An echo state network with improved topology (IESN) is proposed for accurate and efficient time series prediction. In this approach, a tighter bound of the property related to Lipshitz constant reservoir activation function maximum structured singular value firstly researched run model at edge chaos. A smooth composite then designed enhance ESN. The exact solved by computing function. Finally, decoupling matrix eigenvalues distributing uniformly in complex plane built as abundant dynamic...

Fault diagnosis can improve the safety and reliability of diesel engines. An end-to-end method based on a multi-attention convolutional neural network (MACNN) is proposed for accurate efficient engine fault diagnosis. By optimizing arrangement kernel size channel spatial attention modules, feature extraction capability improved, an improved block module (ICBAM) obtained. Vibration signal features are acquired using model alternating between (CNN) ICBAM. The map recombined to reconstruct...

Fatigue crack defects in metallic materials significantly reduce the remaining useful life (RUL) of parts. However, much existing research has focused on identifying single-millimeter-scale cracks using individual nonlinear ultrasonic responses. The identification subtle parameters from complex responses micro-crack groups remains a significant challenge field nondestructive testing. We propose novel multi-harmonic response fusion method integrated with deep learning (DL) model to identify...

The evaluation and fault diagnosis of a diesel engine’s health conditions without disassembly are very important for engine safe operation. Currently, the research on has focused time domain or frequency processing vibration signals. However, early signals mostly weak energy signals, information cannot be completely extracted by analysis. Thus, in this article, novel method valve clearance using improved variational mode decomposition (VMD) bispectrum algorithm is proposed. First,...

As one kind of current main power sources, internal combustion engines require high-reliability to ensure mechanical systems working well in normal operation. This paper studies vibration signals order detect multiple types faults by a single channel signal. First, for the decomposition level variational mode (VMD) which needs be chosen non-automatically, this analyzes features various and optimizes iteration initial values center frequency so as reduce adverse effect inappropriate level....

Engine fault detection is critical to enhancing the reliability of modern equipment. However, it challenging obtain a large number high-quality labeled data for engines, which not conducive improving training accuracy deep learning methods. Therefore, this article proposes method combining adaptive recursive variational mode decomposition (ARVMD) and component energy distribution spectrum (CEDS). The paper first introduces into VMD. Then, dynamically selected according power spectral density...

Transfer learning (TL) is a powerful approach that enhances the generalizability of cross-domain fault diagnosis. However, challenge acquiring high-quality mechanical signals limits its application. This article introduces extreme class imbalance problem in diagnosis, restricting label space target domain while relaxing restrictions unsupervised learning. The study proposes novel generative TL method called fast sparse neural style, which employs representation to capture domain-invariant...

This study experimentally and numerically investigated the nonlinear behavior of resonant bulk waves generated by two-way collinear mixing method in 5052 aluminum alloy with micro-crack damage. When primary longitudinal transverse mixed damage region, numerical experimental results both verified generation if condition ωL/ωT=2κ/(κ−1) was satisfied. Meanwhile, we found that acoustic nonlinearity parameter (ANP) increases monotonously density, size frequency friction coefficient surfaces....

The linear motor feeding system is a typical electromechanical coupling system. Conventional characteristic analyses of often overlook the influence flexible deformation in critical components Moreover, when employing genetic algorithms to optimize servo PID control parameters, slow convergence, nonconvergence, or premature convergence problems may arise. To address these issues, this paper proposes new performance optimization method for uses combination “multi-body theory + finite element”...

Piston aeroengine is a general power device for small aircraft. However, the engine system complex and subjected to strong noise interference in addition changing operating conditions, which make it difficult diagnose faults of engine. This article proposes novel intelligent diagnostic model piston collaboration with acoustic emission (AE) signal rough set theory. First, wavelet packet transform used time-frequency analysis AE signals; then based on theory, select sensitive components...

Engine fault detection is conducive to improving equipment reliability and reducing maintenance costs. In practical scenarios, high-quality data difficult obtain. Usually, only single-sensor available. This paper proposes a method combining Variational Mode Decomposition (VMD) Random Forest (RF). At first, the spectral energy distribution obtained by decomposing statistic engine of multiple working conditions. Based on distribution, overall optimal mode number was identified, quadratic...

Ignition advance angle is one of the important factors affecting performance engine, when it occurs abnormally will make engine power and economy worse, even cause serious damage to engine. Therefore, very necessary recognize abnormal ignition However, system closed has a complex structure, which makes traditional diagnostic methods difficult. This paper proposes an intelligent identification method based on acoustic emission (AE) signals, collects AE signals from surface divides their...

Instantaneous angular speed (IAS) measurement using zebra tape has been presented as an advantageous tool for rotating machinery surveillance. Unfortunately, both butt joint and mounting imperfection of the result in stripes with different widths, which cause outliers appear acquired signal. This paper proposed a solution to accurately identify define displacement values represented by condition periodic fluctuation speed. The procedure is based on K-means clustering method Fourier series...

Abstract The reliability of engines, particularly aero has become increasingly important in recent years. Accurate fault diagnosis can prevent accidents and minimize property damage. Deep neural network methods (DNNs) are commonly used for diagnosis, but their performance relies heavily on large amounts high-quality training data. Unfortunately, obtaining engine data is challenging practice. To address this problem, paper proposes an improved auxiliary classifier generative adversarial...

Style transfer is a heated computer vision topic that has been broadly mentioned in both academic research and practical applications, for example, the style technique behind various filters daily used image processing tools. However, it had major drawback this field its training set required strictly paired images, until introduction of CycleGAN, combination Generating Adversarial Networks (GAN), which also one most popular technologies machine learning since advent. CycleGAN an proposed by...