Abstract: A method for detecting a tension force defect of a steel cable includes the following steps: step S10: providing an excitation sensor on a first position of a steel cable to be detected and providing a detection sensor on a second position of the steel cable; step S20: obtaining a value of a tension force on the steel cable and obtaining a first display diagram corresponding to the value of the tension force; step S30: loading a white noise signal on the excitation sensor through a power amplifier; step S40: acquiring a detection signal collected by the detection sensor; step S50: uploading the detection signal to a personal computer (PC) and performing a Fast Fourier Transform (FFT) process on the detection signal; step S60: determining whether there is a defect in the steel cable, and if so, performing step S70; step S70: determining a defect position of the steel cable.

Abstract: Disclosed are a ground test system and a test method for a space-oriented multi-arm spacecraft system. A spacecraft system simulator floats on an air-floating platform through four porous air feet, a test truss is placed around the air-floating platform, a simulation auxiliary docking device, a simulation crawling truss and a satellite model are arranged in a middle of a ceiling of the test truss, and an assembly test area and a silent air compressor are arranged on sides of the test truss. The application is used to solve the problems that the prior art cannot simulate the movement and crawling of the multi-arm spacecraft system in space, assembly of large space structures, and the prior art cannot simulate the influence of assembling, catching and other actions on a base in a weightless environment.

Abstract: The invention provides a method, apparatus, device and storage medium for predicting a protein binding site. The method comprises the steps of: receiving a protein sequence to be predicted, dividing the protein sequence by using a preset sliding window and sliding step to obtain a plurality of amino acid sub-sequences, building word vectors for the protein sequence according to the amino acid sub-sequences, extracting document features from word elements, building document feature vectors for the protein sequence according to the extracted document features, extracting protein chain biological features from the amino acid sub-sequences, building biological feature vectors for the protein sequence according to the extracted biological features, classifying the amino acid sub-sequences expressed with the document feature vectors and the biological feature vectors by using a preset amino acid residue classification model to obtain amino acid residue types for the protein sequence.

Abstract: A method, a system, and a computer program product for detecting fake news based on a multi-task learning model. In an embodiment, a multi-task learning model is used to perform joint training on authenticity detection and topic classification of news to be detected, and authenticity of the news to be detected and a topic of the news to be detected are returned simultaneously. Through the implementation of the embodiment of the present invention, the authenticity of the news and the topic of the news can be detected simultaneously, and the accuracy of fake news detection and topic classification is improved.

Abstract: A device for gas purification treatment may include: a light oxidation reactor, a light source being disposed in the light oxidation reactor, the light source being configured to emit first light and second light, the light oxidation reactor being configured to perform a first-stage purification treatment on a gas under irradiation of the first light; a catalytic ozone oxidation reactor configured for second-stage purification treatment of the gas; a photocatalytic reactor configured to perform a third-stage purification treatment on the gas under irradiation of the second light; wherein, the photocatalytic reactor is adjacent to the light oxide reactor, and the photocatalytic reactor and the light oxide reactor are separated by a light transmittance component, so that the second light passes through the light transmittance component into the photocatalytic reactor.

Abstract: The present disclosure provides a multimodal fine-grained mixing method and system, a device, and a storage medium. The method includes: extracting data features from multimodal graphic and textual data, and obtaining each composition of the data features, the data features including a visual regional feature and a text word feature; performing fine-grained classification on modal information of each composition of the data features, to obtain classification results; and performing inter-modal and intra-modal information fusion on each composition according to the classification results, to obtain a fusion feature. The method enables a multimodal model to utilize a complementary characteristic of the multimodal data, with no influence by irrelevant information.

Abstract: The present disclosure provides an anomaly detection method and apparatus for multi-type data. According to the anomaly detection method for multi-type data, an adversarial learning network is trained, so that a generator in the adversarial learning network fits a distribution of a normal training sample and learns a potential mode of the normal training sample, to obtain an updated adversarial learning network, an anomaly evaluation function in the updated adversarial learning network is constructed according to a reconstruction error generated during training, and the updated adversarial learning network is constructed into an anomaly detection model, to perform anomaly detection on inputted detection data by the anomaly detection model, to obtain an anomaly detection result. A mode classifier is introduced to effectively resolve difficult anomaly detection when a distribution of detected data is similar to that of normal data, further improving the accuracy of anomaly detection.

Abstract: A nondestructive testing method for detecting and distinguishing internal and external defects of a wire rope includes: acquiring a magnetic flux signal and a MFL signal of a detected wire rope; preprocessing the magnetic flux signal and the MFL signal of the detected wire rope; comparing a preprocessed magnetic flux signal and a preprocessed MFL signal with a preset magnetic flux signal threshold and a preset MFL signal threshold respectively, and calculating a defect position; extracting a magnetic flux signal of a defect and an MFL signal of the defect based on the defect position; calculating a defect width of the detected wire rope based on the magnetic flux signal of the defect and the MFL signal of the defect; calculating a defect cross-sectional area loss of the detected wire rope based on the defect width; and determining whether the defect is the internal defect or the external defect.

Abstract: The present invention relates to a medical wristband with invisible information and a method for manufacturing the same. The conventional developed dyes or inks in the prior art can store information which, however, is visible. In the present invention, polystyrene (PS) microspheres are prepared by micro-emulsification: PS particles are dissolved in a toluene solution and a fluorescent dye is added to obtain a fluorescent solution; an SDS solution is added in and mixed with the fluorescent solution to obtain a system to be ultrasonically emulsified; and the system is ultrasonically emulsified and the toluene is completely volatilized. The ink can be used directly as an invisible ink. The information is written, as two-dimensional codes and by the ink, on a piece of non-fluorescent paper, and plastic-packaged onto a wristband of a patient. Information about a patient can be successfully scanned in the ease of UV on.

Abstract: The present invention belongs to the technical field of material fabrication, and particularly relates to an ultra-sensitive glucose sensor based on a graphene and carbon fiber substrate and a fabrication method thereof. The method includes fabricating a carbon fiber cloth with vertical graphene growth on a surface thereof, performing pretreatment to make the carbon fiber cloth hydrophilic, directly soaking the carbon fiber cloth in a PBS solution of glucose oxidase with the pH of 7.4, and then taking out and drying the carbon fiber cloth at room temperature to obtain a glucose sensor. According to the present invention, the lower limit of glucose detection reaches about 0.1 mM, and the glucose sensor also has multistage corresponding characteristics, so that different detection coefficients and capabilities can be achieved in different glucose concentration ranges. The application range and precision of the glucose sensor are greatly improved.

Abstract: The present disclosure discloses a rotary joint electromagnetic locking device and a rotary joint, the rotary joint electromagnetic locking device comprising a first electromagnetic sucker, a sliding plate, a rotary plate, a connecting shaft connecting the first electromagnetic sucker and the sliding plate, and a second electromagnetic sucker, different from the prior art. When the first electromagnetic sucker and the second electromagnetic sucker have no adsorption force, the sliding plate and the rotary plate can be locked to each other; when the second electromagnetic sucker has an adsorption force, the rotary plate or the first electromagnetic sucker is adsorbed by the second electromagnetic sucker, making the sliding plate and the rotary plate to be further locked; when the first electromagnetic sucker has an adsorption force, the sliding plate can be adsorbed to the first contact surface of the first electromagnetic sucker, making the sliding plate released from the rotary plate.

Abstract: An image privacy perception method based on deep learning, including the following steps: S1, building a privacy classification data set with labeled categories, and training a privacy perception network with a transfer learning method; S2, recognizing a privacy image using a deep convolutional neural network oriented to privacy perception; and S3, extracting an attention profile according to deep convolutional features of the neural network, and locating an attention focusing region to complete the perception of an image privacy region. The method has the following beneficial effects: by completing end-to-end training and testing based on the deep neural network, the privacy image can be accurately distinguished with the privacy region located, facilitating the selective protection of the privacy information in the image.

Abstract: An ultra-long sub-wavelength grating as an optical antenna for optical phased arrays includes a top structure and a bottom structure which are vertically stacked. The bottom structure is made of a material with a refractive index lower than a refractive index of the top structure. The top structure is made of a material with a refractive index higher than that of the bottom structure. A strip waveguide is disposed in the middle of the top structure. subwavelength blocks are disposed periodically on two sides of the straight strip waveguides. The invention has the following beneficial effects. The structure could increase the effective length of the grating; uniform near field distribution can be achieved by controlling the positions of the subwavelength blocks. The structure is simpler with lower fabrication requirements and lower cost.

Abstract: A preparation method of a direct ethanol fuel cell includes synthesizing electrolytes, preparing a cathode and an anode, and clamping the electrolytes between the cathode and the anode to get direct ethanol fuel cell. The electrolytes are synthesized by polymerizing sodium acrylate with an initiator to get a hydrogel, and the hydrogel is soaked in a harsh alkaline solution. The cathode is synthesized by coating N,S codoped carbon catalyst onto a current collector, where the N,S codoped carbon catalyst is synthesized by mixing and preheating silica powder, sucrose and trithiocyanuric acid to get a mixed powder, and mixing and heating the mixed powder with poly tetra fluoroethylene so as to get the N,S codoped carbon catalyst. The anode is synthesized by coating Pt-Ru/C catalyst onto a current collector.

Abstract: A method for determining plane stresses on an in-service steel structure member based on phase spectrum of ultrasonic transverse wave, including: calibrating stress-spectrum parameters k and c of a replica of the in-service steel structure member; determining a first response frequency of a phase difference and a maximum value of a derivative function of the phase difference of an ultrasonic transverse wave echo of the in-service steel structure member, and obtaining a polarization angle of ultrasonic transverse wave components generated by a birefringence effect; solving a plane normal stress difference and a plane shear stress inside the in-service steel structure member; and separating normal stresses by a shear difference method to obtain three independent plane stress components.

Abstract: The invention provides a cloud computing task allocation method and device, an apparatus, and a storage medium. The method comprises: when a cloud task allocation request is received, constructing a cloud task allocation model according to a to-be-allocated cloud computing task in the cloud task allocation request; optimizing cloud computing task allocation for a first preset number of times through the cloud task allocation model and an ant colony algorithm to generate the first preset number of intermediate allocation paths; evolving the intermediate allocation paths for a second preset number of times through a genetic algorithm to generate an optimum allocation path of the cloud computing task; and allocating the cloud computing task to a virtual machine in a cloud environment according to the optimum allocation path, so that the performance of the cloud computing task allocation is effectively improved.

Abstract: The present disclosure proposes a federated speaker verification method based on differential privacy, including: 1. performing, by a server, UBM pre-training to obtain an initial UBM; 2. receiving, by the client, the pre-trained initial UBM, and performing initial UBM learning based on local private speech data; 3. performing, by the client, differential privacy protection based on learned statistics; 4. aggregating, by the server, statistics uploaded by multiple clients, and updating the initial UBM; and 5. receiving, by the client, the updated UBM, performing adjustment based on the local private speech data to obtain a GMM for a user of the client, and determining, based on the updated UBM and the GMM, whether a to-be-verified speech is generated by the user of the client.

Abstract: An image privacy perception method based on deep learning, including the following steps: S1, building a privacy classification data set with labeled categories, and training a privacy perception network with a transfer learning method; S2, recognizing a privacy image using a deep convolutional neural network oriented to privacy perception; and S3, extracting an attention profile according to deep convolutional features of the neural network, and locating an attention focusing region to complete the perception of an image privacy region. The method has the following beneficial effects: by completing end-to-end training and testing based on the deep neural network, the privacy image can be accurately distinguished with the privacy region located, facilitating the selective protection of the privacy information in the image.

Abstract: Disclosed is a method for determining internal uniaxial stress of steel members based on transverse wave phase spectrum, including: manufacturing a replicated steel member of an in-service steel structure member, where the replicated steel member and the in-service steel structure member are the same in material and thickness; loading a test on the replicated steel member to obtain two stress-spectral parameters; performing ultrasonic determination on the in-service steel structure member using an ultrasonic determination device; and collecting transverse wave signals using a signal acquisition system; processing the collected transverse wave signals through an information processing device to obtain a derived curve of the phase spectrum; capturing a first response frequency of the phase spectrum from the phase spectrum derived curve; and obtaining a uniaxial stress of the in-service steel structure member according to the stress-spectral parameters.

Abstract: The present invention relates to a process for preparing a rodlike magnetic ferroferric oxide (Fe3O4) material and use thereof. The preparation includes the following steps: step 1: magnetic Fe3O4 nanoparticle preparation; and step 2: self-assembly of magnetic Fe3O4@SiO2 nanoparticles into a rodlike magnetic material. When in use, the rodlike magnetic Fe3O4 material prepared by the process according to claim 1 is used in micro- and nano-motors, which can implement rotation and deflection in an external magnetic field. The present invention provides a process for preparing a rodlike magnetic Fe3O4 material. The rodlike magnetic ferroferric oxide material prepared by the process is suitable for mass production on an industrial scale, featuring identifiable direction of the magnetic moment, strong magnetism, good magnetic response, simple process, and low cost.