Adaptive average arterial pressure control by multi-agent on-policy reinforcement learning.

The current research introduces a model-free ultra-local model (MFULM) controller that utilizes the multi-agent on-policy reinforcement learning (MAOPRL) technique for remotely regulating blood pressure through precise drug dosing in a closed-loop system. Within the closed-loop system, there exists a MFULM controller, an observer, and an intelligent MAOPRL algorithm. Initially, a flexible MFULM controller is created to make adjustments to blood pressure and medication dosages.

Local corner smoothing based on deep learning for CNC machine tools.

Most of toolpaths for machining is composed of series of short linear segments (G01 command), which limits the feedrate and machining quality. To generate a smooth machining path, a new optimization strategy is proposed to optimize the toolpath at the curvature level. First, the three essential components of optimization are introduced, and the local corner smoothness is converted into an optimization problem.

Non-invasive blood glucose monitoring using PPG signals with various deep learning models and implementation using TinyML.

Accurate and continuous blood glucose monitoring is essential for effective diabetes management, yet traditional finger pricking methods are often inconvenient and painful. To address this issue, photoplethysmography (PPG) presents a promising non-invasive alternative for estimating blood glucose levels. In this study, we propose an innovative 1-second signal segmentation method and evaluate the performance of three advanced deep learning models using a novel dataset to estimate blood glucose levels from PPG signals.

Design of anomalous reflecting metasurface for communication systems.

This paper presents a novel design approach for an anomalous reflector metasurface for communication systems operating at 8 GHz band. The main contribution of this work is the development of a general analytical method that accurately calculates the electromagnetic response of realistic metasurfaces with periodic impedance profiles. The modulated surface impedance is achieved by incorporating appropriately sized conductive patches on a grounded dielectric substrate.

Perovskite retinomorphic image sensor for embodied intelligent vision.

Retinomorphic systems that can see, recognize, and respond to real-time environmental information will extend the complexity and range of tasks that an exoskeleton robot can perform to better assist physically disabled people. However, the lack of ultrasensitive, reconfigurable, and large-scale integratable retinomorphic devices and advanced edge-processing algorithms makes it difficult to realize retinomorphic hardware. Here, we report the retinomorphic hardware prototype with a 4096-pixel perovskite image sensor array as core module to endow embodied intelligent vision functionalities.

pH-Responsive Polyethylene Glycol Engagers for Enhanced Brain Delivery of PEGylated Nanomedicine to Treat Glioblastoma.

The blood-brain barrier (BBB) remains a major obstacle for effective delivery of therapeutics to treat central nervous system (CNS) disorders. Although transferrin receptor (TfR)-mediated transcytosis is widely employed for brain drug delivery, the inefficient release of therapeutic payload hinders their efficacy from crossing the BBB. Here, we developed a pH-responsive anti-polyethylene glycol (PEG) × anti-TfR bispecific antibody (pH-PEG engager) that can complex with PEGylated nanomedicine at physiological pH to trigger TfR-mediated transcytosis in the brain microvascular endothelial cells, while rapidly dissociating from PEGylated nanomedicine at acidic endosomes for efficient release of PEGylated nanomedicine to cross the BBB.

Optical biosensor arrays based on nanozymes for environmental monitoring and food safety detection: principles, design, and applications.

Typical biosensing platforms are based on the "lock-and-key" approach, providing high specificity and sensitivity for environmental and food safety monitoring. However, they are limited in their ability to detect multiple analytes simultaneously. With the use of pattern identification methods, biosensor arrays can detect faint fluctuations caused by multiple analytes with similar properties in complex systems.

Nanomembrane on Graphene: Delamination Dynamics and 3D Construction.

Freestanding nanomembranes fabricated by lift-off technology have been widely utilized in microelectromechanical systems, soft electronics, and microrobotics. However, a conventional chemical etching strategy to eliminate nanomembrane adhesion often restricts material choice and compromises quality. Herein, we propose a nanomembrane-on-graphene strategy that leverages the weak van der Waals adhesion on graphene to achieve scalable and controllable release and 3D construction of nanomembranes.

Rare Earth Complex-Based Functional Materials: From Molecular Design and Performance Regulation to Unique Applications.

ConspectusRare earth (RE) elements, due to their unique electronic structures, exhibit excellent optical, electrical, and magnetic properties and thus have found widespread applications in the fields of electronics, optics, and biomedicine. A significant advancement in the use of RE elements is the formation of RE complexes. RE complexes, created by the coordination of RE ions with organic ligands, not only offer high molecular design flexibility but also incorporate features such as a broad absorption band and efficient energy transfer of organic ligands.

Enhancing the performance of SSVEP-based BCIs by combining task-related component analysis and deep neural network.

Steady-State Visually Evoked Potential (SSVEP) signals can be decoded by either a traditional machine learning algorithm or a deep learning network. Combining the two methods is expected to enhance the performance of an SSVEP-based brain-computer interface (BCI) by exploiting their advantages. However, an efficient strategy for integrating the two methods has not yet been established.

Failure risk study of anchor bolts durability for loess slope under long-term hydro-thermal effect.

The long-term safety and durability of anchor systems are the focus of slope maintenance management and sustainable operation. This study presents the observed temperature, humidity, and anchor bolt stress at varying depths from four-year remote real-time monitoring of the selected loess highway cut-slope. The potential correlation between slope hydrothermal environment and anchor stress is analyzed.

Hybrid electromagnetic and moisture energy harvesting enabled by ionic diode films.

Wireless energy-responsive systems provide a foundational platform for powering and operating intelligent devices. However, current electronic systems relying on complex components limit their effective deployment in ambient environment and seamless integration of energy harvesting, storage, sensing, and communication. Here, we disclose a coupling effect of electromagnetic wave absorption and moist-enabled generation on carrier transportation and energy interaction regulated by ionic diode effect.

π-PrimeNovo: an accurate and efficient non-autoregressive deep learning model for de novo peptide sequencing.

Peptide sequencing via tandem mass spectrometry (MS/MS) is essential in proteomics. Unlike traditional database searches, deep learning excels at de novo peptide sequencing, even for peptides missing from existing databases. Current deep learning models often rely on autoregressive generation, which suffers from error accumulation and slow inference speeds.

Reflective metasurface for 5G & beyond Wireless communications.

This paper presents a scalable reflective metasurface design optimized for 5G and beyond (B5G) wireless communications, featuring a unique combination of passive metasurface elements. The proposed design emphasizes a less complex structural configuration, facilitating easy scalability and cost-effective fabrication. By implementing a single-layer structure, the metasurface enables straightforward integration with existing B5G infrastructure and demonstrates compatibility with emerging intelligent surface technologies, such as Reconfigurable Intelligent Surfaces (RIS).

Quantum machine learning for Lyapunov-stabilized computation offloading in next-generation MEC networks.

Quantum computing and machine learning convergence enable powerful new approaches for optimizing mobile edge computing (MEC) networks. This paper uses Lyapunov optimization theory to propose a novel quantum machine learning framework for stabilizing computation offloading in next-generation MEC systems. Our approach leverages hybrid quantum-classical neural networks to learn optimal offloading policies that maximize network performance while ensuring the stability of data queues, even under dynamic and unpredictable network conditions.

An information gap decision theory and improved gradient-based optimizer for robust optimization of renewable energy systems in distribution network.

In this paper, a robust fuzzy multi-objective framework is performed to optimize the dispersed and hybrid renewable photovoltaic-wind energy resources in a radial distribution network considering uncertainties of renewable generation and network demand. A novel multi-objective improved gradient-based optimizer (MOIGBO) enhanced with Rosenbrock's direct rotational technique to overcome premature convergence is proposed to determine the problem optimal decision variables. The deterministic optimization framework without uncertainty minimizes active energy loss, unmet customer energy, and renewable generation costs.

Remote epitaxy and exfoliation of vanadium dioxide via sub-nanometer thick amorphous interlayer.

The recently emerged remote epitaxy technique, utilizing 2D materials (mostly graphene) as interlayers between the epilayer and the substrate, enables the exfoliation of crystalline nanomembranes from the substrate, expanding the range of potential device applications. However, remote epitaxy has been so far applied to a limited range of material systems, owing to the need of stringent growth conditions to avoid graphene damaging, and has therefore remained challenging for the synthesis of oxide nanomembranes. Here, we demonstrate the remote epitaxial growth of an oxide nanomembrane (vanadium dioxide, VO) with a sub-nanometer thick amorphous interlayer, which can withstand potential sputtering-induced damage and oxidation.

Many-body van der Waals interactions in multilayer structures studied by atomic force microscopy.

Van der Waals interaction in multilayer structures was predicted to be of many-body character, almost in parallel with the establishment of Lifshitz theory. However, the diminishing interaction between layers separated by a finite-thickness intermediate layer prevents experimental verification of the many-body nature. Here we verify the substrate contribution at the adhesion between the atomic force microscopy tip and the supported graphene, by taking advantage of the atomic-scale proximity of two objects separated by graphene.

Game-theoretic planning for multiplayer defense task with online objective function parameter estimation.

This work investigates a game-theoretic path planning algorithm with online objective function parameter estimation for a multiplayer intrusion-defense game, where the defenders aim to prevent intruders from entering the protected area. At first, an intruder is assigned to each defender to perform a one-to-one interception by solving an integer optimization problem. Then, the intrusion-defense game is formulated in a receding horizon manner by designing the objective function and constraints for the defenders and intruders, respectively.

Distributed aggregative optimization with affine coupling constraints.

This paper investigates a distributed aggregative optimization problem subject to coupling affine inequality constraints, in which local objective functions depend not only on their own decision variables but also on an aggregation of all the agents' variables. The formulated problem encompasses numerous practical applications, such as commodity distribution, electric vehicle charging, and energy consumption control in power grids. Hence, there is a compelling need to explore a new neurodynamic approach to address this.

QTypeMix: Enhancing multi-agent cooperative strategies through heterogeneous and homogeneous value decomposition.

In multi-agent cooperative tasks, the presence of heterogeneous agents is familiar. Compared to cooperation among homogeneous agents, collaboration requires considering the best-suited sub-tasks for each agent. However, the operation of multi-agent systems often involves a large amount of complex interaction information, making it more challenging to learn heterogeneous strategies.

A multi-responsive 3D deformable soft actuator with tunable structural color enabled by a graphene/cholesteric liquid crystal elastomer composite.

Intelligent soft robots that integrate both structural color and controllable actuation ability have attracted substantial attention for constructing biomimetic systems, biomedical devices, and soft robotics. However, simultaneously endowing single-layer cholesteric liquid crystal elastomer (CLCE) soft actuators with reversible 3D deformability and vivid structural color changes is still challenging. Herein, a multi-responsive (force, heat and light) single-layer 3D deformable soft actuator with vivid structural color-changing ability is realized through the reduced graphene oxide (RGO) deposition-induced Janus structure of the CLCE using a precisely-controlled evaporation method.

Electrochemical Migration of Zincophilic Metals for Stress Mitigation and Uniform Zinc Deposition in Aqueous Zinc-Ion Batteries.

The propensity of zinc (Zn) to form irregular electrodeposits at liquid-solid interfaces emerges as a fundamental barrier to high-energy, rechargeable batteries that use zinc anodes. So far, tremendous efforts are devoted to tailoring interfaces, while atomic-scale reaction mechanisms and the related nanoscale strain at the electrochemical interface receive less attention. Here, the underlying atomic-scale reaction mechanisms and the associated nanoscale strain at the electrochemical alloy interface are investigate, using gold-zinc alloy protective layer as a model system.

Automatic Segmentation of Vestibular Schwannoma From MRI Using Two Cascaded Deep Learning Networks.

Objective: Automatic segmentation and detection of vestibular schwannoma (VS) in MRI by deep learning is an upcoming topic. However, deep learning faces generalization challenges due to tumor variability even though measurements and segmentation of VS are essential for growth monitoring and treatment planning. Therefore, we introduce a novel model combining two Convolutional Neural Network (CNN) models for the detection of VS by deep learning aiming to improve performance of automatic segmentation.