Fault diagnosis of rolling bearings is a critical task, and in previous research, convolutional neural networks (CNN) have been used to process vibration signals perform fault diagnosis. However, traditional CNN models certain limitations terms accuracy. To improve accuracy, we propose method that combines the Gramian angular difference field (GADF) with residual (ResNet) embeds frequency channel attention module (Fca) ResNet diagnose bearing fault. Firstly, GADF convert into RGB...

Predicting the remaining useful life (RUL) of rolling bearings is crucial for industrial machinery maintenance, non-destructive testing and evaluation (NDT). To address challenges posed by noise interference redundant information, this paper proposes a novel approach utilising residual attention networks multi-scale feature extraction. The method enhances extraction combining shallow deep convolutional layers while employing bidirectional LSTM to capture both short-term long-term...

Abstract This study proposes a power management strategy (PMS) that can efficiently and accurately address the nonlinear dynamics of ship systems for all‐electric ships (AESs). The design PMS be considered naturally in model predictive control framework. However, error caused by mismatch is challenge physical system implementation. proposed based on complex system‐level ensures minimum mismatch. In addition, an efficient optimization algorithm to apply PMS. method verified using real‐time...

Ocean currents pose a significant challenge in the path planning of autonomous underwater vehicles (AUVs), with conventional path-planning algorithms often failing to effectively counter these influences. In response this challenge, we propose algorithm that can consider influences and constraints ocean currents, which leverages strengths two widely employed algorithms, A* genetic (GA), account for on planned paths. Specifically, it enhances initial population generation, formulates fitness...

This study presents a novel image-based visual servoing fault-tolerant control strategy aimed at ensuring the successful completion of tasks despite presence robotic arm actuator faults. Initially, depth-independent model is established to mitigate effects inaccurate camera parameters and missing depth information on system. Additionally, dynamic constructed, which simultaneously considers both multiplicative additive Subsequently, uncertainties, unknown disturbances, coupled faults are...

Summary Rolling bearings often function under complex and non‐stationary conditions, where significant noise interference complicates fault diagnosis by obscuring characteristics. This paper presents an innovative technique using graph convolutional networks (GCN) to address these challenges. Vibration signals are first transformed into the frequency domain through fast Fourier transform (FFT), creating a detailed nodes edges encapsulate signals. The GCN method then extracts node features...

For the problem of large prediction error which is caused by some kind method in constant weight combination forecasting model predicted result mutate, this paper proposes an adaptive variable model. And applied it to ship roll motion prediction. This combined Kalman filter with Volterra series model, recursive least squares identification adopted define weights, established Data real sail test modeling The shows that combining models are more accurate than single and can get better results...

To deal with the problem how to control yaw and roll using rudder together for nonlinear ship stabilization motion, a method based on feedback linearization closed-loop gain shaping algorithm is proposed in this paper. Firstly, used design controller, then get linear law. When motion controlled by rudder, loss of steering gear can be increase, problem, fuzzy applied weighted coefficients optimization. Finally, simulation results show that controller has better adaptation yaw, reduced...

With the development of science and technology, rapid social economy, motor as a new type transmission equipment, in production life people occupies pivotal position. Under computer electronic manufacturing equipment is becoming larger, faster, more continuous, automated. This has resulted complex, expensive, accident-damaging, high-impact for electric motors; even routine maintenance requires significant costs. If fault occurs, it will cause serious damage to entire can have major impact on...

With the in-depth penetration of renewable energy in shipboard power system, uncertainty its output and variability sea conditions have brought severe challenges to control integrated system. In order provide additional accurate signals system eliminate influence uncertain factors, this study proposed an adaptive kernel based online sequential extreme learning machine accurately predict electric fluctuation online. Three factors are introduced, which function scale adaptively ensure accuracy...

According to the problem of small samples and nonlinear feature in fault diagnosis marine diesel engine, comprehensively using methods grey relational analysis kernel fuzzy c-means clustering, a method solving engine is proposed. Firstly, clustering was made on historical dataset. Secondly, preliminary testing by cmeans separately. Finally, final results were got linear weighting matrix membership matrix. The MAN B&W 10L90MC show that this can improve accuracy engine.

Traditional graph neural networks often encounter limitations in fault diagnosis due to insufficient feature extraction at a single scale, particularly complex operational scenarios. To overcome these challenges, we introduce an innovative multi-scale Transformer framework for rolling bearing diagnosis. This incorporates distinctive aggregation mechanism, along with centrality and spatial encoding nodes, enhance structural insights. Leveraging multi-head self-attention, our approach...

According to the characteristics of huge ship motion forced by waves. It is introduced that a real-time modeling and prediction method attitude utilizing automation regressive moving average (ARMA) model. ARMA model AR for wave force bow halls were given. And it discussed in details fast order selection with improved corner condition RLS algorithm online parameter identification. was resolved on-line estimate problem model, calculation time shorten. The suitable under uneasy motion. In...

By using the plane-wave-expansion method, band structure of three-dimension phononic crystals was calculated, in which cuboid scatterers were arranged a host with face-centered-cubic (FCC) structure. The influences few factors such as component materials, filling fraction and ratio (RHL) scatterer's height to its length on band-gaps investigated. It is found that solid FCC structure, optimum case obtain embed high-velocity high-density low-velocity low-density host. maximum value band-gap...

In this paper, an adaptive neural-net control system, in which learning is performed a loop totally independent from the loop, proposed for problem of ship roll stabilization. The modeling and controller are adjusted continuously order to deal with changes dynamic properties caused by disturbances. Based experimental data tank, disturbance model sea wave presented. A recurrent neural network used approaching dynamics ship, real time algorithm described train forward model. This paper...

This paper investigates the problem of full-order and reduced-order fault detection filter (FDF) design under unified linear matrix inequality (LMI) conditions for a class continuous-time singular Markovian jump systems (CTSMJSs) with time-varying delays polytopic uncertain transition rates. By constructing new Lyapunov function, sufficient are firstly provided model error augmented system such that is stochastically admissible an <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Smart grid is regarded as an evolutionary regime of existing power grids. It integrates artificial intelligence and communication technologies to fundamentally improve the efficiency reliability systems. One serious challenge for smart its vulnerability cyber threats. In event a attack, data may be missing; subsequently, load forecast planning that rely on these cannot processed by generation centers. To address this issue, paper proposes transfer learning-based framework scheduling less...

A system, which can capture the real time image of ablative surface under strong stray light environments caused by plasma arc, has been presented. The system nearly one year service in some national key laboratory. It shown that effectively shield influence jet and with temperature over 4,000 K meanwhile, high quality (with resolution 0.01 mm) 3,000 K) be obtained. acquisition provided present paper not only applied to on-line monitoring, but also used other areas require real-time...

This work is concerned with the problem of full-order and reduced-order fault detection filters (FDFs) design in a convex optimization frame for continuous-time singular Markov jump systems (CTSMJSs) complexity transition rates (TRs). A novel Lyapunov function construct approach utilized to cope stochastic admissibility CTSMJSs TRs. In order obtain effective FDFs, we decoupled inequality using presupposed matrix. Owing use theory decoupling method based on polyhedron technique, some...