Search

Due to the peculiar non-locality of fractional order differential and integral operators, Caputo fractional order systems are harder than integer order systems to be discoverd from data. To solve this open problem, we propose a framework of method capable of discovering Caputo fractional order (autonomous and non-autonomous) systems from measurable data. The interior point method and genetic algorithm are embedded respectively in the framework. The former is mainly presented in text, while the latter is implemented for comparison and validation. The framework is designed to dynamically and coordinately update the fractional order and vector field function for the system to be discovered till the difference between the measured and discovered systems is minimized. It is computationally efficient, robust and illustrated by discovering the Caputo fractional order Lorenz system, Chua’s circuit and Duffing’s oscillator hidden in measured data. As thus, this work provides one way to uncover underlying Caputo fractional order mathematical models (or physical laws and governing equations).

Retina segmentation plays an important role in the medical field,In recent years, some proposed networks have some problems, such as single receptive field, huge parameters, and difficulty in training, which affect the segmentation results. In this paper, a U-Net-based DM-Net with deep separable convolution and multi-scale is proposed, a residual multiscale module is designed to reduce the parameters and improve the feature extraction ability. In order to cope with the feature information fusion at different levels and the sudden decrease in the number of feature channels in the decoder, the channel attention mechanism is applied. Experiments on the public data set CHASE_DB 1 show that DM-Net has achieved good results compared with other networks, especially in ACC (0.9748) and SP (0.9882). At the same time, it has few parameters and fast convergence speed.

In this work, a new identification method for the load identification of ball mill based on improved K-means clustering is proposed. Since the external device response signals of ball mill are closely related to its load during the operation, the appropriate parameters are selected. The number of clusters and initial clustering centers are meliorated based on their density and genetic algorithm. The dimensionality of input features are reduced with kernel principal component analysis (KPCA), and the final clustering centers are obtained by iterating in the space with reduced dimensionalities. Tests on datasets have demonstrated the higher accuracy and stability of our improved K-means method compared with other methods. The case-analysis results show that the mill load status can be identified clearly according to the displacement variations of cluster centers calculated by this method. Our research provides a new reference for mill load identification in the viewpoint of multi-source information fusion.

In this study, the anisotropic magnetorheological elastomer(MRE) was prepared by mixing two-component room temperature vulcanized(RTV) silicone rubber, carbonyl iron powder particles and dimethyl silicone oil uniformly, then heated and solidified under 500 mT magnetic field intensity. Five types of magnetic particles with different concentrations of 20%, 30%, 40%, 50%, 60% and five different sizes of 1.7, 2.8, 3.9, 4.6, 7.2 μm, were used to investigate the effects of the concentrations and sizes of the magnetic particles on the microstructure, mechanical properties and the magnetorheological effect (MR-effect) of anisotropic MREs. The results obtained from this study showed that the MR-effect has a significant dependence on the concentrations and particle sizes of the magnetic particles. It increased with the increase of the particle concentrations, and the high particle concentration makes the MRE have a high magnetostrictive modulus and MR-effect. Nevertheless, with the increase of the particle sizes, the MR-effect showed a trend of strengthening firstly and then weakening, there is an optimal particle size of 4.6 μm, at this size, the MR-effect is the most significant.

Transition metal oxides (TMOs) as anode materials have potential for lithium-ion batteries (LIBs). However, the poor rate capacity and cycle stability restrict its application. Herein, we demonstrate a facile one-step hydrothermal method to construct a three-dimensional porous conductive network structure, which consists of thin-layered graphene, ultrafine Co3O4-CoO nanoparticles and nitrogen-doped carbon. This unique structure can effectively prevent particle agglomeration and cracking caused by volume expansion, provide fast passage for lithium ion/electron transport during cycling and improve the electrical conductivity of the electrode. Moreover, the electrochemical kinetic analysis proves that this is a process dominated by pseudocapacitive behavior. Consequently, the N-C@Co3O4-CoO@GO hybrid electrode delivers an ultrahigh capacity of 1 273.1 mA h g-1 at 0.1 A g-1 and superior rate performance (725.1 mA h g-1 at 5 A g-1). Additionally, it exhibits a high reversible cycling capacity of 787.4 mA h g-1 at 1 A g-1 over 600 cycles and even maintains excellent cycling stability for a ultra-long cycles at 5 A g-1. This work provides a feasible strategy for fabricating the N-C@Co3O4-CoO@GO composite as a promising high-performance TMOs anode for LIBs.

A pyridine-containing chiral ionic liquid [BPy][L-Pro] with an amino acid as the anion was synthesized, and its synthesis process is very simple with 86.6% yield and 97% purity. Some important physical properties including specific rotation, refractive index, conductivity and density of [BPy][L-Pro] have been comprehensively determined. The dependence of main properties of the [BPy][L-Pro] aqueous solution at different concentrations and temperatures was also explored and fitted, which can provide useful information for its asymmetric catalytic application.

The segmentation of retinal vessels is greatly significant for doctors to diagnose the fundus diseases. However, existing methods have various problems in the segmentation of the retinal vessels, such as insufficient segmentation of retinal vessels, weak anti-noise interference ability. Aiming to the shortcomings of existed methods, this paper proposes an improved model based on the U-Net networks, which contains densely-connected convolutional network and a novel attention gate (AG) model, referred as Densely-Attention-U-Net (DA-U-Net), to automatically segment the retinal blood vessels. The method can alleviate the vanishing-gradient problem, strengthen feature propagation, substantially reduce the number of parameters, and automatically learn to focus on target structures without additional supervision. By verifying the method on the DRIVE datasets, the segmentation accuracy rate is 96.09%, higher than that of U-Net and R2U-Net.

In recent years, deep learning has made great progress in computer vision, and shown a good application prospect in reading medical images. TAO (thyroid-associated ophthalmopathy) is one of the most common orbital diseases in adults is autoimmune disease, the exact pathogenesis is not clear. In hospitals, CT images are usually used for disease analysis and the distortion of ocular rectus muscles is one of the main causes of TAO, thus the separation of the ocular rectus muscle is of great practical significance. However, the artificial segmentation of the ocular rectus muscle is a time-consuming task and relies heavily on the experience of the operato. In this paper, we improved the traditional U-net with GoogLeNet inception module and apply it on the segmentation of ocular rectus muscles. The experimental results show that the segmentation results of our method are more accurate and efficient than the traditional algorithm, and it is helpful for doctors to diagnose patients’ diseases.

Due to the intrinsic low electrical conductivity and large volume expansion of the CoFe2O4 based active materials, designing more novel structures is still one of the most important challenges for its lithium ion battery application. In this work, the CoFe2O4/reduced graphene oxide/carbon (CFO/rGO/C) composite with integrated multi-layer structure has been synthesized through a facial two-step hydrothermal method. Benefiting from the introduction of the graphene network and amorphous carbon coating layer, as well as the accompanying synergistic effect, this composite can exhibit fast and reversible lithium intercalation/deintercalation reactions. With the aid of a surface-induced capacitive process, the CFO/rGO/C composite delivers a superior specific capacity (945 mA h g-1 at 0.1 A g-1) and excellent long-term cyclic stability (421 mA h g-1 at 4 A g-1 with closely 100% Coulombic efficiency after 2000 cycles). Significantly, at a high current density of 1 A g-1, the reversible capacity exhibits a rapid increasing after 100 cycles and finally shows an ultra-high-capacity of 1430 mA h g-1 over 500 cycles. This method could be generalized to the preparation of other similar transition metal oxide-based materials for the development of high-performance energy storage systems.

To overcome inferior rate capability and cycle stability of MnO-based anode materials for lithium-ion batteries (LIBs), we reported a novel 3D porous MnO@GS/CNT composite, consisting of MnO nanoparticles homogeneously distributed on the conductive interconnected framework based on 2D graphene sheets (GS) and 1D carbon nanotubes (CNTs). The distinctive architecture offers highly interpenetrated network along with efficient porous channels for fast electron transfer and ionic diffusion as well as abundant stress buffer space to accommodate the volume expansion of the MnO nanoparticles. The MnO@GS/CNT anode exhibits an ultrahigh capacity of 1115 mAh g-1 at 0.2 A g-1 after 150 cycles and outstanding rate capacity of 306 mAh g-1 at 10.0 A g-1. Moreover, a stable capacity of 405 mAh g-1 after 3200 cycles can still be achieved, even at a large current density of 5.0 A g-1. When coupled with LiMn2O4 (LMO) cathode, the LMO [Formula: see text] MnO@GS/CNT full cell characterizes an excellent cycling stability and rate capability, indicating the promising application of MnO@GS/CNT anode in the next-generation LIBs.

A novel Ag@GQD (graphene quantum dot) hybrid fabricated by a facile two-step strategy is presented: the GQDs are prepared by citrate acid and AgNO3 is reduced. Catalytic studies showed that the Ag@GQD hybrid exhibited excellent photothermal effect and catalytic performance for 4-nitrophenol (4-NP) reduction, suggesting that the GQDs enhanced the catalytic activity via a synergistic effect and the Ag NPs boosted the catalytic efficiency through SPR-mediated photothermal local heating.

MnCo2O4 nanofibers (MCNFs) were successfully synthesized by electrospinning followed by a calcination process under mild conditions. The structural and morphological characterizations of MCNFs were performed by x-ray powder diffraction (XRD), thermogravimetric analysis (TGA), Raman spectroscopy, scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The electrochemical method was used to measure the electrocatalytic hydrogen evolution reaction (HER) properties of MCNFs. Electrochemical studies show that the developed MCNFs possess excellent electroactivity towards HER with low hydrogen-evolution overpotential and a small Tafel slope. In addition, electrospinning can promote the materials into nano-sized architectures with large surface area, which can effectively enhance the electrocatalytic HER properties of MCNFs.

We perform a kinetically complete measurement on the fragmentation of Coulomb explosion of H2O molecules in intense few-cycle linearly and circularly polarized laser fields. Both the fragmentations of H2O3+ and H2O4+ reveal the concerted pathway of dissociation. The length of the O—H bond prior to the Coulomb explosion of both molecular ions is sensitive to the laser pulse duration and laser intensity. However, the bending angle of H—O—H is less sensitive to the pulse duration and laser intensity. We introduce the mechanism of charge resonance enhanced double ionization to elucidate the triple (or quadruple) dissociative ionization dynamics of H2O, in which two electrons are non-adiabatically localized at the protons of the precursor ion H2O+ (or H2O2+) and are released simultaneously due to the over barrier ionization in the combined laser field and molecular ionic potential. Such charge resonance enhanced multiple ionization is not suppressed in few-cycle laser fields and elliptically polarized laser fields.