High-speed Processing Research Articles

Polar Vortices in Relaxor Ferroelectric Ceramics for High-Efficiency Capacitive Energy Storage.

Polar vortices are predominantly observed within the confined ferroelectric films and the ferroelectric/paraelectric superlattices. This raises the intriguing question of whether polar vortices can form within relaxor ferroelectric ceramics and subsequently contribute to their energy storage performances. Here, we incorporate 10 mol % CaSnO3 into the 0.7NaNbO3-0.3Sr0.7Bi0.2TiO3 matrix, yielding a coexistence of phases: 48.8% orthorhombic P21/ma, 49.1% tetragonal P4bm, and 2.1% tetragonal P42/mnm SnO2, which is confirmed by the combination of X-ray diffraction and transmission electron microscopy. The ceramic features a pronounced core-shell structure with the shell region rich in stripe nanoscale domains of the P21/ma phase and the core region consisting of polar nanoregions deficient in the P21/ma phase, forming polar vortices. Consequently, the ceramic achieves an impressive recoverable energy storage density of 6.83 J cm-3 and an exceptional efficiency of 95.7% at a high breakdown strength of 750 kV cm-1, along with superior stability in frequency, temperature, and cycling. These results not only offer a viable approach for developing high-performance energy storage ceramics through the controlled formation of polar vortices but also offer the potential for direct electric-field control of polar vortices for high-speed data processing and storage.

Experimental study on spatiotemporal distribution characteristics of different cavitation stages based on the flow deviation method

Cavitation may quickly damage the surfaces of the valve core and valve seat, causing noise, and vibration problems. Different cavitation stages will affect the regulating valve's flow characteristics to different degrees. Flow rate is one of the basic parameters in a hydraulic system, which is innovatively used to evaluate the cavitation flow. By analyzing the deviation ratio K between actual and theoretical flow rate, cavitation flows are divided into four stages, and the dynamic behavior in different stages is discussed using a high-speed camera and image processing technology. When K is zero, it is defined as the no-cavitation stage. A slight increase in K is the incipient cavitation stage, whose K is within 2%. A rapid decrease in K is the critical cavitation stage, while a significant increase in K is the severe cavitation stage. In the incipient cavitation stage, bubbles are adhered to the valve orifice forming attached cavities. In the critical cavitation stage, attached cavities develop along the throat wall, with some bubbles detaching to form free cavities. This process is accompanied by high-frequency shedding and collapses, with a dominant frequency of 4266 Hz. In the severe cavitation stage, larger attached cavities are located at the throat, whose front edge of cavitation will fracture into large-scale cavitation clouds. Furthermore, this study proposed the cavitation intensity to elucidate the spatial distribution and density of cavitation flow, the cavitation change rate to highlight the disturbances and nonuniformities caused by cavitation bubbles, and the cavitation coefficient to evaluate the severity of cavitation. Additionally, there is a strong correlation between the growth rate of cavitation length L and the growth rate of flow rate Q in the regulating valve. Both the growth rate of L and Q increase in the critical cavitation stage and decrease in the severe cavitation stage.

Real-Time Interference Mitigation for Reliable Target Detection with FMCW Radar in Interference Environments

Frequency-modulated continuous-wave (FMCW) millimeter-wave (mmWave) radar systems are increasingly utilized in environmental sensing due to their high range resolution and robust sensing ability in severe weather environments. However, mutual interference among radar systems significantly degrades the target detection capability. Recent advancements in interference mitigation utilizing deep learning (DL) approaches have demonstrated promising results. DL-based approaches typically have high computational costs, which makes them unsuitable for real-time applications with strict latency requirements and limited computing resources. In this paper, we propose an efficient solution for real-time radar interference mitigation. A lightweight transformer, which is smaller and faster than the baseline transformer, is designed to reduce interference. The integration of linear attention mechanisms with depthwise separable convolutions significantly reduces the network’s computational complexity while maintaining a comparable performance. In addition, a two-stage knowledge distillation (KD) process is deployed to compress the network and enhance its efficiency. The staged distillation approach alleviates the training difficulties associated with substantial differences between the teacher and student networks. Both simulated and real-world experiments demonstrate that the proposed method outperforms the state-of-the-art methods while achieving high processing speeds.

Intuitionistic Fuzzy Set Guided Fast Fusion Transformer for Multi-Polarized Petrographic Image of Rock Thin Sections

The fusion of multi-polarized petrographic images of rock thin sections involves the fusion of feature information from microscopic images of rock thin sections illuminated under both plane-polarized and orthogonal-polarized light. During the fusion process of rock thin section images, the inherent high resolution and abundant feature information of the images pose substantial challenges in terms of computational complexity when dealing with massive datasets. In engineering applications, to ensure the quality of image fusion while meeting the practical requirements for high-speed processing, this paper proposes a novel fast fusion Transformer. The model leverages a soft matching algorithm based on intuitionistic fuzzy sets to merge redundant tokens, effectively mitigating the negative effects of asymmetric dependencies between tokens. The newly generated artificial tokens serve as brokers for the Query (Q), forming a novel lightweight fusion strategy. Both subjective visual observations and quantitative analyses demonstrate that the Transformer proposed in this paper is comparable to existing fusion methods in terms of performance while achieving a notable enhancement in its inference efficiency. This is made possible by the attention paradigm, which is equivalent to a generalized form of linear attention, and the newly designed loss function. The model has been experimented on with multiple datasets of different rock types and has exhibited robust generalization capabilities. It provides potential for future research in diverse geological conditions and broader application scenarios.

Compact photonic device based on chalcogenide glass loaded lithium niobate on insulator

We propose a hybrid photonic platform based on chalcogenide glass (GeSbSe) integrated with lithium niobate on insulator (LNOI), providing enhanced performance and compactness for integrated photonics. Key photonic components, including grating couplers (GCs), micro-ring resonators, multimode interference (MMI), and the Mach–Zehnder interferometer (MZI), are designed and fabricated on this platform. The micro-ring resonator achieves an intrinsic quality factor of 72,000 and a propagation loss of ∼3 dB/cm. The multimode interference couplers demonstrate minimal excess loss, while the MZI exhibits an extinction ratio exceeding 30 dB over a bandwidth of 50 nm. This hybrid platform, leveraging the unique optical properties of GeSbSe and LNOI, offers scalability and low-loss photonic integrated circuits, with promising applications in high-speed optical communications and signal processing.

Cascaded Feature Fusion Grasping Network for Real-Time Robotic Systems.

Grasping objects of irregular shapes and various sizes remains a key challenge in the field of robotic grasping. This paper proposes a novel RGB-D data-based grasping pose prediction network, termed Cascaded Feature Fusion Grasping Network (CFFGN), designed for high-efficiency, lightweight, and rapid grasping pose estimation. The network employs innovative structural designs, including depth-wise separable convolutions to reduce parameters and enhance computational efficiency; convolutional block attention modules to augment the model's ability to focus on key features; multi-scale dilated convolution to expand the receptive field and capture multi-scale information; and bidirectional feature pyramid modules to achieve effective fusion and information flow of features at different levels. In tests on the Cornell dataset, our network achieved grasping pose prediction at a speed of 66.7 frames per second, with accuracy rates of 98.6% and 96.9% for image-wise and object-wise splits, respectively. The experimental results show that our method achieves high-speed processing while maintaining high accuracy. In real-world robotic grasping experiments, our method also proved to be effective, achieving an average grasping success rate of 95.6% on a robot equipped with parallel grippers.

Energy-Efficient Integrated Electro-Optic Memristors.

Neuromorphic photonic processors are redefining the boundaries of classical computing by enabling high-speed multidimensional information processing within the memory. Memristors, the backbone of neuromorphic processors, retain their state after programming without static power consumption. Among them, electro-optic memristors are of great interest, as they enable dual electrical-optical functionality that bridges the efficiency of electronics and the bandwidth of photonics. However, efficient, scalable, and CMOS-compatible implementations of electro-optic memristors are still lacking. Here, we devise electro-optic memristors by structuring the phase-change material as a nanoscale constriction, geometrically confining the electrically generated heat profile to overlap with the optical field, thus achieving programmability and readability in both the electrical and optical domains. We demonstrate sub-10 pJ electrical switching energy and a high electro-optical modulation efficiency of 0.15 nJ/dB. Our work opens up opportunities for high-performance and energy-efficient integrated electro-optic neuromorphic computing.

High speed and low power hybrid adder using 22 nm SG-TG FinFET

Abstract This paper demonstrates a new 22-T (transistor) hybrid adder for low-power applications using 22 nm shorted gate (SG) tri-gate (TG) FinFET technology. Large conducting channels and increased gate control make FinFET suitable for high-speed low-power computing circuits. The proposed hybrid adder is designed using pass transistor and CMOS logic. Also, it adopts the state-of-the-art techniques of dynamic leakage balancing on the ground path and stacking ground transistors that reduce parasitic effects and increase the speed of the logic transitions compared to traditional designs. Extensive simulations were executed to characterize for PDP (Power Delay Product), power dissipation, and delay of the hybrid adder circuit. From the Monte Carlo analysis, the robustness and efficiency of the circuit are also observable for minimum variation in power and delay due to process voltage and temperature (PVT) variations. Besides, concerning the supply variability and temperature, the circuit shows excellent stability by keeping the power consumption and delay performance consistent. The proposed 22-T hybrid adder using 22 nm SG-TG FinFET are very well-suited in low power and high-speed processors showing highly stable behaviour with reduced process variability.

Study on Static Explosion Shock Wave Test and Numerical Simulation of Warhead

Abstract In order to study the distribution law of free field and ground reflection shock wave overpressure in the static explosion test of warhead, based on high-speed photography and image processing technology, the free field shock wave overpressure test at different blast center distance was carried out, and the ground shock wave overpressure test at different blast center distance was carried out with ground shock wave measuring device. The static value simulation of the battle department was performed through the limited element analysis software. The results show that the measurement value of shock wave overpressure will be disturbed by the ballistic wave produced by the fragment passing over the sensor sensitive surface and the vibration of the measuring device. In the free field and ground shock wave tests, the absolute values of the average relative errors of theoretical and measured data are less than 25% and 20% respectively. In the numerical simulation, the absolute values of the average relative errors of the theoretical data of free field and ground shock wave overpressure and the simulation data are less than 16% and 34% respectively, and the absolute values of the average relative errors of the theoretical data of ground shock wave overpressure and the simulation data are less than 25% and 18% respectively. The shock wave overpressure values obtained by test and numerical simulation show that at the same detonation center distance, the ground and free field shock wave overpressure values increase with the increase of the warhead loading ratio, and the closer the distance to the detonation center, the more obvious the increase of shock wave overpressure value with the increase of the loading ratio.

Hanbury Brown and Twiss-type optical secret sharing

Towards growing challenges of information security and authentication, various optical techniques based on holography, diffraction, interference, metasurfaces, etc., deliver promising solutions with low energy consumption and parallel high-speed information processing. Here, we report on a new dimension–second-order coherence found in the well-known Hanbury Brown and Twiss (HBT) effect–for performing optical authentication and secret sharing. We develop a method to generate a pair of correlated phase-only masks, each of which is distributed to a shareholder and can produce a specific pattern as authentication under coherent illumination, while two secret images are encrypted in the mutual information of two masks. By combining two masks in two configurations, two secret images can be extracted through spatially cascaded display under coherent illumination and intensity correlation under incoherent illumination, respectively. Conspicuously, two extremes of coherence–spatially coherent or incoherent–will enable the encoding and decoding of two different images with the same phase masks, indicating that the first-order and second-order coherence can be two independent channels for optical cryptography just like other degrees of freedom of light (e.g., polarization). Moreover, we demonstrate a polarization-multiplexing scheme to achieve polarization-selective HBT-type optically secret-sharing with increased capacity, and this type of polarization-phase masks can be readily replaced with metasurfaces.

Morphology and Properties of the Laser-Irradiated Composition “Chrome Coating — Copper Substrate”

Introduction. During pulsed laser processing and modification of the surface of non-ferrous alloys and coatings based on them, several still unresolved issues arise. In particular, the extreme thermal deformation conditions of laser processing are not linked to the peculiarities of structure formation and formation of properties in irradiated “coating — copper substrate” compositions. A metal physical analysis of the possibility and reasons for increasing the adhesion strength of coatings to a metal (copper) substrate during high-speed laser processing is insufficiently substantiated and evidence-based. To make a reasonable choice of technological parameters for the surface hardening mode of non-ferrous alloy products, as well as for obtaining high-quality workable composite layers on their surface, it is necessary to solve the above issues and tasks. The aim of this article is to determine the possibility and conditions for increasing the adhesion strength of a chrome coating to a copper substrate under laser irradiation of the composition.Materials and Methods. Metal physical studies in the work were carried out on samples of non-ferrous alloys of the Cu–Zn system with a chrome electrochemical coating with a thickness of 20 μm. The “copper substrate — chrome coating” composition was irradiated at a Kvant-16 installation with a radiation power density of 70–250 MW/m2. Metallographic structural analysis, scanning probe microscopy, and durometric studies were used in the work.Results. It has been calculated that the dynamic and thermal stresses arising in the laser-irradiated compositions “chrome coating — copper substrate” were about 320 MPa. Metal physical studies revealed that, in extreme thermal deformation conditions of laser treatment, the effect of contact melting was manifested at the boundary of the coating with the copper base. Dynamic recrystallization occurred in the surface layers of the irradiated L62 copper alloy, resulting in the formation of grains with a size of 4.5–5.0 μm on the surface of the alloy with an initial grain size of 25 μm.Discussion and Conclusion. It has been found that the adhesion strength of a chrome coating to a copper alloy substrate increased laser irradiation at a radiation power density of 150 MW/m2. This was due to the formation of a transition region 2–4 μm deep in the contact zone with a structure consisting of sections of mutually insoluble solid solutions based on chromium and copper. Based on the analysis of the copper —chromium state diagram and the model of the temperature field under laser irradiation of the chromium coating, it was suggested that contact melting occurred in the transition zone from the coating to the copper substrate. It was shown that thermostrictive stresses, the calculated quantitative values of which were about 320 MPa, had an initiating effect on the observed processes of structure formation in the laser irradiation zones. It was found that such a level of stresses arising in copper alloys under laser irradiation was sufficient for plastic deformation and dynamic recrystallization of the metal and contributed to the formation of a fine-grained structure (4.5–5.0 μm) with an initial grain size of 25 μm. An analysis of the results of studies of irradiated compositions "coating — copper substrate" allowed us to conclude that they expanded the technological capabilities of the laser method of hardening materials and ensure guaranteed high performance of irradiated products with coatings

Nonreciprocal spin wave channeling in ferromagnetic/heavy-metal nanostrips

Nonreciprocity, unidirectionality, and channeling are essential concepts for potential magnonic applications. Nonreciprocity and unidirectionality ensure the efficient propagation of spin waves along predetermined paths with preferential directions, disrupting the symmetry of counterpropagating waves. Channeling fosters the development of intricate spin-wave networks, enabling more sophisticated functionalities. Integrating these concepts into practical applications will shape the future of spin-wave-based information processing devices. This article theoretically studies the dynamics of spin waves in a ferromagnetic strip coupled to a heavy-metal strip, where the nonreciprocity, unidirectionality, and channeling effects are analyzed. Both backward volume (BV) and Damon–Eshbach (DE) configurations are considered, where the lateral dimensions of the heavy-metal and ferromagnetic strips can differ. Calculations show notable nonreciprocal channeling of spin waves in both DE and BV modes. In the BV configuration, the dispersion is reciprocal with nontrivial localization of lateral confined modes. It is shown that the waves can be channeled into the zones in contact with the HM, where the Dzyaloshinskii–Moriya interaction is active. In the DE configuration, the waves exhibit nonreciprocal spin-wave dispersion, allowing unidirectional and channeled spin-wave propagation. The main results are compared to micromagnetic simulations, where an excellent agreement between both methods is obtained. These findings are relevant for envisioning advanced magnonic devices, enabling precise control over spin-wave propagation for innovative, low-power, high-speed information processing.

Wear of End Mills with Carbon Coatings When Aluminum Alloy A97075 High-Speed Processing

It is recommended to use high-speed milling to maintain an effective material removal rate and the required cutting-edge geometry. However, on the other hand, high speed increases wear, so the surface of the cutters is modified by deposition functional coatings. The wear of end mills made of CTS12D and H10F tungsten carbides during the high-speed processing of aluminum A97075 (B95T1) was compared. To increase the durability of the tools, well-proven technologies for deposition diamond-like and polycrystalline diamond coatings in microwave plasma with different film structures, which were determined by the coating growth conditions, were used. The milling cutter corner was mostly worn out, but the nature of the wear had its characteristics. It was revealed that at a forced cutting mode of about 1000 m/min, cutters made of CTS12D alloy with a nanocrystalline diamond coating with a “cauliflower” structure and with a diamond-like film showed 10% higher resistance. The primary wear mechanism was adhesive. Images of worn cutting edges were obtained using a 3D optical digital image processing system.

Exploring the potential of self-pulsing optical microresonators for spiking neural networks and sensing

Photonic platforms are promising for implementing neuromorphic hardware due to their high processing speed, low power consumption, and ability to perform parallel processing. A ubiquitous device in integrated photonics, which has been extensively employed for the realization of optical neuromorphic hardware, is the microresonator. The ability of CMOS-compatible silicon microring resonators to store energy enhances the nonlinear interaction between light and matter, enabling energy efficient nonlinearity, fading memory and the generation of spikes via self-pulsing. In the self-pulsing regime, a constant input signal can be transformed into a time-dependent signal based on pulse sequences. Previous research has shown that self-pulsing enables the microresonator to function as an energy-efficient artificial spiking neuron. Here, we extend the experimental study of single and coupled microresonators in the self-pulsing regime to confirm their potential as building blocks for scalable photonic spiking neural networks. Furthermore, we demonstrate their potential for introducing all-optical long-term memory and event detection capabilities into integrated photonic neural networks. In particular, we show all-optical long-term memory up to at least 10 μs and detection of input spike rates, which is encoded into different stable self-pulsing dynamics.

Iono-Magnonic Reservoir Computing Utilizing Interfered Spin Wave Manipulated By Ion-Gating

Physical reservoir computing is a promising candidate for implementing high-performance artificial intelligence devices, mimicking biological system by using nonbiological components. The reservoir computing network has fewer learning parameters than conventional deep learning, and the advantage leads to high-speed processing and low electric power consumption. Since the physical reservoir plays a vital role in mapping input information depending on past state nonlinearly into a high dimensional space, the physical device must possess nonlinearity, short-term memory, and high dimensionality. However, some issues (i.e., high electrical power consumption, insufficient computational performance, and large volumes) still need to be addressed in achieving practical physical reservoir.Our previous work revealed that nonlinear interfered spin wave multi-detection exhibits high computational performance on a second-order nonlinear autoregressive moving average (NARMA2) task.[1] While the spin wave is a Joule-loss-free information carrier, and its interference gives a reservoir rich expressive power (e.g., chaos), the computational performance is still inferior to an optoelectrical reservoir. The best way to overcome this problem is to manipulate the spin wave in situ.Herein, we fabricated an iono-magnonic reservoir device to let the research fields of ionics and magnonics collaborate. This scheme is built on the basis of ion-gating, which can drastically manipulate magnetic property through redox reaction triggered by gate voltage (V G) application.[2,3] This study is the first study of manipulating chaotic spin wave interference as the information carrier, achieved with a solid-state electrolyte and its application for high-performance reservoir computing.[4]Figure A shows the fabricated physical reservoir, which consists of a Y3Fe5O12 (YIG) single crystal and a proton-conducting Nafion, and its measurement configuration. Two exciters for interference and two detectors for multi-detection are deposited on the YIG. Protons in Nafion migrate to the YIG by V G application, and in situ magnetic property manipulation due to electron doping is expected.Figure B shows the depth dependence of energy loss measured by electron energy-loss spectroscopy. While the energy loss of a Nafion/YIG to which V G was applied ('biased') at the bulk region was in good agreement with that of a Nafion/YIG to which V G was not applied (‘unbiased’), the energy loss at the Nafion/YIG interface region shifted to the lower energy side, indicating that Fe ion near the interface was electrochemically reduced from trivalent to divalent states. V G dependence of M S and H a is shown in Fig. C. M S of 1984.6 Gauss at V G = 0.0 V is in good agreement with M S of 1984.0 Gauss obtained from magnetization measurement in a pristine YIG single crystal. H a at V G = 0.0 V is 1586.7 Oe. Increasing V G, corresponding to electron doping, decreases both M S and H a. The change saturates at the region of V G ≥ 1.6 V. Spin wave frequency variations at various V G are summarized in Fig. D. The frequencies under magnetic fields of 170 mT increase from approximately 1.3 GHz to 1.32 GHz, corresponding to a shift of 17.9 MHz, and the increase ratios are 1.38 %. This modulated spin wave can achieve a variety of reservoirs and may contribute to improving the computational performance.Figure E shows a schematic concept of reservoir computing. The network was constructed by reservoir layers connected in parallel by utilizing the spin wave property modulated by ion-gating. Input time-series data is transformed nonlinearly to waveforms by 800 nodes (X 1 - X 800) in reservoir layers connected in parallel through utilizing spin waves modulated by eight V G states. Then, these nodes are crossed by 800 output weights W out to generate a reservoir output. The computational error of the NARMA2 task is 9.53 x 10-3, which corresponds to reducing 47.3 %, compared to the computational performance of the nonlinear interfered spin wave multi-detection, and the computational error is much lower than that of the reservoir utilizing the optoelectronic system, as shown in Fig. F. This drastic improvement results from excellent nonlinearity of the chaotic spin wave interference and the ability to map in higher dimensional space by ion-gating, and the iono-magnonic reservoir updated the best performance in other physical reservoirs reported thus far. The iono-magnonic reservoir leads to achievement in high-performance artificial intelligence devices.This work was partially supported by Innovative Science and Technology Initiative for Security Grant Number JPJ004596, ATLA, Japan, JST PRESTO (JPMJPR23H4), and JSPS KAKENHI Grant Numbers JP22H04625 and JP19H05814 (Grant-in-Aid for Scientific Research on Innovative Areas “Interface Ionics”).

Emergent Self-Adaptation in an Integrated Photonic Neural Network for Backpropagation-Free Learning.

Plastic self-adaptation, nonlinear recurrent dynamics and multi-scale memory are desired features in hardware implementations of neural networks, because they enable them to learn, adapt, and process information similarly to the way biological brains do. In this work, these properties occurring in arrays of photonic neurons are experimentally demonstrated. Importantly, this is realized autonomously in an emergent fashion, without the need for an external controller setting weights and without explicit feedback of a global reward signal. Using a hierarchy of such arrays coupled to a backpropagation-free training algorithm based on simple logistic regression, a performance of 98.2% is achieved on the MNIST task, a popular benchmark task looking at classification of written digits. The plastic nodes consist of silicon photonics microring resonators covered by a patch of phase-change material that implements nonvolatile memory. The system is compact, robust, and straightforward to scale up through the use of multiple wavelengths. Moreover, it constitutes a unique platform to test and efficiently implement biologically plausible learning schemes at a high processingspeed.

Comparison of CNN-Based Architectures for Detection of Different Object Classes

(1) Background: Detecting people and technical objects in various situations, such as natural disasters and warfare, is critical to search and rescue operations and the safety of civilians. A fast and accurate detection of people and equipment can significantly increase the effectiveness of search and rescue missions and provide timely assistance to people. Computer vision and deep learning technologies play a key role in detecting the required objects due to their ability to analyze big volumes of visual data in real-time. (2) Methods: The performance of the neural networks such as You Only Look Once (YOLO) v4-v8, Faster R-CNN, Single Shot MultiBox Detector (SSD), and EfficientDet has been analyzed using COCO2017, SARD, SeaDronesSee, and VisDrone2019 datasets. The main metrics for comparison were mAP, Precision, Recall, F1-Score, and the ability of the neural network to work in real-time. (3) Results: The most important metrics for evaluating the efficiency and performance of models for a given task are accuracy (mAP), F1-Score, and processing speed (FPS). These metrics allow us to evaluate both the accuracy of object recognition and the ability to use the models in real-world environments where high processing speed is important. (4) Conclusion: Although different neural networks perform better on certain types of metrics, YOLO outperforms them on all metrics, showing the best results of mAP-0.88, F1-0.88, and FPS-48, so the focus was on these models.

High-speed and low-power molecular dynamics processing unit (MDPU) with ab initio accuracy

Molecular dynamics (MD) is an indispensable atomistic-scale computational tool widely-used in various disciplines. In the past decades, nearly all ab initio MD and machine-learning MD have been based on the general-purpose central/graphics processing units (CPU/GPU), which are well-known to suffer from their intrinsic “memory wall” and “power wall” bottlenecks. Consequently, nowadays MD calculations with ab initio accuracy are extremely time-consuming and power-consuming, imposing serious restrictions on the MD simulation size and duration. To solve this problem, here we propose a special-purpose MD processing unit (MDPU), which could reduce MD time and power consumption by about 103 times (109 times) compared to state-of-the-art machine-learning MD (ab initio MD) based on CPU/GPU, while keeping ab initio accuracy. With significantly-enhanced performance, the proposed MDPU may pave a way for the accurate atomistic-scale analysis of large-size and/or long-duration problems which were impossible/impractical to compute before.

Research on Autonomous Navigation of Unmanned Aerial Vehicles Based on Correlation-Based Image Comparison Methods

Relevance. Autonomous navigation of unmanned aerial vehicles (UAVs) is one of the key challenges in the modern aerospace industry. Specifically, for small UAVs, the task of autonomous navigation becomes even more complex due to limitations in computational resources and sensor capabilities. Optimizing navigation methods under such conditions is a pressing issue that requires solutions capable of providing high performance with minimal resource consumption.Objective. Improving the efficiency of autonomous navigation for small UAVs by using correlation methods for image comparison. Achieving this objective is linked to the development and evaluation of algorithms that provide high-speed and accurate navigation with limited computational resources. The work uses methods such as the autocorrelation function, Pearson correlation, the structural similarity index, and a simple neural network for image comparison.Solution. The research showed that the autocorrelation-based approach demonstrates the best performance under low computational resources. It ensures high-speed processing of the entire image and shows optimal detection accuracy. Compared to other methods presented in the study, the autocorrelation approach is capable of working not only with noise in the "reference" map but also of using alternative areas with altered patterns as detected regions.The scientific novelty of the study is determined by the systematic comparison of various methods applied to the task of image comparison for small UAVs with limited computational resources. Unlike well-known works in the field of correlation-extreme systems, this research focuses on the use of a "reference map" and a search area, representing two different images of the same terrain taken from different sources. This is a key difference, as most methods are not highly efficient in processing such images where patterns may differ significantly.The practical significance of the developed algorithm lies in the fact that the proposed autocorrelation-based method can be used by developers of autonomous small UAV control systems to reduce computational load and increase data processing speed.

Dynamic deformation and fracture of brass: Experiments and dislocation-based model

In this work, we perform a comprehensive study of the dynamic deformation and fracture of brass, including Taylor tests with classical and profiled cylinders and ball throwing experiments reaching the strain rates of about (0.1−1)/μs, as well as atomistic and continuum-level numerical modeling. Molecular dynamics (MD) simulations are used to construct the equation of state (EOS) of brass and to study its fracture characteristics at shear deformation under negative pressure. An original model of fracture under combined tensile-shear loading is formulated, which takes into account both the accumulation of empty volume in the process of lattice loosening due to the lattice defect production in the course of plastic deformation and further mechanical growth of voids controlled by the dislocation plasticity. This atomic-scale model is transmitted to the macroscopic experiment-scale level and embedded into 3D dislocation plasticity model to describe the dynamic deformation and fracture of brass using the numerical scheme of smoothed particle hydrodynamics (SPH). A part of experimental data is used to find the optimal parameters of the dislocation plasticity model by means of the Bayesian global optimization method accelerated with the help of artificial-neural-network (ANN)-based emulator of the 3D model. Another part of experimental data is used to fit the fracture model parameter. The remaining experimental data, which are not used in the parameterization, are applied to verify the parameterized model. The developed physical-based model provides correct and meaningful description of the dynamic deformation and fracture of brass, while the developed formalized approach to its parameterization opens a way to wider use of this type of models in the engineering applications, including studies on dynamic performance and high-speed processing technologies.