Vibration signal analysis is regarded as a fundamental approach in diagnosing faults in rolling bearings, and recent advancements have shown notable progress in this domain. However, the presence of substantial background noise often results in the masking of these fault signals, posing a significant challenge for researchers. In response, an adaptive denoising autoencoder (ADAE) approach is proposed in this paper. The data representations are learned by the encoder through convolutional layers, while the data reconstruction is performed by the decoder using deconvolutional layers. Both the encoder and decoder incorporate adaptive shrinkage units to simulate denoising functions, effectively removing interfering information while preserving sensitive fault features. Additionally, dropout regularization is applied to sparsify the network and prevent overfitting, thereby enhancing the overall expressive power of the model. To further enhance ADAE’s noise resistance, shortcut connections are added. Evaluation using publicly available datasets under scenarios with known and unknown noise demonstrates that ADAE effectively enhances the signal-to-noise ratio in strongly noisy backgrounds, facilitating accurate diagnosis of faults in rolling bearings.

Blast waves generated from the muzzles of weapons may exert negative effects, such as shock waves and impulse noise. If the weapon is fired with a muzzle brake, these effects are recognized to be more severe. This paper discusses the influence of the muzzle brake on certain aeroacoustic noise characteristics based on numerical simulations and a corresponding experiment. The impulse noise, which is induced by complex jet flows discharging from small caliber rifles with muzzle brakes, is focused in this study. Computational fluid dynamics (CFD) and computational aeroacoustics (CAA) are combined to calculate the muzzle flow field and jet noise for cases with and without a muzzle brake, and then the data sets are carefully compared. The simulations indicate that the muzzle brake alters the muzzle flow field and directional distribution of the jet noise compared to a rifle sans muzzle brake. Deviations less than 7.6% between experimental data and simulation results validate the simulation model. The results presented in this paper may provide a workable reference for the prediction of muzzle noise and the optimization of muzzle brake designs.

The performance of electromagnetic launch (EML) systems is generally evaluated in terms of two key parameters: the velocity of the projectile and the overall energy utilization efficiency. However, existing multistage EML circuits are constrained by their topology, which limits the timely reutilization of residual energy and thereby restricts further improvements in both projectile velocity and system efficiency. To address this issue, this study proposes a sequential in-process energy recovery topology (SIPERT). The proposed circuit topology is capable of collecting the residual energy remaining in the circuit after the completion of acceleration in each stage and reutilizing this energy in the subsequent stages of the launch process. Specifically, once the upstream coil conduction is completed, the induced electromotive force is used to recharge the downstream capacitor, thereby both accelerating the attenuation of ineffective freewheeling current and increasing the initial voltage of the subsequent capacitor. The fundamental principle of SIPERT is elucidated in detail, and its performance is validated by experiments on a dedicated test platform. Comparative simulations are further conducted against conventional, bridge-type, and final-stage recovery circuits. The results demonstrate that, relative to the bridge-type circuit, SIPERT achieves a 10.02% increase in projectile velocity and a 14.08% improvement in system efficiency. These findings provide valuable insights for the future design and optimization of EML systems.

Vibration signal analysis is regarded as a fundamental approach in diagnosing faults in rolling bearings, and recent advancements have shown notable progress in this domain. However, the presence of substantial background noise often results in the masking of these fault signals, posing a significant challenge for researchers. In response, an Adaptive Denoising Autoencoder (ADAE) approach is proposed in this paper. The data representations are learned by the encoder through convolutional layers, while the data reconstruction is performed by the decoder using deconvolutional layers. Both the encoder and decoder incorporate adaptive shrinkage units to simulate denoising functions, effectively removing interfering information while preserving sensitive fault features. Additionally, dropout regularization is applied to sparsify the network and prevent overfitting, thereby enhancing the overall expressive power of the model. To further enhance ADAE's noise resistance, shortcut connections are added. Evaluation using publicly available datasets under scenarios with known and unknown noise demonstrates that ADAE effectively enhances the signal-to-noise ratio in strongly noisy backgrounds, facilitating accurate diagnosis of faults in rolling bearings.