Metal fiber surface combustion technology has become one of the most promising low nitrogen combustion technologies for gas boilers because of its advantages of stable combustion process and uniform temperature distribution. Based on the low-nitrogen combustion experimental data of metal fiber surface in limited space, a three-dimensional physical model of metal fiber surface combustion was established, and the full-premixed combustion numerical solution was carried out based on the porous medium resistance model, metal fiber turbulence model and EDC combustion model. The flame combustion conditions and the velocity characteristics of hot smoke flow field of metal fiber surface combustion in limited space were obtained. Temperature field distribution and fuel distribution in furnace. Considering the influence of different excess air coefficient, the emission characteristics and average generation rate distribution of fast burning NOx and thermal NOx on metal fiber surface were obtained. The results show that under different excess air coefficients, the generation rate of fast NOx is faster than that of thermal NOx. However, limited by the effective reaction space and reaction duration, the volume occupied by fast NOx in the furnace is much smaller than that of thermal NOx.With the increase of excess air coefficient, the generation rate and emission of thermal NOx gradually decrease. When α=1.6, the emission of NOx from flue gas outlet is 22.55 mg/m³, which meets the standard of low nitrogen combustion. Therefore, in addition to changing the flame temperature, oxygen concentration and combustion mode, the generation of thermodynamic NOx can be suppressed by controlling the excessive air coefficient, and ultra-low nitrogen combustion of industrial gas boilers can also be effectively achieved.

This study conducts a numerical simulation of a coaxial dual-chamber plasma generator based on MHD theory. The research aims to reveal the distribution characteristics of the arc-plasma and the behavior of the flow field, further analyzing the relationship between external parameters and arc behavior. An axisymmetric model is adopted, incorporating coupled calculations of the flow field and electromagnetic field to investigate the correlations among arc voltage, cathode spot distribution, and airflow parameters. The simulation results indicate that arc voltage is relatively insensitive to radial airflow; within the studied parameter range, fluctuations in radial airflow pressure have a maximum impact of only 4.2% on arc voltage. In contrast, arc voltage exhibits a strong positive correlation with axial airflow velocity, and a fitted correlation equation has been obtained. Temperature and velocity distribution analyses show that the maximum nozzle temperature exceeds 3500 K, ensuring sufficient ignition capability for low-quality coal and stable combustion performance. Moreover, the results confirm that with appropriate airflow settings, the coaxial dual-chamber structure enables the plasma igniter to achieve both high power and extended electrode lifespan. The study reveals the anti-ablation mechanism of this structure, wherein the alternating sweeping effects of the two airflow paths play distinct roles: axial airflow primarily regulates output power, while radial airflow controls arc root movement through periodic oscillations, effectively preventing single-point ablation and prolonging electrode life. Furthermore, an optimized design analysis has identified the equilibrium point between the two airflow paths, providing theoretical support for future research on the longevity of plasma generators.

To address the challenges of current over-limitation and limited voltage support capability in grid-forming energy storage converters during low-voltage ride-through (LVRT) under grid fault-induced voltage sags, this study proposes a reactive power coordination control strategy for a hybrid Virtual Synchronous Generator (VSG) and Static Var Generator (SVG) system incorporating an LVRT current limiting strategy. Initially, the typical control methodology of VSG is presented, followed by a transient characteristic analysis of its power-angle curve under various grid voltage dip scenarios. Subsequently, an improved q-axis priority current limiting strategy is developed, where the active power reference value is proportionally reduced according to the diminished d-axis current, resulting in an adaptive active power reference q-axis priority current limiting approach. Furthermore, to overcome the restricted voltage support capacity of standalone VSG during severe grid voltage sags, a collaborative control framework integrating SVG with VSG is established, accompanied by a reactive power allocation strategy. The effectiveness of the proposed methodology is conclusively validated through SIMULINK simulation results. The experimental results demonstrate that the proposed current-limiting strategy effectively restricts the output current, while the hybrid system control methodology significantly enhances voltage recovery performance during grid voltage sags.

A new power line communication (Power line communication, PLC) scheme named full carrier index differential chaos shift keying (Differential chaos shift keying, DCSK) with noise reduction (FCI-DCSK-NR) is proposed in the paper, which integrates both noise reduction technology and index modulation with DCSK to tackle the problem caused by complex power line channel, especially the impulsive noise. In the proposed scheme, all the carriers except the reference carrier are firstly divided into two major categories, i.e., the first active carriers and the second active carriers, where the first active carriers are chosen by the index bits and the remained carriers are the second active carriers. Then, the transmitted bits are delivered through the chaotic signal and its Hilbert transform successively by the first active carriers and the second active carriers. Besides, the scheme achieves noise reduction by transmitting the signal many times at the transmitter and averaging the received signals at the receiver. The bit error rate (Bit error rate, BER) expression of FCI-DCSK-NR over power line channel is derived, and its correctness is validated through Monte Carlo simulations. Both theoretical analysis and simulation results demonstrate that the proposed scheme can deal with the impulsive noise of power line channel effectively. It is shown that the proposed scheme can enhance the BER performance and data rate compared to traditional DCSK and generalized carrier index DCSK (Generalized carrier index DCSK, GCI-DCSK).

Short-term power load multi-step forecasting for commercial buildings plays a pivotal role in urban orderly power consumption and virtual power plant scheduling. The power load time series in commercial buildings is characterized by strong stochasticity, non-stationarity, and nonlinearity, and traditional iterative multi-step power load forecasting strategy suffer from error accumulation effects that degrade prediction accuracy., a short-term power load multi-step forecasting method based on Frequency Enhanced Channel Attention Mechanism (FECAM) -Sparrow Search Algorithm (SSA) -Informer is proposed. Based on the time-domain features output by the Informer encoder, the method use FECAM to adaptively model the frequency dependence between feature channels, further extracting the frequency-domain features of multi-dimensional input sequences. The decoder then integrates both time-frequency domain information to directly generate future multi-step load sequences. Furthermore, due to the lack of theoretical basis for the improved Informer hyperparameter settings, the SSA algorithm is used to optimize model hyperparameters such as learning rate, batch size, fully connected dimensions, and dropout rate. Experimental validation using annual load data from a commercial building in Guangzhou demonstrates that, compared with other deep learning models, the proposed model significantly improved prediction accuracy across varying forecast horizons (48-step, 96-step, 288-step, 480-step, 672-step), exhibiting superior performance in short-term power load multi-step forecasting.

In high-mobility scenarios, wireless communications undergo time and frequency doubly selective fading, making channel estimation essential for accurately obtaining channel state information (CSI), which in turn enhances the performance of communication systems. The Time-Frequency Doubly Selective Channel is a channel model that characterizes signal fading with selective properties in both time and frequency dimensions. To address the challenges of channel estimation in such environments, deep learning methods have been widely adopted in recent years. Networks that originally excelled in computer vision and natural language processing, such as Convolutional Neural Networks (CNN) and Long Short-Term Memory networks (LSTM), have been applied to channel estimation techniques. However, due to significant differences in data characteristics and task objectives between channel estimation and image processing, these approaches still face numerous challenges. This paper introduces a novel channel estimation deep learning algorithm based on an Channel Enhanced Deep Horblock Network (CEHNet). The proposed algorithm treats the time-frequency grid of the doubly selective channel as a two-dimensional image and employs a Super-Resolution (SR) network to reconstruct the CSI. Additionally, a data augmentation preprocessing method that increases amplitude features is utilized to expand the dataset, and Lasso regression is incorporated as a constraint to accelerate the network’s convergence speed. Experimental results demonstrate that, across various channel models, the proposed CEHNet algorithm outperforms traditional channel estimation methods such as Minimum Mean Square Error (MMSE) and conventional Super-Resolution Convolutional Neural Networks (SRCNN) when the number of pilots is limited. Furthermore, CEHNet exhibits significantly faster convergence rates, achieving a fourfold performance improvement over SRCNN at a signal-to-noise ratio (SNR) of 22 dB. These findings highlight the effectiveness of deep learning-based channel estimation techniques in complex time-frequency selective environments, offering substantial improvements in estimation accuracy and convergence efficiency.

In wastewater treatment processes, efficient modeling of key water quality-related parameters is crucial for achieving process control optimization, anomaly detection, and decision support. However, process data typically exhibit complex temporal dependencies, multivariable coupling nexus, and non-stationary characteristics, which pose significant challenges for modeling. To address these issues, this paper proposes a Lightweight Multi-Parameter Time Series Prediction Method for Wastewater Treatment based on the Smooth Wavelet Transform (SWT) and Collaborative Attention (CA). This method first performs multi-scale decomposition on wastewater data and uses the stationary wavelet transform to extract features from the data at different scales. Then, it constructs a collaborative attention mechanism based on geometric attention and sparse attention to efficiently capture the complex coupling relationships and temporal features among key water quality parameters. Finally, a dual projection layer maps the reconstructed features obtained from the inverse wavelet transform to the final prediction outputs. The model is trained and evaluated using real-world data collected from a wastewater treatment plant in Dongguan. Experimental results show that, in 12-step forecasting tasks, the proposed model achieves a reduction of 9.15% to 37.70% in RMSSD compared to baseline models. In other tasks, its accuracy is second only to TimesNet, which has a significantly larger parameter size. These findings demonstrate the model’s strong balance between lightweight design and predictive performance, validating its effectiveness for time series forecasting in wastewater treatment.

Recent advances in monaural speech separation leveraging self-attention mechanisms have demonstrated substantial improvements. Despite their superior capabilities in modeling long-range contextual dependencies, self-attention architectures exhibit limitations in preserving subtle acoustic characteristics, including temporal or frequency continuity, spectral structure, and timbre. Furthermore, existing single-paradigm attention frameworks lack effective mechanisms for multi-scale feature integration. To address these challenges, this paper proposes TCANet, an end-to-end time-domain comprehensive attention network incorporating local and global attention modules for monaural speech separation. Local modeling employs S&C-SENet-enhanced Conformer blocks to meticulously capture short-term spectral structures, timbral features, and other fine-grained acoustic details. Global modeling incorporates improved Transformer blocks with relative position embeddings to explicitly learn long-range dependencies within dynamic speech contexts. Furthermore, a dimension transformation mechanism bridges intra-block local features with inter-block global representations, thereby achieving cross-scale feature co-optimization. Extensive experimental results on benchmark datasets (LRS2-2Mix, Libri2Mix and EchoSet) show that the proposed method outperforms other end-to-end speech separation methods in terms of scale-invariant signal-to-noise ratio improvement (SI-SNRi) and signal-to-distortion ratio improvement (SDRi).

Ultra wideband (UWB) technology, as a model of the new generation of indoor positioning technology, is often combined with inertial navigation systems (INS) to optimize the non line of sight error (NLOS) problem in positioning in practical applications. However, centralized information processing methods cannot effectively distinguish the sources of NLOS errors, resulting in redundant positioning anchor points, information waste, and increased costs. This paper proposes a UWB/INS indoor positioning method based on self resetting genetic particle filter (SGPF) to address the problem of non line of sight error identification and elimination in indoor positioning. The method uses the SGPF algorithm as a medium to trace the NLOS error in the measured values through INS system estimation, improving the tracking stability in NLOS environment. Firstly, by grouping physical anchor points and combining them with virtual anchor points to partition likelihood regions; Then, the NLOS error identification strategy is used to preliminarily estimate the high probability areas through the INS system and eliminate the NLOS anchor group and measurement values; Finally, considering whether to enable genetic resampling to optimize the diversity of the particle set and improve the robustness of the algorithm based on the effective number of particles to determine the state of the particle set. The SGPF algorithm combines the structures of standard particle filtering and genetic algorithm, which can effectively alleviate the problems of particle degradation and impoverishment, and achieve higher robustness with lower particle count and time consumption. Experiments have shown that in line of sight environments, the SGPF algorithm only requires 30% of the particle count of the PF algorithm to achieve the positioning effect of standard particle filtering, and the time consumption is much lower than that of traditional genetic particle filtering. In non line of sight environments, the average positioning error of SGPF algorithm is 0.0552m, which is 56.97% and 48.94% higher than traditional particle filtering and traditional genetic particle filtering, respectively.

In the aerospace field, rigid-flexible coupling structures are widely used due to their high structural efficiency. However, the existence of the rigid-flexible coupling effect brings huge challenges to active vibration control. To solve this difficult problem, this paper conducts research on active vibration control as the research object. In terms of research methods, a vibration measurement and control platform for the three-flexible beam coupling system was first established. The detection and suppression of vibration signals were achieved by using piezoelectric sensors and drivers. On this basis, the design of vibration measurement and control algorithms was carried out. Subsequently, the system dynamics model was established by combining the finite element method with the Hamiltonian variational principle. The main modal shapes of the free vibration of the system were determined in the simulation environment. After introducing the modal coordinates, the modal truncation method was adopted to obtain the state space equation of the system. Meanwhile, in view of the uncertainty of model parameters, wavelet analysis and jumping spider optimization algorithm were used to accurately identify the parameters of the system state space equation. Furthermore, considering the nonlinearity and parameter uncertainty of the system, a fuzzy logic controller based on Gaussian membership function was designed to suppress the vibration of the flexible beam. The simulation and experimental results show that within the same control saturation voltage period, the fuzzy logic controller performs better than the large gain PD(Proportional and Derivative) control in suppressing the vibration of the coupled three-flexible beam. While rapidly suppressing large-value vibrations, it can suppress small-value vibrations at a faster speed. It effectively shortens the time for the system to reach a stable state and significantly improves the vibration control effect. To sum up, the fuzzy logic controller based on Gaussian membership function designed in this paper overcomes the problems of nonlinearity and parameter uncertainty in the vibration control of rigid-flexible coupling structures. Compared with the traditional high-gain PD control, it shows stronger adaptability and higher control efficiency.

Elevator traction machines generate high sound pressure level noise during the braking process. This study proposes a vibration and noise reduction solution based on particle dampers. Firstly, the vibration characteristics of the brake wheel and brake pads are investigated through finite element analysis. Key vibration modes are identified by correlating the principal vibration frequencies obtained from whole-machine vibration and noise tests. Based on these findings, and considering the symmetry and spatial layout of the brake wheel structure, an innovative cavity design is introduced within the brake wheel to accommodate particle dampers. The study employs coupled EDEM-ADAMS simulation technology to optimize the parameters of the solid particles, with a focus on addressing three critical technical issues: (1) To avoid interference from magnetic fields, pure aluminum is ultimately selected as the damping particle material; (2) The energy dissipation process of the particle damper within the cavity is simulated using discrete element analysis; (3) The particle radius and filling ratio of the damper are optimized by integrating multi-body dynamics simulations. Experimental validation is conducted in a semi-anechoic chamber, where a timed braking control strategy with a 5-second cycle is implemented. A triaxial sensor array is used to collect vibration signals, while sound pressure level data are recorded synchronously. The test results indicate that, following the installation of particle dampers, the average sound pressure level during the braking process of the traction machine is reduced by 20.7%, thereby confirming the effectiveness of the proposed solution. This research provides a novel technical approach for noise control in electromagnetic braking systems and demonstrates significant engineering application value.

Lightweight and high-performance aluminum alloys are crucial for the weight reduction design of aerospace equipment. Consequently, spray forming with rapid solidification technology has garnered increasing attention for the fabrication of high-strength aluminum alloys. To meet the demands of large-scale aerospace components, a multi-nozzle collaborative system is required to achieve larger billet diameters. During the scanning deposition process, where atomization cones from multiple nozzles intersect the deposition interface at specific inclination angles, it is essential to ensure uniform distribution of the molten material and maintain a flat, stable growth of the billet’s top surface. These factors are key to producing high-quality billets with dense and uniform microstructures. The process parameters associated with multi-nozzle configurations directly influence the scanning trajectories of atomized droplets and the material deposition state at the interface, playing a decisive role in billet growth. Aiming to produce large-scale billets with consistent morphology and uniform deposition quality, a multi-nozzle Deposition Surface Behavior Model (DSBM) is established based on microscale scanning deposition heights. Additionally, the GA-DSBM intelligent optimization method is employed to simulate, analyze, and optimize key process parameters during deposition. BY the optimized parameters, spray forming experiments were conducted, which shows that the surface unevenness was lower than 7.52 mm when spray forming a billet with a diameter of 600 mm. It meets the design requirements and confirming the feasibility of the intelligent optimization method.

As the primary actuator in hydraulic systems, motors generate significant noise radiation that increasingly fails to meet low-noise requirements. Furthermore, the lack of clarity regarding the primary noise sources and the low localization accuracy in current methodologies have resulted in persistently unsatisfactory noise reduction effectiveness. To clarify the main sources of motor noise and improve the localization accuracy, a multi-physics approach was implemented. First, a fluid dynamics model of the motor was established using Pumplinx to analyze fluid-induced vibration forces at the valve plate. Co-simulation of ADAMS and AMESim was conducted to capture vibration forces generated by piston-cylinder collisions during motor operation. Transient dynamic analysis in ANSYS was then employed to obtain vibration displacement responses on the motor housing and rear cover surfaces. These vibration results were applied as acoustic boundary conditions in LMS Virtual.Lab, combined with the Boundary Element Method (BEM), to simulate the motor’s acoustic field, identifying the primary noise sources and dominant regions. A dedicated acoustic intensity test bench was designed to acquire noise distribution maps, validating the multi-physics co-simulation results. The Regularized Orthogonal Matching Pursuit (ROMP) algorithm, integrating observation matrices, sparse representation, and reconstruction techniques, was further utilized to refine noise localization accuracy. Final experimental verification confirmed the feasibility of the optimized algorithm. Results demonstrated the accuracy of the multi-physics model, revealing that the main noise originates from pressure impacts at the valve plate and piston collisions, with the valve plate region identified as the primary noise source. The refined localization precision reached 25 mm, achieving enhanced determination and spatial resolution of motor noise sources.