As a key enabling technology for sixth-generation (6G) wireless communication systems, the simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) can dynamically reconfigure its electromagnetic elements to effectively extend wireless network coverage. The deployment of STAR-RIS in vehicle-to-vehicle (V2V) communications effectively mitigates vehicle non-line-of-sight transmission issues and improves the reliability of vehicular communication. This paper investigates two typical multiple access schemes, namely non-orthogonal multiple access (NOMA) and rate splitting multiple access (RSMA). Considering both perfect and imperfect successive interference cancellation (SIC) conditions, we adopt hybrid automatic repeat request (HARQ) technology to further enhance communication reliability. We analyze the outage performance of receiving vehicles in HARQ-aided STAR-RIS-NOMA/RSMA systems and derive the closed-form expressions for outage probability. Numerical simulations validate the correctness of the theoretical derivations and reveal the following results. First, for both multiple access schemes, RSMA delivers better system performance than NOMA, and the systems under perfect SIC outperform those under imperfect SIC. Second, the introduction of HARQ can significantly boost the performance of STAR-RIS-NOMA/RSMA V2V communication systems.
This paper deals with the direct torque control (DTC) of an induction motor. This control allows torque flux to be decoupled without the need for coordinate transformations. It has the advantage of being robust to parametric variations of the machine in tracking mode, and to load disturbances in regulation mode. However, in addition to torque and flux ripples, this control in its conventional version proves insufficient at low speeds due to the demagnetizing effect of the stator flux caused by the ohmic drop across the stator resistor. This demagnetization considerably increases torque and flux ripples and causes current distortions. To alleviate this problem, it is necessary to shift back the flux positions in the (α, β) plane by an appropriate angle. The main objective of this study is to find this offset angle without affecting the entire order in its original version. In speed loop, the parameters of the PI controller are optimized using the Particle Swarm Optimization (PSO) algorithm.
This study evaluates the potential of the lead-free double perovskites CsGeI3 and CsSnI3 for high-performance photovoltaic applications through a combined approach integrating SCAPS–1D solar cell modeling with density functional theory (DFT). Using the Au/CuI/CsGeI3/CsSnI3/ZnSe/FTO device configuration, SCAPS–1D simulations identified optimal absorber-layer parameters, including thicknesses of 1000 nm for CsGeI3 and 700 nm for CsSnI3, a doping level of 1019 cm−3, and a defect density of 1014 cm−3. Under these optimized conditions, the device achieved high performance, delivering a Voc of 1.17534 V, Jsc of 30.54 mA·cm−2 FF of 88.08%, and a PCE of 31.61%. To complement these device-level results, DFT calculations were performed to examine the structural, mechanical, thermodynamic, and electronic characteristics of both materials. The optimized lattice parameters, elastic constants, and derived mechanical moduli confirmed structural integrity and ductile behavior, while the calculated sound velocities and Debye temperatures supported their thermal stability. Electronic-structure analyses revealed narrow mBJ band gaps of 0.71 eV for CsGeI3 and 0.69 eV for CsSnI3, indicating strong absorption capabilities across the visible-NIR region. This work presents the first application of a CsGeI3-CsSnI3 dual-absorber solar cell, offering a promising strategy for improving light absorption, carrier transport, and overall device performance. Furthermore, these combined results demonstrate that CsGeI3 and CsSnI3 possess favorable optoelectronic properties and high simulated photovoltaic efficiency, making them excellent candidates for future eco-friendly solar-energy technologies.
This research explores machine learning (ML) techniques, with an emphasis on neural networks (NNs), to predict household energy consumption in the context of smart grids. The study evaluates algorithms such as CNN, LSTM, RF, AdaBoost, and CatBoost with a primary focus on hybrid models to enhance forecast precision. The research adopts a comprehensive data collection strategy while dividing the time series data into training (80%) and testing (20%) segments for rigorous model evaluation. Hybrid approaches that combine several algorithms significantly outperform single models, and LSTM-CatBoost is at the forefront by exhibiting lower error rates and higher R2 values. To avoid overfitting, the research employs cross-validation in conjunction with early stopping, thus, producing robust and trustworthy models. The paper provides a wealth of information on energy forecasting that paves the way for the energy efficiency and sustainability of smart grids, which are essentially energy management systems based on the latest technology and consumer feedback mechanisms.
This study proposes an anomaly detection framework that combines CNN–LSTM feature extraction with a boosting-based ensemble strategy to improve the reliability of photovoltaic (PV) system monitoring. Real multi-source PV operational data are first preprocessed using the ISODATA clustering algorithm, which automatically adjusts the number of clusters and reduces redundancy. Principal component analysis (PCA) is then applied to lower data dimensionality while retaining key variability. A hybrid CNN-LSTM network is developed, where CNNs extract spatial features from heterogeneous PV measurements and LSTMs capture temporal dependencies in power sequences. Based on the learned representations, an ensemble model integrates the outputs of Gaussian Mixture Models (GMM), Isolation Forest (IF), and Interquartile Range (IQR) through a boosting-inspired weighting mechanism to enhance robustness under complex operating conditions. Experiments conducted on real PV datasets show that the proposed method achieves nearly 97% anomaly detection accuracy, with an average F1-score of 0.89 ± 0.03 and a recall rate of 0.91 ± 0.02. Compared with single-model baselines, the framework provides more stable performance and maintains a false positive rate below 2.1%, demonstrating its practical value for real-world PV anomaly detection.
