The global automotive industry is in the phase where Internal Combustion vehicles are in decline and witnessing a shift towards sustainable development. The major parameter of a successful EV is an efficient battery p...
详细信息
Halftone classification is a primary requisite for the perfect reconstruction of binary patterns during inverse halftone process. Majority of the halftone classification techniques are either limited to error diffused...
详细信息
Along with the flourishing of brain-computer interface technology,the brain-to-brain information transmission between different organisms has received high attention in recent ***,specific information transmission mod...
详细信息
Along with the flourishing of brain-computer interface technology,the brain-to-brain information transmission between different organisms has received high attention in recent ***,specific information transmission mode and implementation technology need to be further *** this paper,we constructed a brain-to-brain information transmission system between pigeons based on the neural information decoding and electrical stimulation encoding *** system consists of three parts:(1)the“perception pigeon”learns to distinguish different visual stimuli with two discrepant frequencies,(2)the computer decodes the stimuli based on the neural signals recorded from the“perception pigeon”through a frequency identification algorithm(neural information decoding)and encodes them into different kinds of electrical pulses,(3)the“action pigeon”receives the Intracortical Microstimulation(ICMS)and executes corresponding key-pecking actions through discriminative learning(electrical stimulation encoding).The experimental results show that our brain-to-brain system achieves information transmission from perception to action between two pigeons with the average accuracy of about 72%.Our study verifies the feasibility of information transmission between inter-brain based on neural information decoding and ICMS encoding,providing important technical methods and experimental program references for the development of brain-to-brain communication technology.
This article introduces a unified structure for broadband and multi-band graphene-based absorbers. The absorption process relies on multiple circular graphene disks positioned on a glass surface. The absorber’s funct...
详细信息
To address issues such as poor initial population diversity, low stability and local convergence accuracy, and easy local optima in the traditional Multi-Objective Artificial Hummingbird Algorithm (MOAHA), an Improved...
详细信息
To address issues such as poor initial population diversity, low stability and local convergence accuracy, and easy local optima in the traditional Multi-Objective Artificial Hummingbird Algorithm (MOAHA), an Improved MOAHA (IMOAHA) was proposed. The improvements involve Tent mapping based on random variables to initialize the population, a logarithmic decrease strategy for inertia weight to balance search capability, and the improved search operators in the territory foraging phase to enhance the ability to escape from local optima and increase convergence accuracy. The effectiveness of IMOAHA was verified through Matlab/Simulink. The results demonstrate that IMOAHA exhibits superior convergence, diversity, uniformity, and coverage of solutions across 6 test functions, outperforming 4 comparative algorithms. A Wilcoxon rank-sum test further confirmed its exceptional performance. To assess IMOAHA’s ability to solve engineering problems, an optimization model for a multi-track, multi-train urban rail traction power supply system with Supercapacitor Energy Storage Systems (SCESSs) was established, and IMOAHA was successfully applied to solving the capacity allocation problem of SCESSs, demonstrating that it is an effective tool for solving complex Multi-Objective Optimization Problems (MOOPs) in engineering domains.
Graph neural networks (GNNs) have gained increasing popularity, while usually suffering from unaffordable computations for real-world large-scale applications. Hence, pruning GNNs is of great need but largely unexplor...
详细信息
Graph neural networks (GNNs) have gained increasing popularity, while usually suffering from unaffordable computations for real-world large-scale applications. Hence, pruning GNNs is of great need but largely unexplored. The recent work Unified GNN Sparsification (UGS) studies lottery ticket learning for GNNs, aiming to find a subset of model parameters and graph structures that can best maintain the GNN performance. However, it is tailed for the transductive setting, failing to generalize to unseen graphs, which are common in inductive tasks like graph classification. In this work, we propose a simple and effective learning paradigm, Inductive Co-Pruning of GNNs (ICPG), to endow graph lottery tickets with inductive pruning capacity. To prune the input graphs, we design a predictive model to generate importance scores for each edge based on the input. To prune the model parameters, it views the weight’s magnitude as their importance scores. Then we design an iterative co-pruning strategy to trim the graph edges and GNN weights based on their importance scores. Although it might be strikingly simple, ICPG surpasses the existing pruning method and can be universally applicable in both inductive and transductive learning settings. On 10 graph-classification and two node-classification benchmarks, ICPG achieves the same performance level with 14.26%–43.12% sparsity for graphs and 48.80%–91.41% sparsity for the GNN model.
This article is concerned with the multilevel secure warning problem for lithium-ion battery systems (LBSs) subject to multi-fault scenarios. First, the long short-term memory (LSTM) network is employed to design a se...
详细信息
This article presents an initial solution based selective harmonic elimination (SHE) method for multilevel inverter (MLI) that aims to solve SHE problem with high accuracy while significantly reducing the number of it...
详细信息
The traditional plasma etching process for defining micro-LED pixels could lead to significant sidewall *** near sidewall regions act as non-radiative recombination centers and paths for current leakage,significantly ...
详细信息
The traditional plasma etching process for defining micro-LED pixels could lead to significant sidewall *** near sidewall regions act as non-radiative recombination centers and paths for current leakage,significantly deteriorating device *** this study,we demonstrated a novel selective thermal oxidation(STO)method that allowed pixel definition without undergoing plasma damage and subsequent dielectric *** annealing in ambient air oxidized and reshaped the LED structure,such as p-layers and InGaN/GaN multiple quantum ***,the pixel areas beneath the pre-deposited SiO_(2)layer were selectively and effectively *** was demonstrated that prolonged thermal annealing time enhanced the insulating properties of the oxide,significantly reducing LED leakage ***,applying a thicker SiO_(2)protective layer minimized device resistance and boosted device efficiency *** the STO method,InGaN green micro-LED arrays with 50-,30-,and 10-μm pixel sizes were manufactured and *** results indicated that after 4 h of air annealing and with a 3.5-μm SiO_(2)protective layer,the 10-μm pixel array exhibited leakage currents density 1.2×10^(-6)A/cm^(2)at-10 V voltage and a peak on-wafer external quantum efficiency of~6.48%.This work suggests that the STO method could become an effective approach for future micro-LED manufacturing to mitigate adverse LED efficiency size effects due to the plasma etching and improve device ***-LEDs fabricated through the STO method can be applied to micro-displays,visible light communication,and optical interconnect-based *** planar pixel geometry will provide more possibilities for the monolithic integration of driving circuits with ***,the STO method is not limited to micro-LED fabrication and can be extended to design other III-nitride devices,such as photodetectors,laser diodes,high-electron-mobility transistors
Potential malicious cyber-attacks to power systems which are connected to a wide range of stakeholders from the top to tail will impose significant societal risks and *** timely detection and defense are of crucial im...
详细信息
Potential malicious cyber-attacks to power systems which are connected to a wide range of stakeholders from the top to tail will impose significant societal risks and *** timely detection and defense are of crucial importance for safe and reliable operation of cyber-physical power systems(CPPSs).This paper presents a comprehensive review of some of the latest attack detection and defense ***,the vulnerabilities brought by some new information and communication technologies(ICTs)are analyzed,and their impacts on the security of CPPSs are *** malicious cyber-attacks on cyber and physical layers are then analyzed within CPPSs framework,and their features and negative impacts are ***,two current mainstream attack detection methods including state estimation based and machine learning based methods are analyzed,and their benefits and drawbacks are ***,two current mainstream attack defense methods including active defense and passive defense methods are comprehensively ***,the trends and challenges in attack detection and defense strategies in CPPSs are provided.
暂无评论