RepDNet:A re-parameterization despeckling network for autonomous underwater side-scan sonar imaging with prior-knowledge customized convolution

Side-scan sonar(SSS)is now a prevalent instrument for large-scale seafloor topography measurements,deployable on an autonomous underwater vehicle(AUV)to execute fully automated underwater acoustic scanning imaging along a predetermined trajectory.However,SSS images often suffer from speckle noise caused by mutual interference between echoes,and limited AUV computational resources further hinder noise suppression.Existing approaches for SSS image processing and speckle noise reduction rely heavily on complex network structures and fail to combine the benefits of deep learning and domain knowledge.To address the problem,Rep DNet,a novel and effective despeckling convolutional neural network is proposed.Rep DNet introduces two re-parameterized blocks:the Pixel Smoothing Block(PSB)and Edge Enhancement Block(EEB),preserving edge information while attenuating speckle noise.During training,PSB and EEB manifest as double-layered multi-branch structures,integrating first-order and secondorder derivatives and smoothing functions.During inference,the branches are re-parameterized into a 3×3 convolution,enabling efficient inference without sacrificing accuracy.Rep DNet comprises three computational operations:3×3 convolution,element-wise summation and Rectified Linear Unit activation.Evaluations on benchmark datasets,a real SSS dataset and Data collected at Lake Mulan aestablish Rep DNet as a well-balanced network,meeting the AUV computational constraints in terms of performance and latency.

Experimental and Numerical Analysis of Particle Migration and Patterning Behavior in a Gravel Pack

Due to its long lifespan and high sand-removal efficiency,gravel packing is one of the most applied sand control methods during the recovery of reservoirs with sanding problems.The blockage and retention of injected sand in a gravel pack is a complex process affected by multiple mechanisms.The majority of existing studies based on the phenomenological deep bed filtration(DBF)theory focused on the gravel pack’s overall permeability damage and failed to obtain the inner-pore particle distribution pattern.In this work,experiments and simulations were carried out to reveal the particle distribution in a gravel pack during flooding.In particular,through real-time monitoring of particle migration,the penetration depth and distribution pattern of invaded particles with different gravel-sand particle ratios,fluid viscosities and injection rates could be determined.By simplifying each unit bed element(UBE)into a pore-throat structure with four tunnels(two horizontals for discharge and two verticals for sedimentation),a new network simulation method,which combines deep bed filtration with a particle trajectory model,was implemented.Cross comparison of experimental and numerical results demonstrates the validity and accuracy of the model.

Dynamical evolution of cross phase of edge fluctuations and transport bifurcation

The dynamical evolution of edge turbulence during a transport bifurcation is explored using a flux-driven nonlinear fluid model with a geometry relevant to the plasma edge region.The simulations show that the self-generated mean shear flows can dramatically modify the phase angle between turbulent fluctuations.The changes in phase differences and amplitudes of edge fluctuations give rise to the modifications of turbulent edge transport.The statistical properties of flux and fluctuations are also investigated before and after edge shear flow generation.

Photogated proton conductivity of ZIF-8 membranes co-modified with graphene quantum dots and polystyrene sulfonate

Smart proton conductive metal-organic framework(MOF) membranes with dynamic remote control over proton conduction show high potential for use in advanced applications, such as sensors and bioprocesses. Here, we report a photoswitchable proton conductive ZIF-8 membrane by coencapsulating polystyrene sulfonate and graphene quantum dots into a ZIF-8 matrix(GQDs-PSS@ZIF-8) via a solidconfined conversion process. The proton conductivity of the GQDs-PSS@ZIF-8 membrane is 6.3 times higher than that of pristine ZIF-8 and can be reversibly switched by light due to photoluminescence quenching and the photothermal conversion effect, which converts light into heat. The local increase in temperature allows water molecules to escape from the porous channels, which cuts off the proton transport pathways and results in a decrease in proton conductivity. The proton conductivity is restored when the light is off owing to regaining water molecules, which act as proton carriers, from the surroundings. The GQDs-PSS@ZIF-8 membrane responds efficiently to light and exhibits an ON/OFF ratio of 12.8. This photogated proton conduction in MOFs has potential for the development and application of MOF-based protonic solids in advanced photoelectric devices.

Highly conductive and transparent metal-organic frameworks thin film

Energy crisis has aroused global concern and needs to be solved immediately. A series of serious environmental problems caused by the unlimited use of trad让ional fossil fuels like petroleum and charcoal, such as rising sea levels and melting glaciers, have dangerously threatened human destiny [ 1 ]. It is an effective choice to look for new energy sources to replace the lim让ed fossil fuel reserves. Among these new energy sources, solar energy stands out and attracts tremendous attention because of its environmental friendliness, stability, renewability, and universality [2-5]. Transparent and conductive electrode is the vital part in photoelectric conversion devices [6]. Indium tin oxide (ITO) is the most popular transparent and conductive electrode material at present [7]. However, the limited reserves of indium, a component in ITO has resulted in the high price of ITO electrode, restricting its wide application.