generalized integrator based estimator; the new startup procedure avoided the over-current and reduced the settling time; the AN estimator presented high performances even under distorted and unbalanced grid voltages.

The importance of looking into microgrid security is getting more crucial due to the cyber vulnerabilities introduced by digitalization and the increasing dependency on information and communication technology (ICT) systems. Especially with a current academic unanimity on the incremental significance of the microgrid’s role in building the future smart grid, this article addresses the existing approaches attending to cyber-physical security in power systems from a microgrid-oriented perspective. First, we start with a brief descriptive review of the most commonly used terms in the latest relevant literature, followed by a comprehensive presentation of the recent efforts explored in a manner that helps the reader to choose the appropriate future research direction among several fields.

For lithium-ion batteries, active balancing can bring advantages compared to passive balancing in terms of lifetime and available capacity. Most known balancing techniques suffer from a low efficiency or high complexity and cost. This paper proposes a new technical solution based on a non-isolated DC/DC converter and a low-speed switching matrix to overcome efficiency and power limitations of present balancing methods. The proposed circuit allows balancing current paths which are not possible with previous methods and results in balancing efficiencies of over 90%. The performance comparison is based on batch numerical simulations using calculated and measured efficiency values. The simulation results are compared to the approved active balancing method cell-to-cell and to passive balancing. A study case with eight randomly distributed battery cells shows an improvement in overall balancing efficiency of up to 47.4% and a reduction in balancing time of up to 36.9%. The available capacity increases from 97.2% in a passively balanced system to 99.7% for the new method. A hardware prototype was set up to demonstrate the working principle and to verify the numerical simulations.

In this paper, a new neural network-based virtual flux (NN-VF) estimator is proposed for sensorless control of a pulse-width modulation rectifier under unbalanced and distorted grid conditions. In this estimator, the VF vector is reconstructed through an emulated ideal integrator. Thereafter, positive and negative sequence VF components are extracted using a NNbased positive and negative sequence components separator. Lyapunov’s theory-based convergence analysis is performed for optimal tuning of the NN-VF estimator. Accordingly, accurate and fast estimation is achieved. A VF-based predictive direct power control (VF-PDPC) model including the estimated VF positive sequence components is formulated. A startup procedure is considered for smooth starting under unbalanced grid conditions. Feasibility and robustness of the developed VF-PDPC are verified by experiments. Mainly, obtained results demonstrate that the startup procedure reduces settling time and avoids over currents; the VF-PDPC achieves better current waveforms than those obtained by the conventional PDPC under unbalanced grid conditions; and the NN-VF estimator presents similar dynamic performances as the second-order generalized integrator which uses grid voltages measurement.