Unique Matrix Research Articles

High-performance Si@C anode for lithium-ion batteries enabled by a novel structuring strategy.

Si anode has drawn growing attention because of its features of large specific capacity, low electrochemical potential, and high natural abundance. However, it suffers from severe electrochemical irreversibility due to its large volume change during cycling. In spite of the achievement of improved electrochemical performance after compositing with carbon materials, most of the reported Si/C composite anodes lack a simple preparation process. To obtain a promising Si-based anode material, both simple preparation process and improved performance are necessary. Herein, inspired by the structure of shock proof foam, a novel structure of Si-based composite (Si@FeNO@P), consisting of Si nanoparticles embedded within a highly graphitized Fe3C/Fe3O4 hybrid nanoparticle-interspersed foam-like porous carbon matrix, has been constructed using a simple method, consisting of simple mixing, drying, and carbonization processes. Thus, the well-designed composite structure effectively mitigates issues resulting from volumetric change of the Si during cycle and hence improves its performance significantly. The research results confirm outstanding performance of the Si@FeNO@P anode in the aspects of cycle durability, specific capacity, and rate capability, with 1116.1 (250th cycle), 858.1 (500th cycle), and 503.1 (500th cycle) mA h g-1 at 100, 1000, and 5000 mA g-1, respectively. By comparing the performance and structure of Si@FeNO@P with other control samples, it was substantiated that the outstanding performances of the Si@FeNO@P anode depend on the synergistic effects of the well-designed unique carbon matrix, conductive Fe3C, and Fe3O4-in situ derived metallic Fe nanoparticles during cycling. The outstanding electrochemical performance and simple preparation route make the Si@FeNO@P anode promising for lithium-ion battery applications. This work also gives useful insights into the development of high-performance Si-based anodes with simple practical methods.

Pericyte heterogeneity identified by 3D ultrastructural analysis of the microvessel wall.

Confident identification of pericytes (PCs) remains an obstacle in the field, as a single molecular marker for these unique perivascular cells remains elusive. Adding to this challenge is the recent appreciation that PC populations may be heterogeneous, displaying a range of morphologies within capillary networks. We found additional support on the ultrastructural level for the classification of these PC subtypes-"thin-strand" (TSP), mesh (MP), and ensheathing (EP)-based on distinct morphological characteristics. Interestingly, we also found several examples of another cell type, likely a vascular smooth muscle cell, in a medial layer between endothelial cells (ECs) and pericytes (PCs) harboring characteristics of the ensheathing type. A conserved feature across the different PC subtypes was the presence of extracellular matrix (ECM) surrounding the vascular unit and distributed in between neighboring cells. The thickness of this vascular basement membrane was remarkably consistent depending on its location, but never strayed beyond a range of 150-300nm unless thinned to facilitate closer proximity of neighboring cells (suggesting direct contact). The density of PC-EC contact points ("peg-and-socket" structures) was another distinguishing feature across the different PC subtypes, as were the apparent contact locations between vascular cells and brain parenchymal cells. In addition to this thinning, the extracellular matrix (ECM) surrounding EPs displayed another unique configuration in the form of extensions that emitted out radially into the surrounding parenchyma. Knowledge of the origin and function of these structures is still emerging, but their appearance suggests the potential for being mechanical elements and/or perhaps signaling nodes via embedded molecular cues. Overall, this unique ultrastructural perspective provides new insights into PC heterogeneity and the presence of medial cells within the microvessel wall, the consideration of extracellular matrix (ECM) coverage as another PC identification criteria, and unique extracellular matrix (ECM) configurations (i.e., radial extensions) that may reveal additional aspects of PC heterogeneity.

Delay-Dependent Distributed Kalman Fusion Estimation With Dimensionality Reduction in Cyber-Physical Systems.

This article studies the distributed dimensionality reduction fusion estimation problem with communication delays for a class of cyber-physical systems (CPSs). The raw measurements are preprocessed in each sink node to obtain the local optimal estimate (LOE) of a CPS, and the compressed LOE under dimensionality reduction encounters with communication delays during the transmission. Under this case, a mathematical model with compensation strategy is proposed to characterize the dimensionality reduction and communication delays. This model also has the property of reducing the information loss caused by the dimensionality reduction and delays. Based on this model, a recursive distributed Kalman fusion estimator (DKFE) is derived by optimal weighted fusion criterion in the linear minimum variance sense. A stability condition for the DKFE, which can be easily verified by the exiting software, is derived. In addition, this condition can guarantee that the estimation error covariance matrix of the DKFE converges to the unique steady-state matrix for any initial values and, thus, the steady-state DKFE (SDKFE) is given. Note that the computational complexity of the SDKFE is much lower than that of the DKFE. Moreover, a probability selection criterion for determining the dimensionality reduction strategy is also presented to guarantee the stability of the DKFE. Two illustrative examples are given to show the advantage and effectiveness of the proposed methods.

Blue lacy matrix in giant cell tumour of bone with or without denosumab therapy.

Giant cell tumour of bone (GCTB) is genetically characterised by an H3F3A mutation. GCTB is treated with curettage or resection, and denosumab may be administered. Herein, we retrospectively analysed a large cohort of GCTB and identified a previously uncharacterised distinct blue matrix. Among 127 archival GCTB cases positive for the H3F3A G34 mutation, a blue lacy matrix was observed in 10 cases (7.9%). Five patients were previously treated with denosumab. Although a focal observation, the matrix was multifocal or relatively conspicuous in a subset. It appeared as blue mottles or bands, in which a delicate meshwork of basophilic, focally calcified, lacy material intricately surrounded the H3.3 G34W-positive mononuclear cells. The matrix was associated with depletion of osteoclast-like giant cells and fascicular proliferation of spindle cells, regardless of the history of denosumab therapy. Recognising this unique matrix will help avoid confusion with other bone tumours with different clinical management.

Fast Flexible Direct Power Flow for Unbalanced and Balanced Distribution Systems

The study proposes a fast flexible direct power flow solution for radial distribution systems and a fast flexible direct weakly meshed power flow solution for weakly meshed distribution systems. The algorithm is based on the direct forward sweep power flow solution, which is an updated version of the backward/forward sweep solution. The fast flexible direct power flow uses a unique conversion matrix (CM) to rapidly determine the power flow solution. The inverted conversion matrix and its slide-modified matrix are used to solve the power flow problem in a single forward sweep step, which is a novel feature of this work. To ensure the invertibility of the conversion matrix, it is constructed to have a small condition number and a determinant of minus one, and all of its eigenvalues must be equal to that of minus one. Additionally, by modifying the conversion matrix to accommodate the loop branch using the break-point idea, a new weakly meshed conversion matrix (WMCM) is generated with the same following modification as for the radial network and employed in the fast flexible direct weakly meshed power flow (FFDWMPF) solution for the weakly meshed distribution network. The usage of a single matrix in the power flow solution and advanced direct techniques decreases the number of iterations and CPU execution time when MATLAB programming is executed. Furthermore, the proposed method is flexible enough to incorporate any changes in the radial or weakly meshed distribution system just by incorporating the changes in the CM and WMCM for any radial or weakly meshed system. Moreover, the robustness of FFDPF and FFDWMPF is evaluated under various loading scenarios on balanced radial and weakly meshed distribution networks. Finally, to validate the proposed algorithm, the proposed strategy is applied to numerous balanced and unbalanced distribution systems.

Autoencoder Neural Network-Based STAP Algorithm for Airborne Radar with Inadequate Training Samples

Clutter suppression is a key problem for airborne radar, and space-time adaptive processing (STAP) is a core technology for clutter suppression and moving target detection. However, in practical applications, the non-uniform time-varying environments including clutter range dependence for non-side-looking radar lead to the training samples being unable to satisfy the sample requirements of STAP that they should be independent identical distributed (IID) and that their number should be greater than twice the system’s degree of freedom (DOF). The lack of sufficient IID training samples causes difficulty in the convergence of STAP and further results in a serious degeneration of performance. To overcome this problem, this paper proposes a novel autoencoder neural network for clutter suppression with a unique matrix designed to be decoded and encoded. The main challenges are improving the accuracy of the estimation of the clutter-plus-noise covariance matrix (CNCM) for STAP convergence, designing the form of the data input to the network, and making the network successfully explored to the improvement of CNCM. For these challenges, the main proposed solutions include designing a unique matrix with a certain dimension and a series of covariance data selections and matrix transformations. Consequently, the proposed method compresses and retains the characteristics of the covariances, and abandons the deviations caused by the non-uniformity and the deficiency of training samples. Specifically, the proposed method firstly develops a unique matrix whose dimension is less than half of the DOF, meanwhile, it is based on a processing of the selected clutter-plus-noise covariances. Then, an autoencoder neural network with l2 regularization and the sparsity regularization is proposed for the unique matrix to be decoded and encoded. The training of the proposed autoencoder can be achieved by reducing the total loss function with the gradient descent iterations. Finally, an inverted processing for the autoencoder output is designed for the reconstruct ion of the clutter-plus-noise covariances. Simulation results are used to verify the effectiveness and advantages of the proposed method. It performs obviously superior clutter suppression for both side-looking and non-side-looking radars with strong clutter, and can deal with the insufficient and the non-uniform training samples. For these conditions, the proposed method provides the relatively narrowest and deepest IF notch. Furthermore, on average it improves the improvement factor (IF) by 10 dB more than the ADC, DW, JDL, and original STAP methods.

Structural Characterization of Polycrystalline Titania Nanoparticles on C. striata Biosilica for Photocatalytic POME Degradation.

The biosilica shell of marine diatoms has emerged as a unique matrix for photocatalysis, owing to its sophisticated architecture with hierarchical nanopores and large surface area. Although the deposition of titania nanoparticles on diatom biosilica has been demonstrated previously, their photocatalytic activity has been tested only for degradation of pure compounds, such as dyes, nitrogen oxide, and aldehydes. The efficiency of such photocatalysts for degradation of mixtures, for instance, industrial wastewaters, is yet to be investigated. Furthermore, reports on the lattice structures and orientation of nanotitania crystals on biosilica are considerably limited, especially for the underexplored tropical marine diatoms. Here, we report an extensive characterization of titania-loaded biosilica from the tropical Cyclotella striata diatom, starting from freshly grown cell cultures to photodegradation of wastewaters, namely, the palm oil mill effluent (POME). As Indonesia is the largest palm oil producer in the world, photocatalytic technology could serve as a sustainable alternative for local treatment of POME. In this study, we achieved a 54% loading of titania on C. striata TBI strain biosilica, as corroborated by XRF analyses, which was considerably high compared to previous studies. Through visualization using HR-TEM, supported by SAED and XRD analyses, nanocrystal TiO2 appeared to be trapped in an anatase phase with polycrystalline characteristics and distinct crystallographic orientations. Importantly, the presence of C. striata biosilica lowered the band gap of titania from 3.41 eV to around 3.2 eV upon deposition, enabling photodegradation of POME using a broad-range xenon lamp as the light source, mimicking the sunlight. Kinetic analyses revealed that POME degradation using the photocatalysts followed quasi-first-order kinetics, in which the highest titania content resulted in the highest photocatalytic activity (i.e., up to 47% decrease in chemical oxygen demand) and exhibited good photostability throughout the reaction cycles. Unraveling the structure and photoactivity of titania-biosilica catalysts allows transforming marine diatoms into functional materials for wastewater photodegradation.

High-Speed Dicing of SiC Wafers with 0.048 mm Diamond Blades via Rolling-Slitting.

In this study, an innovative fabrication method called rolling-slitting forming, which forms ultra-thin diamond blades, was presented for the first time. Furthermore, the feasibility of the rolling-slitting forming method when applied to silicon carbide wafer dicing blades was investigated; moreover, the cold-pressing blade samples were manufactured through the conventional process under the same sintering conditions to compare and analyze the manufacturing efficiency, organization and performance. The results show that the new method achieves high-precision and low-thickness dicing blades through continuous production without molds-with the thinnest blades being 0.048 mm thick. Furthermore, the rolling-slitting blade has a unique multiporous heat-conductive matrix structure and in-situ generated amorphous pyrolytic carbon, which can reduce the dicing resistance and contribute to a better cutting quality. In addition, the effects of the dicing parameters on SiC were investigated by using indications of spindle current, dicing chipping size and kerf width during the high dicing process. For a dicing depth of 0.2 mm, the ideal performance of dicing SiC with an ultra-thin blade was achieved at a spindle speed of 22,000 rpm and a feed rate of 5 mm/s. This research provides a new idea for the manufacturing of dicing blades, which can satisfy the demand for ultra-narrow dicing streets of high integration of ICs.

Swimming with PFAS in public and private pools

Per- and polyfluoroalkyl substances (PFAS) are a class of anthropogenic chemicals used to provide water and stain resistance in many consumer products. Their widespread use, nearly ubiquitous presence across multiple environments, and growing list of adverse health effects has raised concerns among communities. PFAS have been frequently detected and quantified globally in wastewater, groundwater, surface and drinking water; however, the presence of PFAS in swimming pool water – a unique matrix in which constituents may concentrate through evaporation and which also may present a high risk of direct human exposure – has not been reported. Here, ultra-high performance liquid chromatography - tandem mass spectrometry (UHPLC-MS/MS) was used to monitor 92 PFAS in 54 water samples collected from city, apartment, hotel, and personal swimming pools in six Florida cities. In total, 14 PFAS were detected with six perfluoroalkyl acids – perfluorobutanoic acid (PFBA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA) and perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), and perfluorobutane sulfonate (PFBS) – detected in every sample. PFHxA accounted for 49% of all PFAS quantified in this study. PFAS profiles were compared between sites as a function of pool type, rate of use, and geographic location. Total ΣPFAS concentrations were similar across pool types, with both the highest (633 ng/L) and lowest (1.9 ng/L) measurements found in public city pools. Between sites, higher PFAS levels were observed in city pools in Miami, Melbourne and Tampa compared to Naples, Orlando and Gainesville. Our findings highlight the potential exposure of PFAS in an underexplored and yet important exposure pathway in communities.

Different skeletal protein toolkits achieve similar structure and performance in the tropical coral Stylophora pistillata and the temperate Oculina patagonica

Stony corals (order: Scleractinia) differ in growth form and structure. While stony corals have gained the ability to form their aragonite skeleton once in their evolution, the suite of proteins involved in skeletogenesis is different for different coral species. This led to the conclusion that the organic portion of their skeleton can undergo rapid evolutionary changes by independently evolving new biomineralization-related proteins. Here, we used liquid chromatography-tandem mass spectrometry to sequence skeletogenic proteins extracted from the encrusting temperate coral Oculina patagonica. We compare it to the previously published skeletal proteome of the branching subtropical corals Stylophora pistillata as both are regarded as highly resilient to environmental changes. We further characterized the skeletal organic matrix (OM) composition of both taxa and tested their effects on the mineral formation using a series of overgrowth experiments on calcite seeds. We found that each species utilizes a different set of proteins containing different amino acid compositions and achieve a different morphology modification capacity on calcite overgrowth. Our results further support the hypothesis that the different coral taxa utilize a species-specific protein set comprised of independent gene co-option to construct their own unique organic matrix framework. While the protein set differs between species, the specific predicted roles of the whole set appear to underline similar functional roles. They include assisting in forming the extracellular matrix, nucleation of the mineral and cell signaling. Nevertheless, the different composition might be the reason for the varying organization of the mineral growth in the presence of a particular skeletal OM, ultimately forming their distinct morphologies.

Review of the genus Pleurotroppopsis Girault (Hymenoptera: Eulophidae) with interspecific phylogenetic relationships based on morphological characters.

This study is dedicated to the late Dr. John LaSalle, and reviews the world species of Pleurotroppopsis Girault (Hymenoptera: Eulophidae); fourteen species are treated, of which two are newly described: P. dactylispae Cao & Zhu sp. nov. from China and P. peukscutella Cao & Zhu sp. nov. from Malaysia. On the basis of morphological characters, tentative relationships among genera allied to Pleurotroppopsis are discussed. A revised definition of Pleurotroppopsis is presented based on study of type specimens of all species and a critical review of previous studies on the genus. In addition, parsimony analyses were conducted to infer a phylogeny of Pleurotroppopsis species based on a unique data matrix of morphological characters. Keys to genera allied to Pleurotroppopsis and to known species of Pleurotroppopsis are provided.

Module-assembly of injectable cellular DNA hydrogel via clickable cells and DNA scaffolds

In regenerative therapy, artificial extracellular matrix (ECM)-embedded cells are implanted to exert therapeutic effects in vivo. Often, artificial ECM fails to achieve the required tissue compatibility to promote cell-to-cell communication and retention of cells at the target site. To address this challenge, we introduced a unique breathing therapeutic matrix fabricated by the covalent unit assembly of a DNA microscaffold (DNA microscaf) with therapeutic mammalian cells. In our system, metabolically engineered cells serve as active building units for the final construct as well as therapeutic agents. In contrast, a DNA microscaffold with a pre-assigned clickable moiety serves as a depot for cells to maintain biological functions for subsequent in vivo localization. Notably, the final construct has ultra-soft mechanical properties, enabling the injection of an intact therapeutic matrix without surgery. The active linkages between the cells and scaffolds are gradually diluted as the cells proliferate, allowing the dislodging of cells. The subsequent slow disintegration of the cellular DNA hydrogel also allows for the successful replacement of the damaged tissue.

Targeted Contrast Agents for Magnetic Resonance Molecular Imaging of Cancer.

Magnetic resonance imaging (MRI) is a clinical imaging modality that provides high-resolution images of soft tissues, including cancerous lesions. Stable gadolinium(III) chelates have been used as contrast agents (CA) in MRI to enhance the contrast between the tissues of interest and surrounding tissues for accurate diagnostic imaging. Magnetic resonance molecular imaging (MRMI) of cancer requires targeted CA to specifically elucidate cancer-associated molecular processes and can provide high-resolution delineation and characterization of cancer for precision medicine. The main challenge for MRMI is the lack of sufficient sensitivity to detect the low concentration of the cellular oncogenic markers. In addition, targeted CA must satisfy regulatory safety requirements prior to clinical development. Up to now, there is no FDA-approved targeted CA for MRMI of cancer.In this Account, we discuss the latest developments in the design and development of clinically translatable targeted CA for MRMI of cancer, with an emphasis on our own research. The primary limitation of MRMI can be overcome by designing small molecular targeted CA to target abundant cancer-specific targets found in the tumor microenvironment (TME). For example, aggressive tumors have a unique extracellular matrix (ECM) composed of oncoproteins, which can be used as targetable markers for MRMI. We have designed and prepared small peptide conjugates of clinical contrast agents, including Gd-DTPA and Gd-DOTA, to target fibrin-fibronectin clots in tumors. These small molecular CA have been effective in enhancing MRMI detection of solid tumors and have demonstrated the ability to detect submillimeter cancer micrometastases in mouse tumor models, exceeding the detection limit of current clinical imaging modalities. We have also identified extradomain B fibronectin (EDB-FN), an oncofetal subtype of fibronectin, as a promising TME target to leverage in the design and development of small peptide targeted CA for clinical translation. The expression level of EDB-FN is correlated with invasiveness of cancer cells and poor patient survival of multiple cancer types. ZD2 peptide with a sequence of seven amino acids (TVRTSAD) was identified to specifically bind to the EDB protein fragment. Several ZD2 conjugates of macrocyclic GBCA, including Gd-DOTA and Gd(HP-DO3A), have been synthesized and tested in mouse tumor models. ZD2-N3-Gd(HP-DO3A) (MT218) with a high r1 relaxivity was selected as the lead agent for clinical translation. The physicochemical properties and preclinical assessments of MT218 are summarized in this Account. MRMI of EDB-FN with MT218 can effectively detect invasive tumors of multiple cancers with risk-stratification and monitor tumor response to anticancer therapies in mouse models. Currently, MT218 is in clinical trials for precision cancer MRMI. Herein, we will show that using targeted MRI contrast agents specific to abundant TME biomarkers is a pragmatic solution for effective precision cancer imaging in high spatial resolution. And thus, we illustrate a replicable approach for CA development that is vital for cancer MRMI.

A receptance based method for the calculation of nonlinear normal modes of large ordered structures with distributed localized nonlinearities

In recent years, the concept of nonlinear normal modes (NNMs) has gained interest for interpreting a broad range of nonlinear dynamic phenomena. Present study introduces a simple and efficient computational framework to compute NNMs of large order nonlinear structures overcoming difficulties associated with commonly used time-integration and discretization methods. In this paper, a Receptance Based Nonlinear Normal Mode Calculation Method (RBNM) is developed to determine the NNMs of large ordered realistic finite element models in frequency domain. Describing Function Method (DFM), which makes it possible to write the nonlinear internal forcing vector as a nonlinear displacement dependent stiffness matrix multiplied by displacement vector, is used to model the internal nonlinear forcing vector in frequency domain. A unique matrix manipulation based on dynamic stiffness and receptance concepts is used, as a result of which the number of governing nonlinear equations becomes independent from the total number of degrees of freedom of the model and it only depends on the number of degrees of freedom associated with the nonlinear elements. This brings a significant computational advantage over the classical reduced order modeling techniques. A simple 2-degree of freedom (DOF) system, the results of which are available in the literature, is used to verify the proposed method in determining NNMs. A 20-DOF lumped parameter system model and later a complex finite element model of a large structure are used to demonstrate the applicability of RBNM in determining NNMs of larger systems with several different types of nonlinear elements distributed among the degrees of freedom.

Constraint-Induced Symmetric Nonnegative Matrix Factorization for Accurate Community Detection

As a fundamental characteristic of an undirected network, community reveals its networking organization and functional mechanisms, making community detection be a highly-interesting issue in network representation learning. With great interpretability, a symmetric and nonnegative matrix factorization (SNMF)-based approach is frequently adopted to tackle this issue. However, it only adopts a unique feature matrix for describing the symmetry of an undirected network, which unfortunately results in a reduced feature space that evidently impairs its representation learning ability. Motivated by this discovery, this paper proposes a novel Constraint-induced Symmetric Nonnegative Matrix Factorization (C-SNMF) model that adopts three-fold ideas: a) Representing a target undirected network with multiple latent feature matrices, thus preserving its representation learning capacity; b) Incorporating a symmetry-regularizer into its objective function, which preserves the symmetry of the learnt low-rank approximation to the adjacency matrix, thereby making the resultant detector precisely illustrate the target network's symmetry; and c) Introducing a graph-regularizer that preserves local invariance of the network's intrinsic geometry into its learning objective, thus making the achieved detector well-aware of community structure within the target network. Note that the regularization coefficients are selected according to the modularity of the learnt community structure on the training data only, thereby greatly improving the achieved model's practical significance for real applications. Experimental results on six real-world networks demonstrate that the proposed C-SNMF model significantly outperforms the benchmarks and state-of-the-art models in achieving highly-accurate community detection results.

Mechanical Characterization and Microstructural Analysis of Hybrid Composites (LM5/ZrO2/Gr)

Hybrid composites recently developed as highly effective, high-strength structural materials that are increasingly used. Aluminum matrix hybrid composites strengthened with ceramic particulates are commonly used in marine, aerospace, and defense applications because of their exceptional properties. Zirconia-reinforced composites are favored because these composites display high refractory properties, excellent abrasion resistance, and chemical resistance compared to composites of other reinforcements.For applications where lightweight and superior performance is paramount, such as parts for spacecraft, fighter aircraft, and racecars, graphite compositesare the material of choice. In this research work, an effort was made to combine the properties of zirconia and graphite by producing a unique metal matrix composite of LM5 aluminum alloy reinforced with 6% zirconium dioxide (zirconia), using the stir casting process by changing the percentage of the weight of graphite to 2%, 3%, and 4%. The test specimens were prepared and evaluated in compliance with ASTM standards to study micro- and macrohardness, and impact, tensile, and compressive strength. Microstructural studies of composites performed through optical microscopy and SEM expose the unvarying dispersal of particulates of ZrO2/graphite in the aluminum matrix. The hardness, impact, and compressive strength are enhanced due to the addition of reinforcement.

Hybrid Model Compression for Multi-Task Network

In recent years, deep neural networks have made significant achievements in various fields in science and engineering. To achieve outstanding performance, neural network models with very deep structures and an extensive number of parameters are usually used, leading to huge memory storage and computational resources requirements of the hardware systems, and thus preventing them from being deployed in devices with hardware limits. In particular, such problem will be further amplified when multiple neural network models for several tasks are to be integrated into one single platform at the same time. In this paper, a hybrid joint-network optimization model is proposed to solve this multi-task multi-model compression problem. The idea of hard parameter sharing in multi-task learning is adopted. Specifically, we turn the sub-network of the same structure of multiple models into a hybrid network model, in which the same parameters are shared for all the tasks. In addition, a unique sparse matrix for each task is added to each task independently. Experimental results show that the proposed compression methods can not only achieve a high compression ratio but also get performance improvement for some tasks.

Smart Active Vibration Control System of a Rotary Structure Using Piezoelectric Materials.

A smart active vibration control (AVC) system containing piezoelectric (PZT) actuators, jointly with a linear quadratic regulator (LQR) controller, is proposed in this article to control transverse deflections of a wind turbine (WT) blade. In order to apply controlling rules to the WT blade, a state-of-the-art semi-analytical solution is developed to obtain WT blade lateral displacement under external loadings. The proposed method maps the WT blade to a Euler–Bernoulli beam under the same conditions to find the blade’s vibration and dynamic responses by solving analytical vibration solutions of the Euler–Bernoulli beam. The governing equations of the beam with PZT patches are derived by integrating the PZT transducer vibration equations into the vibration equations of the Euler–Bernoulli beam structure. A finite element model of the WT blade with PZT patches is developed. Next, a unique transfer function matrix is derived by exciting the structures and achieving responses. The beam structure is projected to the blade using the transfer function matrix. The results obtained from the mapping method are compared with the counter of the blade’s finite element model. A satisfying agreement is observed between the results. The results showed that the method’s accuracy decreased as the sensors’ distance from the base of the wind turbine increased. In the designing process of the LQR controller, various weighting factors are used to tune control actions of the AVC system. LQR optimal control gain is obtained by using the state-feedback control law. The PZT actuators are located at the same distance from each other an this effort to prevent neutralizing their actuating effects. The LQR shows significant performance by diminishing the weights on the control input in the cost function. The obtained results indicate that the proposed smart control system efficiently suppresses the vibration peaks along the WT blade and the maximum flap-wise displacement belonging to the tip of the structure is successfully controlled.

Tuning the Size of Large Dense-Core Vesicles and Quantal Neurotransmitter Release via Secretogranin II Liquid-Liquid Phase Separation.

Large dense‐core vesicles (LDCVs) are larger in volume than synaptic vesicles, and are filled with multiple neuropeptides, hormones, and neurotransmitters that participate in various physiological processes. However, little is known about the mechanism determining the size of LDCVs. Here, it is reported that secretogranin II (SgII), a vesicle matrix protein, contributes to LDCV size regulation through its liquid–liquid phase separation in neuroendocrine cells. First, SgII undergoes pH‐dependent polymerization and the polymerized SgII forms phase droplets with Ca2+ in vitro and in vivo. Further, the Ca2+‐induced SgII droplets recruit reconstituted bio‐lipids, mimicking the LDCVs biogenesis. In addition, SgII knockdown leads to significant decrease of the quantal neurotransmitter release by affecting LDCV size, which is differently rescued by SgII truncations with different degrees of phase separation. In conclusion, it is shown that SgII is a unique intravesicular matrix protein undergoing liquid–liquid phase separation, and present novel insights into how SgII determines LDCV size and the quantal neurotransmitter release.

Implementation of a Binary Neural Network on a Passive Array of Magnetic Tunnel Junctions

The increasing scale of neural networks and their growing application space have produced demand for more energy- and memory-efficient artificial-intelligence-specific hardware. Avenues to mitigate the main issue, the von Neumann bottleneck, include in-memory and near-memory architectures, as well as algorithmic approaches. Here we leverage the low-power and the inherently binary operation of magnetic tunnel junctions (MTJs) to demonstrate neural network hardware inference based on passive arrays of MTJs. In general, transferring a trained network model to hardware for inference is confronted by degradation in performance due to device-to-device variations, write errors, parasitic resistance, and nonidealities in the substrate. To quantify the effect of these hardware realities, we benchmark 300 unique weight matrix solutions of a 2-layer perceptron to classify the Wine dataset for both classification accuracy and write fidelity. Despite device imperfections, we achieve software-equivalent accuracy of up to 95.3 % with proper tuning of network parameters in 15 x 15 MTJ arrays having a range of device sizes. The success of this tuning process shows that new metrics are needed to characterize the performance and quality of networks reproduced in mixed signal hardware.