Abstract The phase-locked loops (PLLs) are one of the frequently employed algorithms to precisely estimate amplitude of voltage, frequency and phase of single phase grid connected Solar PV (GCSPV) array based system. However, there is deviation in PLL’s output during extraction of fundamental voltage component being hampered by presence of grid side harmonics, DC Offset and other disturbances occurred. Therefore, for the first time Notch Filter (NF) based modified cascaded second order generalized integrators (MCSOGI)-PLL with progressive control over frequency is designed for removing the error present in phase undisturbing its response which must be dynamic in nature. The proposed method systematically embeds the notch filtering mechanism within the multi-cascaded SOGI framework itself. Also, this work limelight’s the effect of proposed algorithm on DC Offset and harmonics through transfer function modeling via stability analysis (bode, polar & root locus plots) while selecting appropriate value of damping factor. The effectiveness of proposed technique is confirmed by thorough analysis through various cases of grid disturbances via Matlab/Simulink simulations. Results revealed that the proposed method has been effectively enhancing the power quality by complying with the IEEE 519 standards (THD grid side 10.96% & load end 2.65%). Hence, the proposed work approaches to significant improvement in stability and providing dynamic response via better filtering abilities.

Abstract The train control system is a critical piece of equipment in the field of high-speed railway signaling. Timely and comprehensive mastery of its health status, shifting daily maintenance from breakdown maintenance to condition-based maintenance, plays a vital role in guaranteeing safe train operation and improving equipment life. A CNN-Transformer encoder-BiLSTM serial-parallel neural network is proposed to predict and evaluate the health status of train control equipment. First, the collected multi-dimensional degradation status information is subjected to multi-scale local feature extraction by a Convolutional Neural Network (CNN). The improved Transformer encoder module performs feature extraction through parallel multi-head attention. By introducing three different attention mask mechanisms into the Transformer encoder, it focuses on data at different positions respectively to extract their correlations. Secondly, the Bidirectional Long Short-Term Memory (BiLSTM) module extracts, memorizes, and processes the degradation information with positional encoding input to the Transformer encoder to extract the correlation of historical sequences, thereby improving prediction accuracy. Meanwhile, an improved particle swarm optimization algorithm is introduced to establish dynamic nonlinear inertia weights to find the optimal hyperparameter combination of the model. Finally, the regression layer outputs the predicted value of the Health Index (HI) percentage. The model is capable of conducting real-time and accurate health status prediction within the entire service life of the equipment, which will assist the ground maintenance site in timely maintaining equipment triggering health warnings, thereby ensuring train operation safety while significantly reducing maintenance workload and financial costs. Experiments indicate that the prediction accuracy of the model reaches over 96%, providing a new perspective for the field of health status prediction based on deep learning models, and possessing important reference value for the safe maintenance of train control equipment and the assurance of train safe operation.

Abstract In engineering practice, surrounding rock is commonly dissected by joints and thus discontinuous, which can readily trigger instability and failure. Prestressed rock bolts provide effective reinforcement for jointed surrounding rock; however, current design still relies largely on experience and engineering analogy. Therefore, a quantitative evaluation of their support effectiveness is necessary to guide engineering practice. Through the combination of theoretical analysis and numerical simulation, the stress compensation of the rock bolt is analyzed quantitatively. The results show that: (1) The vertical stress compensation of surrounding rock in the two models with and without joints of a single rock bolt is 16kPa and 20kPa respectively. (2) The vertical stress compensation of surrounding rock in the two models with and without joints of multiple rock bolts is 41% and 43% higher than that of single rock bolt, respectively. (3) Compared with the single rock bolt model, the normal and tangential stresses of the joint surface of the multiple rock bolts model increased by 48% and 57%, respectively. (4) The stress compensation of the system rock bolt to the jointed surrounding rock of the tunnel is 0.2~0.3MPa. (5) Although the direct stress compensation value of prestressed rock bolt for jointed surrounding rock is not high, which is far lower than the rock strength, through the direct shear test, the mechanism of prestressed rock bolt is obtained: under the condition of multiple rock bolts and multiple sets of joint surfaces, 100kN prestress is applied, the cohesion of rock mass is increased by about 13.5%, and the internal friction angle is increased by about 40.3%. Under the action of prestressed rock bolt, the overall stability of surrounding rock is greatly improved.

Abstract This study presents the development of a Bidirectional KY (BKY) converter tailored for integration into electric vehicle (EV) drivetrain systems. The converter operates in continuous conduction mode (CCM), supplying power from a battery pack to a BLDC motor via the DC bus. Voltage gain analysis is conducted with consideration of non-idealities in semiconductor devices and passive components. System stability is validated through Bode and phase margin plots. Due to frequent dynamic variations in EV drivetrain operation, the DC bus voltage tends to fluctuate, necessitating robust regulation. To address this, an Average Current Mode Control (ACMC) strategy is employed, utilizing state-space modelling to derive realistic loop gain transfer functions for both current and voltage control loops. Furthermore, a passivity-based control (PBC) scheme integrated with a nonlinear disturbance observer (NDO) is proposed to reduce steady-state error and improve voltage regulation. Comparative analysis reveals that the PBC+NDO approach offers superior transient response and sustained voltage stability across varying operating conditions. The proposed converter and control strategies are implemented and evaluated using MATLAB/Simulink and dSpace under diverse driving scenarios.

Abstract The rapid development of Cloud-IoT computing environments enables intelligent services, but raises serious privacy and trust challenges due to massive distributed data generation. This paper proposes a verifiable multi-layer privacy-preserving Cloud-IoT computing framework that integrates differential privacy, secret sharing, and gradient masking within a cloud-edge-end collaborative architecture. An adaptive differential privacy mechanism dynamically adjusts noise intensity according to data sensitivity and training dynamics, while edge intelligence supports efficient pre-aggregation and privacy measurement. Extensive experiments in a real Cloud-IoT environment with 200 terminal devices demonstrate that the proposed framework improves model convergence speed by 37.8%, reduces communication overhead by 89.1%, and decreases privacy leakage risk by up to 82.9% compared with the DP-FedAvg and SecAgg baselines. Meanwhile, it maintains 91.3% model accuracy, suppresses membership inference attack success rates to 52.1%, which is close to the random-guessing baseline (50%), indicating that the attacker’s advantage is largely suppressed. The framework introduces only 3.2% additional verification overhead through a lightweight zero-knowledge proof mechanism. These results indicate that the proposed approach effectively balances privacy protection, verifiability, and system efficiency, providing a practical solution for large-scale Cloud-IoT applications in privacy-sensitive domains such as healthcare and financial services.

Abstract Laser-induced graphene (LIG) electrodes have gained interest for non-enzymatic analyte sensing applications due to their low cost, flexibility, and favorable electrochemical properties. In this study, we investigate the effects of Ag and Pt nanoparticle (NP) surface modification on the capacitive glucose sensing performance of LIG electrodes. LIG was fabricated at varying laser fluences and subsequently decorated with different NP loadings to evaluate interfacial capacitance changes in 1 M KCl electrolyte. Electrochemical characterization revealed that a 10 μg/mL NP loading concentration yielded the highest sensitivity -0.0449 and 0.052 mF/mM•cm² for Ag-and Pt-modified LIG, respectively -with the lowest detection limit of 1.29 µM achieved by Ag-LIG. Although the study is limited to controlled electrolyte conditions without interference or selectivity tests, the results demonstrate a measurable enhancement in glucose sensitivity compared to pristine LIG. These findings support the potential of metal-NP loaded LIG for cost-effective glucose sensing in high-concentration environments such as food fermentation monitoring.

Abstract Target detection based on unmanned aerial vehicle (UAV) in complex scenes has increasingly become a research hotspot. This paper proposes a new lightweight infrared target detection algorithm based on YOLOv5n for UAV imagery. The proposed algorithm increases the detection accuracy by improving the network structure and loss function. First, a lightweight FasterNet block is introduced into the backbone network to amplify and refine the feature information, and meanwhile the network depth is increased through block stacking to enhance the feature extraction capability. Second, a new module called efficient convolutional block (ECB) is designed and then integrated into the neck network to strengthen the feature fusion capability, which is beneficial for complex environments including dense and small targets. Finally, the complete intersection over union (CIoU) in the original loss function is replaced by the minimum point distance intersection over union (MPDIoU) to effectively estimate the positional error of the target bounding box, so as to improve the localization ability. Experimental results demonstrate that the proposed network contains only 3.1M parameters yet achieves significant performance gains over the baseline across two public benchmarks: on the DroneVehicle dataset, it improves mAP@0.5 by 4.0% and mAP@0.5:0.95 by 3.6%; similarly, on the HIT-UAV dataset, it achieves 5.0% and 1.9% improvements in mAP@0.5 and mAP@0.5:0.95 respectively. Furthermore, our method outperforms several state-of-the-art lightweight networks in consideration of detection performance and resource requirements, demonstrating the potential of the proposed algorithm for UAV visual perception tasks in infrared scenes, such as disaster rescue, traffic monitoring and wildfire detection.

Abstract The current investigation comprises the characterization of waste Lintz Donawitz (LD) slag/polyester composites to obtain a sustainable material. The effect of filler loading in the polyester resin is studied by preparing the sample with LD slag content varied from 0 wt. % to 40 wt. %. The prepared samples undergo various testing, which includes evaluation of different properties. The density of the polyester increases with LD Slag content. The experimental density is compared with the theoretical one and found that the difference between the two increases with the LD slag content, reaching a maximum value of 5.85% which depicts the presence of void content. The polyester filled with LD slag absorbs more water than unfilled polyester, and the increment rate depends on LD slag content and immersion time. Among the different filler loadings, the optimized tensile and flexural strength is obtained for 30 wt. % LD Slag, which are 37.8 MPa and 59.2 MPa, respectively. The compressive strength shows an increment of 26.21 %, and for hardness, it is 13.15 %, for 40 wt. % filler. The Taguchi method is used to design the experiment for conducting a wear test as per the L25 orthogonal array. The experiment reveals that the filler quantity and sliding velocity significantly govern wear, whereas the distance of sliding and applied load show a minor effect. To understand the wear behaviour beyond the experimental domain, a three-layer neural network model is developed with 10 hidden layers. The deteriorated part of the specimen post-sliding wear was analyzed to understand the mechanism of material removal.

Abstract In strongly coupled coal port environments, traditional scheduling models face a fundamental conflict between search efficiency and multi-objective global optimality. Existing methods often struggle because the feasible solution space is sparse, leading standard evolutionary algorithms to expend excessive computational resources on invalid solutions. To resolve this, this study proposes a Topology-Aware Cooperative Optimization Framework (HSA-ENSGA-II). Unlike simple hybridizations, this framework establishes a hierarchical "guidance-evolution" mechanism. First, a specialized A*-Beam Search (A*BS-BFS) module constructs a high-quality feasible subspace by pruning invalid search branches in the task-resource graph. This process effectively projects the search onto a sparse subgraph of valid solutions, preventing the combinatorial explosion typical of constrained scheduling. Second, an enhanced NSGA-II, incorporating Adaptive Hierarchical Selection (AHS) and Variable Neighborhood Operators (VNOs), performs global optimization within this refined space. Crucially, a dynamic feedback mechanism reinjects unexplored heuristic branches when evolutionary stagnation occurs, distinguishing this approach from simple heuristic initialization. Theoretical complexity analysis and ablation studies confirm that the heuristic guidance acts as a computational catalyst, reducing total runtime by ~20% not merely through initialization, but by actively reshaping the search landscape. Experimental validation on DTLZ benchmarks (IGD: 0.008) and proprietary port data demonstrates that the model achieves a hypervolume of 0.928. By reducing average waiting time to 29.8 minutes and fuel consumption to 603 L, the proposed framework successfully resolves the conflict between search efficiency and global optimality.

Abstract Municipal solid waste incineration (MSWI) fly ash poses environmental risks due to its high soluble chloride content and heavy-metal leachability. Conventional organic chelation treatments often exhibit poor long-term stability and occupy large landfill volumes. To overcome these limitations, this study developed an inorganic mineralization and mechanical compaction treatment, and established a service-life prediction approach based on carbonation behavior. Chemical and mineralogical characteristics were analyzed using XRF, XRD, and SEM, while heavy-metal leaching tests were performed in accordance with GB 16889. The resulting volume-reduced solidified bodies (VRSB) exhibited a density of 2.0 t/m3, a compressive strength of 4.19 MPa, and a volume reduction rate of 65%. Heavy-metal leachate concentrations were reduced below regulatory limits following treatment. Accelerated carbonation tests showed a carbonation depth of 2.6 mm after 28 days, and modeling predicted a natural carbonation depth of 21.2 mm after 100 years, equivalent to approximately 2.1% of the specimen dimension. These findings indicate that the VRSB maintain long-term heavy-metal immobilization and exhibit favorable durability under landfill conditions. Overall, the proposed method enhances landfill space efficiency, mitigates groundwater contamination risks, and provides a scalable and practical solution for long-term MSWI fly ash stabilization.