Abstract: The invention discloses a three-dimensional (3D) measurement method based on end-to-end deep learning for speckle projection. First, the speckle pattern was projected by the projector and collected simultaneously by the stereo camera. The speckle images after stereo rectification are fed into the stereo matching network. A feature extraction sub-network based on shared weights processes the speckle images to obtain a series of low-resolution 3D feature tensors, The feature tensor is fed into the saliency object detection sub-network to detect foreground information in the speckle images, producing a full-resolution valid mask map. A 4D matching cost volume is generated using the feature tensor of both views based on the candidate disparity range, filtered by a series of 3D convolutional layers to achieve cost aggregation, so that the initial disparity map is obtained by disparity regression.

Abstract: A high-temperature disaster forecast method based on a directed graph neural network is provided, and the method includes the following steps: S1, performing standardization processing on meteorological elements respectively to scale the meteorological elements into a same value range; S2, taking the meteorological elements as nodes in the graph, and describing relationships among the nodes by an adjacency matrix of graph; then learning node information by a stepwise learning strategy and continuously updating a state of the adjacency matrix; S3, training the directed graph neural network model after determining a loss function, obtaining a model satisfying requirements by adjusting a learning rate, an optimizer and regularization parameters as a forecast model, and saving the forecast model; and S4, inputting historical multivariable time series into the forecast model, changing an output stride according to demands, and thereby obtaining high-temperature disaster forecast for a future period of time.

Abstract: The present disclosure discloses a method for removing chlorinated hydrocarbons in groundwater through step-by-step electrocatalytic dechlorination degradation. A double-chamber electrolyzer reactor is used to carry out step-by-step electrocatalytic dechlorination degradation to remove chlorinated hydrocarbons in groundwater. The double-chamber electrolyzer reactor comprises a cathode chamber, a proton exchange membrane, an anode chamber and an intermediate processing unit, wherein the cathode chamber is separated from the anode chamber through the proton exchange membrane, and the intermediate processing unit is connected between the cathode chamber and the anode chamber through a cathode chamber water outlet, an anode chamber water inlet and pipelines.

Abstract: The disclosure provides a method for detecting and defending a Distributed Denial of Service attack in an SDN environment. The method includes: building data messages acquired as feature messages by a proxy module; sending the feature messages to a pre-built detection model to obtain a detection result; making a decision instruction based on the detection result; and performing control operations by the proxy module based on the decision instruction.

Abstract: This invention proposes a single-vector-based finite control set model predictive control for two parallel power converters, which adopts a centralized control structure to achieve accurate control of overall performance. It establishes predictive models for line currents and three phase-circulating currents and constructs a novel cost function that uses these currents as performance indices to implement the predictive control algorithm based on the proposed predictive models. The invention proposes dynamic weighting coefficients and adjustment principles to improve system control performance. A finite set output signal matrix containing important characteristic information of all alternative vectors is constructed to avoid redundant calculations in each control horizon, reducing computation time during practical implementation.

Abstract: A method for assessing microbial drug resistance multi-level risks of antibiotic residues in water environments, belonging to the technical field of water environment assessment, comprises the following steps: S1, environment monitoring; S2, preliminary screening of antibiotics: S2-1, determination of n-octanol/water partition coefficient, and S2-2, determination of antibiotic environment concentration; S3, assessment of microbial drug resistance; and S4, high-level assessment. The assessment method of the present disclosure conducts a step-by-step assessment of target antibiotics or target antibiotic derivatives in water environment with risks.

Abstract: The present disclosure provides a meteorological big data fusion method based on deep learning, including the following steps: constructing multi-source meteorological data samples; according to an original resolution of different climate variables, selecting a corresponding super-resolution multiple to obtain an optimized super-resolution module under the constraint of maximizing information retention efficiency; constructing a spatial-temporal attention module using a focused attention mechanism, and selecting a corresponding time stride according to periodic characteristics of different climate variables; constructing a meteorological data fusion model in combination with the optimized super-resolution model and the spatial-temporal attention module; taking a minimum resolution of climate variables as a loss function, and training the meteorological data fusion model with the multi-source meteorological data samples; and importing the acquired real-time meteorological data from multiple data sources into t

Abstract: A semantic segmentation method for aircraft point cloud based on voxelization and three views, including: filtering a collected point cloud followed by centralization to obtain a centralized point cloud; inputting the centralized point cloud into a T-Net rotation matrix network; rotating the centralized point cloud to a front side followed by voxelization to obtain a voxelized point cloud; subjecting the voxelized point cloud to voxel filling to obtain a voxel-filled point cloud; calculating thickness maps of three views of the voxel-filled point cloud, followed by sequentially stitching and inputting to the point cloud semantic segmentation network to train the point cloud semantic segmentation network; inputting the collected point cloud into the trained point cloud semantic segmentation network; and predicting a result semantic segmentation of a 3D point cloud of the aircraft.

Abstract: A three-dimensional (3D) spectrum situation completion method and device based on a generative adversarial network includes performing graying and coloring preprocessing based on incomplete 3D spectrum situations from historical or empirical spectrum data obtained by a UAV through sampling a target region, obtaining three-channel incomplete 3D spectrum situation maps displayed in colors, forming a training set based on the incomplete 3D spectrum situation maps; training the generative adversarial network based on the training set and obtaining a trained generator network in the generative adversarial network, performing graying and coloring preprocessing based on a measured incomplete 3D spectrum situation obtained by the UAV through sampling a specified measurement region, obtaining a three-channel measured incomplete 3D spectrum situation map displayed in colors, and using the measured incomplete 3D spectrum situation map as input data to the generator network to obtain a three-channel measured complete 3D

Abstract: A method for preparing an electrochemiluminescent nanoprobe according to an embodiment includes: adding a metal complex ion to an inorganic oxide nanoparticle to provide a metal-doped inorganic oxide nanoparticle; and binding a secondary antibody to the metal-doped inorganic oxide nanoparticle to provide a metal-doped inorganic oxide nanoparticle modified with the secondary antibody. The secondary antibody is configured to identify a specific antibody against nucleic acid-specific site modification.

Abstract: This application discloses a semantic learning-based down-sampling method of point cloud data of an aircraft, including: (S1) constructing a multi-input encoder based on feature learning according to point cloud semantic learning principle; inputting the point cloud data of the aircraft and feature point data into the multi-input encoder for feature fusion followed by decoding using a decoder the multi-input feature fused data to obtain to-be-measured data; (S2) constructing and training a point cloud feature weight calculation network based on semantic learning to acquire a feature weight of each point in the to-be-measured data; and (S3) performing spatial weighted sampling on the feature weight of each point in the to-be-measured data followed by down-sampling based on Gaussian distribution-based spatial sampling principle.

Abstract: The present disclosure discloses a method and apparatus for generating an optical frequency comb. The specific generation method comprises: receiving a pump laser that matches a thermally stable state of a nonlinear optical resonant cavity and causing the pump laser to oscillate in the nonlinear optical resonant cavity, such that a Brillouin gain corresponding to the pump laser coincides with a target longitudinal mode in the nonlinear optical resonant cavity; continuously generating a Brillouin laser at the target longitudinal mode in the case that a pump power of the pump laser exceeds a threshold for generating the Brillouin laser; and generating an optical frequency comb by using the Brillouin laser through a Kerr nonlinear four-wave mixing process. According to the technical solution of the present disclosure, the nonlinear optical resonant cavity with the Brillouin gain can generate an optical frequency comb in its thermally stable region.

Abstract: A few-shot defect detection method based on metric learning, including: (S1) performing data enhancement on a to-be-detected few-shot defect data set through a G2-Generative adversarial network (G2-GAN); (S2) extracting features of a defect data set similar to the to-be-detected few-shot defect data set based on an adaptive convolution kernel-based convolutional neural network (SKM-CNN) to generate a pre-training model; and (S3) transferring the pre-training model to a few-shot defect detection network (S2D2N) based on metric learning; and performing target feature extraction and metric learning in sequence to realize rapid identification and location of defects.

Abstract: An active feedback control method for a quantum communication system based on machine learning is disclosed. In the transmission process of a quantum key distribution system, the present invention uses a pre-trained double-layer LSTM network to predict, according to a real-time ambient temperature, humidity and laser light intensity fluctuation, as well as voltage changes in the past moment, a zero-phase voltage value of a phase modulator at a receiving end at the next moment, and updates the network at a fixed time interval, so that the LSTM network can accurately predict for a long time, ensuring that the quantum key distribution system operates stably and efficiently for a long time. The present invention greatly improves the transmission efficiency of the quantum key distribution system by method of active prediction and feedback control.

Abstract: The invention discloses a miniaturized, low-cost, multi-contrast label-free microscopic imaging system. The imaging system is based on an inverted microscopic structure, a highly integrated optical system is designed by adopting a micro lens having a fixed focal length, and a complex optical system of a traditional microscope system is replaced, such that the whole microscope is highly integrated. The system uses a programmable LED array as an illumination light source the LED array is controlled by a computer to display different illumination modes, six imaging functions of a bright field, a dark field a rainbow dark field, Rheinberg optical dyeing, differential phase contrast, and quantitative phase imaging are achieved; and diversified unmarked imaging methods are provided for biological applications.

Abstract: A heavy-load and high-precision transmission mechanism applicable to sheet-metal bending equipment, which includes pressing arms, connecting rods, hinged supports, pressing arm lifting assemblies and pressing arm lifting driving assemblies. The pressing arms are symmetrically arranged on a top part of a rack. A front end part of each pressing arm facing an upper cross beam is hinged with the cross beam through the connecting rod. An intermediate part of each pressing arm or a rear end part of each pressing arm away from the cross beam is hinged on the rack through the hinged support. The rear end part or the intermediate part of each pressing arm is provided with a set of pressing arm lifting assemblies. Each set of pressing arm lifting assemblies is connected with a set of pressing arm lifting driving assemblies. Each set of pressing arm lifting driving assembly includes an all-electric servo motor.

Abstract: A method for automatically processing a structure-reinforcing member of an aircraft, including: (S1) acquiring, by a handheld laser scanner, data of an area to be reinforced of the aircraft; (S2) controlling a robotic arm to automatically grab the reinforcing member for automatic scanning; (S3) setting a cutting path in a computer aided design (CAD) digital model followed by registration with real data to obtain an actual cutting path, and cutting the reinforcing member; (S4) controlling the robotic arm to guide a cut reinforcing member to a scanning area for automatic scanning; and (S5) subjecting point cloud data of the cut reinforcing member and the area to be reinforced to virtual assembly and calculating a machining allowance to determine whether an accuracy requirement is met; if yes, ending a task; otherwise, grinding the reinforcing member automatically, and repeating steps (S4)-(S5).

Abstract: A subtle defect detection method based on coarse-to-fine strategy, including: (S1) acquiring data of an image to be detected via a charge-coupled device (CCD) camera; (S2) constructing a defect area location network and preprocessing the image to be detected to initially determine a defect position; (S3) constructing a defect point detection network; and training the defect point detection network by using a defect segmentation loss function; and (S4) subjecting subtle defects in the image to be detected to quantitative extraction and segmentation via the defect point detection network.

Abstract: A driver fatigue detection method based on combining a pseudo-three-dimensional (P3D) convolutional neural network (CNN) and an attention mechanism includes: 1) extracting a frame sequence from a video of a driver and processing the frame sequence; 2) performing spatiotemporal feature learning through a P3D convolution module; 3) constructing a P3D-Attention module, and applying attention on channels and a feature map through the attention mechanism; and 4) replacing a 3D global average pooling layer with a 2D global average pooling layer to obtain more expressive features, and performing a classification through a Softmax classification layer. By analyzing the yawning behavior, blinking and head characteristic movements, the yawning behavior is well distinguished from the talking behavior, and it is possible to effectively distinguish between the three states of alert state, low vigilant state and drowsy state, thus improving the predictive performance of fatigue driving behaviors.

Abstract: A method and a device for measuring a four-dimensional (4D) radiation pattern of an outdoor antenna based on an unmanned aerial vehicle (UAV) are provided. The device includes a measurement path planning unit, a UAV platform unit, a radiation signal acquisition unit, a data command processing unit, and a ground data processing unit. The measurement path planning unit, the radiation signal acquisition unit, and the data command processing unit each are suspended from the UAV platform unit by using a pod. The present disclosure applies to the radiation pattern measurement of an outdoor antenna.