Underwater sensor networks are facing a great challenge in designing routing protocol with longer network lifetime and higher packet delivery rate under the complex underwater environment. In this paper, we propose an energy-aware void-avoidable (EAVARP). EAVARP includes layering phase data collection phase. During phase, concentric shells built around sink node, nodes distributed on different shells. Sink node performs hierarchical tasks periodically to ensure validity real-time of...

In the computer vision field, underwater object detection has been a challenging task. Due to attenuation of light in medium and scattering by suspended particles water, optical images often face problems color distortion target feature blurring, which greatly affect accuracy detection. Although deep learning-based algorithms have achieved state-of-the-art results field detection, most them cannot be applied practice because limited computing capacity low-power processor embedded unmanned...

Abstract Antifouling technologies have attracted considerable attention in recent years, as numerous fouling phenomena pertaining to inorganic, organic, bio-, and composite foulants substantially affect daily life. Poly(dimethyl siloxane) (PDMS) has several practical applications; however, it possesses limited resistance or biofoulants such proteins bacteria. Among the antifouling strategies reported thus far, induced by surface wettability (AFISW) is an exceptional strategy with potential....

Abstract Core@shell catalyst composited of dual aluminosilicate zeolite can effectively regulate the distribution acid sites to control hydrocarbon conversion process for stable formation target product. However, diffusion restriction reduces accessibility inner active and affects synergy between core shell. Herein, hollow ZSM‐5 nanoreactor with inverse aluminum double shells are prepared employed methanol aromatization. It is demonstrated that intershell cavity alleviated steric hindrance...

Accurate and robust simultaneous localization mapping (SLAM) systems are crucial for autonomous underwater vehicles (AUVs) to perform missions in unknown environments. However, directly applying deep learning-based SLAM methods environments poses challenges due weak textures, image degradation, the inability accurately annotate keypoints. In this paper, a deep-learning visual system is proposed. First, feature generator named UWNet designed address texture degradation problems extract more...

The reverse osmosis method has developed extremely rapidly in recent years and become the most competitive seawater desalination technology world, it been widely used all aspects. Large-scale plants cannot provide fresh water resources areas with insufficient limited space. Therefore, this paper proposes a research plan for small device based on osmosis, which is mainly suitable handling emergencies, disaster relief, desert outdoor activities other needs timely freshwater resources. It...

The prediction of tidal fields is crucial in coastal and marine hydrodynamic analyses, particularly complex environments, as it plays an essential role disaster warning fisheries management. However, monitoring the entire field impractical, harmonic analysis numerical simulation methods continue to face challenges accuracy efficiency for large-scale predictions. To address these issues, this paper proposes a method based on Long Short-Term Memory (LSTM) networks. A physics-based model...

ABSTRACT Underwater object detection serves as a crucial means for autonomous underwater vehicles (AUVs) to gain awareness of their surroundings. Currently, AUVs predominantly depend on optical cameras or sonar sensing techniques furnish vital information sources subsequent tasks such rescue and mining exploration. However, the influence light attenuation significant background noise often leads failure either acoustic sensor. Consequently, traditional single‐modal network, which relies...

Submarine pipelines, as arteries for offshore oil and gas transportation, play a particularly important role in the exploitation of resources. Since world’s first submarine pipelines were laid Gulf Mexico, numerous failures have been caused by pipeline free-spanning. This paper provides review causes treatment measures free span pipeline. Various factors cause including wave flow scouring, fluctuations seabed topography, residual stress or thermal human activities. The scour current is main...

The environmental adaptability of autonomous underwater vehicles is always a problem for its path planning. Although reinforcement learning can improve the adaptability, slow convergence caused by multi-behavior coupling, so it difficult vehicle to avoid moving obstacles. This article proposes critic algorithm applied planning overcome problems associated with oscillating amplitudes and low efficiency in early stages training which are common traditional actor–critic algorithms. Behavior...

As multichannel medium access control (MAC) protocol has higher network performance, it becomes an important issue in underwater sensor networks (UWSNs). The design of MAC faces many challenges caused by long-delay, low bandwidth and triple hidden terminal problems. We propose a Distributed Receiver-oriented Adaptive Multichannel (DRAMAC) for UWSNs this paper. DRAMAC contains two key schemes: channel negotiation process based on cooperative correction (NPCC) receiver-oriented dynamic...

Abstract Based on the characteristics of high-frequency swing during fast swimming fish, this paper designs a bionic fish-driven joint based electromagnetic drive to achieve swing. Aiming at characteristic parameters control, Fourier transform is used separate and then compared driving accuracy joints in open-loop closed-loop. The comparison results show that closed-loop control performed after transform. Under same conditions, method can improve accuracy. Then fish robot composed three...

Reinforcement learning (RL) is known for its efficiency and practicality in single-agent planning, but it faces numerous challenges when applied to multi-agent scenarios. In this paper, a Super Sampling Info-GAN (SSIG) algorithm based on Generative Adversarial Networks (GANs) proposed address the problem of state instability Multi-Agent Learning (MARL). The SSIG model allows pair GAN networks analyze previous dynamic system predict future consecutive pairs. A (MAS) can deduce complete all...

In order to clean underwater silt in artificially constructed rivers, lakes, and fish ponds, for which no suitable tool exists, a has been developed that imitates the structure movement of tortoise’s legs, designs four-legged dredging robot can adapt complex environment. The article uses transformation matrix analyze kinematics robot, determines sequence outriggers according principle stability, analyzes characteristics three gait modes. Then, we combined control function foot trajectory...

Visual perception technology is of great significance for underwater robots to carry out seabed investigation and mariculture activities. Due the complex environment, it often necessary enhance image when detecting targets by optical sensors. Most traditional methods involve enhancement then target detection. However, this method greatly increases timeliness in practical application. To solve problem, we propose a feature-enhanced detection network, Global-Net (G-Net), which combines with...