Zhejiang Tongji Vocational College of Science and Technology

With the rapid development of the Chinese power grid, its dynamic behavior becomes increasingly complex. How to prevent voltage instability and voltage collapse at the receiving-end system has become an urgent problem at present. To prevent voltage collapse, the key is to determine the stability margin at the operating point of the power system, and the load margin is the most basic and widely accepted voltage stability margin indicator. Considering the role of medium and long-term slow-acting elements in the actual system that affect the system voltage stability, this paper combines the static voltage stability analysis with the transient, medium, and long-term dynamic voltage stability analysis based on whole process simulation, and uses time-domain simulation methods to evaluate the voltage stability of the actual system. By analyzing the various dynamic characteristics of generator overexcitation limits and transformer on-load regulation on the development process of voltage-collapsing accidents, this paper proposes an evaluation index for the whole process dynamic stability margin of the receiving-end system. On this basis, the whole process dynamic defense system of the power grid is also constructed. The proposed method is validated by using a provincial power grid in China as an example.

Feature selection plays an essential role in the field of computer vision. Current research involves the adjustment of feature subsets to minimize the dissimilarity between the feature space and selected feature subset, thus enhancing the quality of the chosen features. However, the optimization process encounters challenges related to extended computation times. To address this challenge, this study introduces an $$L_{2,0}$$ sparse constraint and presents a greedy feature selection approach utilizing a residual downhill strategy to enhance computational efficiency, while not compromising model accuracy. The residual downhill can quickly reduce the loss of the objective function, and the sparse regularization can filter out irrelevant features. This study aims to rigorously evaluate the effectiveness of the model by utilizing various data sets, including six publicly gene datasets related to diseases (e.g., Leukemia etc.), four additional classification challenge datasets (e.g., Arcene, etc.), and five image datasets (e.g., VOC-2007, etc.). KNN and SVM with five-fold cross validation are implemented as classifiers to assess their efficacy. The performance of the model is comprehensively assessed by three evaluation metrics such as classification accuracy. Finally, the results are tested for significance to verify that the improvement is significant based on P-value. Due to the efficacy of this method, it is necessary to further investigate it and explore its applications in other domains.

Algal blooms not only lead to significant water pollution but also produce algal organic matter (AOM) from algal cells, which serves as precursors for disinfection byproducts. To address the inefficiency of existing algae removal technologies in eliminating AOM, this study investigates the feasibility of a combined technique utilizing chitosan quaternary ammonium salt (HTCC) and activated carbon (AC) for simultaneous algae removal and AOM elimination. First, the release of AOM during algae removal by standalone HTCC was examined. Second, AC with optimal AOM removal efficiency was selected, followed by evaluating the performance of the HTCC–AC combined technique in algae removal and AOM elimination. The disinfection byproduct formation potential before and after treatment was analyzed. Mechanism insights were elucidated through characterization techniques including Zeta potential, scanning electron microscope, Brunauer–Emmett–Teller, and FT-IR. Results indicated that standalone HTCC treatment poses a risk of AOM release, with released AOM originating from both intracellular and extracellular algal components. Wood-based activated carbon exhibited the highest AOM adsorption capacity at 6.43 mg/g. The HTCC–AC combined technique achieved removal efficiencies of 97.32%, 96.11%, and 91.13% for algal density, turbidity, and dissolved organic carbon, respectively. Notably, trihalomethane formation decreased by 94.56% compared with the control group, resulting in a posttreatment THM ratio of 0.15, significantly below the 1.0 limit specified in drinking water standards. The mechanisms of algal removal through the combined technique involve adsorption bridging and charge neutralization, while the removal of AOM occurs via π–π conjugation and hydrogen bonding interactions.

To improve the accuracy of deformation prediction in concrete buildings and large-scale infrastructures such as dams, this study proposes an Attention-BiLSTM model integrated with a parameter-optimized Variational Mode Decomposition (VMD). Specifically, the Pelican Optimization Algorithm (POA) is employed to optimize VMD parameters, enhancing signal decomposition efficiency for structural deformation time series. The optimized VMD is then coupled with a BiLSTM network embedded with an attention mechanism, forming a hybrid prediction framework that captures both temporal dependencies and key feature weights in monitoring data. Using three sets of engineering-measured deformation datasets, the proposed model is validated through comparative analyses with conventional single models (e.g., standalone BiLSTM and VMD-BiLSTM without attention). Results demonstrate that the developed model achieves superior accuracy and stability, significantly outperforming all comparative methods, with the highest R2 reaching 0.996, while reducing MAE and RMSE by over 60% and 30%, respectively. Quantitative evaluation indicators (e.g., RMSE, MAE, and R2) confirm that the approach effectively captures both short-term fluctuations and long-term trends of structural deformation. These findings verify its reliability and applicability for intelligent safety monitoring of concrete buildings and infrastructures.

Context Nitrogen and phosphorus contaminants from paddy field drainage flow into the receiving bodies of water, posing a risk of eutrophication. To mitigate this pollution risk, several technologies, including constructed wetlands, ecological revetments and ecological floating beds, have been introduced in drainage ditches. However, the capacity of these technologies to intercept pollutants need to reconstruct the existing ditch structure. Aims This study aimed to optimise the interception and purification capabilities of drainage ditches through matrix dams. Methods A hydrodynamic and water-quality numerical model was established using the Mike 21 software, and the comprehensive ability of purifying pollutants by the interception of matrix dams and drainage ditches was evaluated. Key results We found that the overall average flow rate was reduced by 84% in the paddy field drainage ditch after introducing matrix dams, and pollutant concentrations decreased progressively from upstream to downstream. Total nitrogen concentrations were reduced by exceeding 60%, with a maximum reduction in total dissolved nitrogen concentrations of 84.3%. Conclusions Results indicated that the the combined deployment of drainage ditches and matrix dams is effective in removing nitrogen and phosphorus pollutants from paddy field drainage. Implications This study can provide a theoretical foundation for the practical implementation of matrix dams.

This paper proposes a Hard Constrained Physics-informed Neural Network (HardPINN) integrated with Slow Feature Analysis (SFA) to address the challenging problem of fault diagnosis in pump-turbines under load rejection transient conditions. During load rejection, the strong multi-physics field coupling characteristics of the water-machine -electric system result in inaccurate transient flow patterns modeling using traditional physical models. Meanwhile, existing data-driven methods suffer from poor generalization performance and insufficient extraction of weak fault features due to neglecting physical mechanism constraints. To address these issues, HardPINN embeds core physical laws, such as the continuity and momentum equations, into the neural network loss function as strong constraints, enabling dynamic modeling of multi-physics field coupling. Additionally, SFA is introduced to suppress transient high-frequency noise interference and enhance sensitivity to slow-varying fault features. Experiments on pump-turbine datasets of load rejection condition validate the effectiveness of HardPINN, which achieves an average accuracy of 90.72% in diagnosing three typical faults (penstock rupture, guide vane asynchrony, and runner imbalance) and outperforms SFA (85.87%), CNN (80.69%), and GMM (38.66%). By implementing mechanism-data collaborative modeling, HardPINN resolves the gradient conflict between transient responses and long-term degradation features, providing a robust solution for the intelligent operation and maintenance of hydropower systems.

In response to the urgent need for high-precision 3D modeling of complex underground pipelines and dense ground buildings in cities, this article innovatively proposes and implements the "Space Ground Intelligent Connected Centimeter Twin System".The system deeply integrates the robot dog with three major technology platforms: a handheld laser scanner (underground), a gantry laser scanner (ground), and an unmanned aerial vehicle airborne laser radar (aerial), breaking through key technologies such as real-time fusion of multi-source heterogeneous point clouds, unified global spatial benchmarks, and AI-driven intelligent diagnosis.Efficient and blind spot collection of GNSS free underground pipe galleries is achieved through autonomous navigation of robotic dogs and SLAM technology; Establish millimeter level ground benchmarks using gantry scanners and coordinate calibration of drone and robotic dog data; Innovatively adopting a multi-source data confidence dynamic weighted fusion algorithm to achieve cross scale spatiotemporal alignment and realtime fusion of "air ground underground".

Four aquatic plants and three veterinary antibiotics were selected to construct a hydroponic test system for analyzing the tolerance, removal efficiency, and mechanisms of antibiotics. The results indicated that antibiotic concentrations ranging from 0.2 to 200 μg·L−1 promoted plant height and leaf chlorophyll content, while concentrations of 600 μg·L−1 and 1200 μg·L−1 had inhibitory effects. The antibiotic removal efficiency from the hydroponic solution by different plants followed this order: Iris wilsonii (16.0%–57.3%) > Hydrilla verticillata (10.0%–48.0%) > Lythrum salicaria (9.0%–45.0%) > Nymphoides peltata (8.0%–40.2%). The plants that exhibited the highest removal of ciprofloxacin, sulfamethopyridazine, and oxytetracycline after 16 days of hydroponic cultivation with 200 μg·L−1 antibiotics were Hydrilla verticillata, Iris wilsonii, and Iris wilsonii, with removal rates of 48.0%, 51.1%, and 57.3%, respectively. Antibiotic accumulation in different plant tissues followed the order: root > stem > leaf, with accumulation increasing over time. The diversity of rhizosphere microorganisms decreased as antibiotic treatment concentrations increased, while the abundance of Aeromonas, Bacillus, Lysinibacillus, and Staphylococcus exhibited an increasing trend. These findings suggest that both antibiotic uptake by plants and the dynamics of the rhizosphere microbial community contribute synergistically to antibiotic removal.

Let G be the unramified unitary group [Formula: see text] over a non-archimedean local field F of odd residue characteristic p, and let B be the standard Borel subgroup of G. In this paper, we study the problem of the restriction of irreducible smooth [Formula: see text]-representations of G to B, and we prove results which are analogous to that of Paškūnas on [Formula: see text] [V. Paškūnas, On the restriction of representations of [Formula: see text] to a Borel subgroup, Compos. Math. 143(6) (2007) 1533–1544, MR 2371380 (2009a:22013)].