12results about How to "Help capture" patented technology

The invention relates to the technical field of photosynthesis power generation, and discloses a device and a method for realizing photosynthesis power generation based on a liquid metal-semiconductor junction. The negative electrode comprises a metal-semiconductor junction (MSJs) consisting of liquid metal and a silicon wafer; the thylakoid suspension is positioned between the positive electrode and the negative electrode and is extracted from spinach. The liquid metal is eGaInSn room temperature liquid metal. The method comprises the following steps: extracting crushed chloroplasts and thylakoid which are rich in photosynthetic devices such as a photosystem I (PSI), a photosystem II (PS) and the like in spinach leaves by using a salt solution separation method; in combination with flexible transparent indium tin oxide (ITO), the photosynthesis power generation device with the structure of Si / LMs / BBY / ITO is manufactured. The liquid metal and the silicon wafer are combined to obtain the MSJs as the negative electrode, the electron transfer efficiency of the power generation device is enhanced, the preparation is simple, the performance is excellent, and green power generation is realized.

The present invention discloses a method for identifying transient DIM in a power system, comprising the following steps: 1) obtaining a number of bus voltage curves and power angle curves of the power system; 2) using the K-means clustering algorithm to perform typical fault curve screening and weight calculation on the bus voltage curve and power angle curve of the power system to obtain the bus voltage instability curve and the power angle instability curve; 3) using the MultiRocket algorithm to extract the multidimensional time series characteristics of the bus voltage instability curve and the power angle instability curve; 4) based on the multidimensional time series characteristics, constructing a basic model, and using the Bagging integrated learning and integrated optimization method to optimize the basic model to obtain a DIM identification model; 5) using the DIM identification model to complete the transient DIM identification of the power system. The present invention combines the error-bifurcation theory and proposes a DIM intelligent identification framework based on integrated optimization to improve the prediction accuracy and credibility.

The present invention relates to the technical field of alloy visual recognition, and is a method and system for detecting the morphology of an aerospace-grade intermediate alloy based on visual recognition, comprising: photographing an alloy sample to be tested to obtain a grayscale alloy image set, sequentially extracting grayscale alloy images from the grayscale alloy image set, obtaining a grayscale detection accuracy group according to the grayscale alloy image, obtaining a layered alloy image group according to the grayscale detection accuracy group, performing feature analysis on the layered alloy image to obtain an image grayscale mean difference, judging whether the image grayscale mean difference is less than a standard grayscale mean difference, if not, returning to the step of obtaining the grayscale detection accuracy group, if less than, recording the layered alloy image as a target alloy image, performing morphology analysis according to the target alloy image set, and obtaining an alloy damage range. The present invention can improve the accuracy and efficiency of aerospace-grade intermediate alloy morphology detection.

The invention relates to a stirring casting preparation method suitable for a superfine SiCp reinforced aluminum-based composite material. The problems that in a traditional technology, wettability of SiCp and aluminum melt is poor, agglomeration is prone to occurring, an Al4C3 brittle phase is prone to occurring, and the melt is prone to being polluted by impurities and high in porosity are solved. The preparation method comprises the following steps: performing in-situ growth of graphene on the surface of SiCp through pulse laser-methane micro-explosion to form a composite reinforced phase; carrying out aluminum alloy fractional smelting and step-by-step stirring to obtain a composite melt; an external coil-iron core assembly is adopted to carry out pulse magnetic oscillation on the melt, and mechanical stirring and argon degassing are synchronously matched; and finally casting and molding. According to the method, generation of Al4C3 can be avoided, uniform dispersion of SiCp is achieved, the density of the obtained composite material is larger than or equal to 99.2%, the 0.2% yield strength is larger than or equal to 220 MPa, the heat conductivity retention rate is larger than or equal to 99%, the increment of the iron content of the melt is smaller than or equal to 50 ppm, ton-level mass production can be achieved based on conventional equipment, and the method is suitable for industrial preparation of the high-performance aluminum-based composite material in the fields of aerospace and automobiles.

The present invention relates to the field of encryption technology, and in particular to a method for detecting encrypted malicious traffic. The present invention can generate diversified training data by combining random noise vectors with malicious traffic samples, which helps the model learn more features. The bidirectional model can simultaneously consider context information and extract richer feature representations, which helps capture complex patterns and time series relationships in encrypted traffic. By combining bidirectional feature extraction, forged sample generation and adversarial training, a multi-level learning framework is formed to avoid dependence on a single algorithm. By combining time series features and local spatial features, the model can effectively capture the dynamic changes and spatial correlation of traffic, improve the accuracy of detection, and through real-time traffic monitoring and confidence scoring, the method effectively solves the limitations of existing encrypted malicious traffic detection and enhances the real-time, dynamic adaptability and detection accuracy of the system.

The present invention relates to the field of intelligent speech recognition, and in particular to an artificial intelligence speech recognition method based on the industrial Internet. A speech receiver cluster is set. When a recognition device detects that a product to be detected has passed, a trigger instruction is issued to perform a quality test. The speech receiver pre-learns the quality test audio corresponding to each product to be detected to form audio pre-learning data. An audio learning model generates a quality test distribution map, which is compared with a quality test threshold to form a quality test result. The position of the trigger element is adjusted again according to the quality test result, and the quality test result of the product to be detected is obtained again. Faulty products that have not passed the quality test are screened, a focusing instruction is issued to the recognition device, and a high-definition image of the faulty product is taken and uploaded to a recognition platform. Through speech recognition and machine learning, automated quality inspection of assembly line products is realized, and the inspection strategy can be adjusted in real time to adapt to different products and quality requirements, thereby improving inspection efficiency and product quality.

The present invention relates to the field of image processing technology, and in particular to an infrared and weak visible light image registration method and system based on color edge enhancement, the method comprising: building an image registration model, the image registration model comprising two branches, the first branch performing feature extraction and registration processing on an infrared image and a weak visible light image, the second branch performing feature extraction and registration processing on an infrared image and an enhanced weak visible light image, and combining the registration results of the two branches to obtain a more accurate infrared and weak visible light image registration effect; when extracting features of the infrared image and the weak visible light image of the two branches, additionally integrating edge features extracted by a method based on an attention mechanism, so that the edge features of the infrared image are fully utilized; by performing color enhancement processing on the weak visible light image, the comprehensiveness of the feature expression of the weak visible light image is improved, and the alignment accuracy between the infrared image and the visible light image is improved.

The invention provides a dynamic penetration testing device and method based on effective penetration energy, and relates to geotechnical engineering reconnaissance, the device comprises a support, a displacement sensor, a guide rod, a center penetrating hammer, a hammer seat, a probe rod, a pipe wrench, a cable, a processor, an energy collection connector and the like, and a strain and acceleration sensor is arranged in the energy collection connector; the strain and acceleration sensor is used for collecting the strain and acceleration fed back by the penetration of the sounding probe into the soil body, and calculating the effective penetration energy of the sounding probe in the hammering process based on the data. The method has the beneficial effects that the effective penetration energy in the successive hammering process is accurately obtained, and fine division of a stratum structure is facilitated; absorption and dissipation of hammering energy by hole depth, side friction resistance and drill rod vibration are eliminated in measurement data, and the measurement data are not limited by the penetration depth in the industry specification any more; proper hammer weight and drop distance can be flexibly selected according to the actual test depth; the automation degree is high, automatic data acquisition and processing can be achieved, and manual errors are effectively avoided.

The application relates to a high-function autism screening method based on a clinician-child dialogue, and relates to the technical field of natural language processing and medical auxiliary diagnosis. The application aims to solve the problem that existing screening relies on expensive equipment or manual features and is difficult to capture subtle abnormalities in children's language. The implementation steps are as follows: obtaining a doctor-child question and answer dialogue text; adopting a sliding window to combine multiple continuous question and answer pairs into an input sequence; using a pre-trained BERT model to extract word-level deep semantic representation; dynamically weighting the contribution of each word by an adaptive word aggregation module to obtain sequence-level aggregated features; performing paragraph-level classification; averaging the prediction results of all paragraphs of the same child to obtain individual screening results. The application realizes low-cost and high-efficiency early screening of HFA by analyzing the semantic, syntactic and pragmatic abnormalities in the dialogue language, and highlights fine-grained language clues with diagnostic value.

The invention discloses an oral cavity digital impression instrument image acquisition method and system, and the method comprises the steps: carrying out the down-sampling of oral cavity images collected by an oral cavity digital impression instrument, and weighting the frequency of a gray value in each down-sampling image according to the size ratio of each down-sampling image to the oral cavity image, the method comprises the following steps of: performing edge detection on an oral cavity image under gray levels of different gray scales, calculating the prominence degree of a pixel point according to gradient amplitudes of the pixel point under different gray scales and the length of an edge to which the pixel point belongs, correcting the weighting frequency of a gray value, obtaining a mapping interval of the gray value according to the correction frequency, and obtaining the mapping interval of the gray value according to the mapping interval; according to the protruding degree of the pixel point and the gray values of other pixel points in the neighborhood of the pixel point, the mapping gray value of the pixel point is determined from the mapping interval of the gray values of the pixel point, and the mapping gray values of all the pixel points in the oral cavity image form an enhanced oral cavity image so as to reconstruct 3D information of teeth. According to the invention, the accuracy and trueness of 3D information reconstruction are improved.

The application discloses a mixed geometric representation-based pre-training generalization operation method and system of a mechanical arm, and belongs to the field of generalization operation of the mechanical arm. The method comprises the following steps: constructing a 3D-MOV data set containing an object, a fine-grained scene and a desktop 3D point cloud; pre-training a mixed geometric representation encoder based on the data set; rendering the 3D point cloud from multiple perspectives, mapping to a hyperbolic space, optimizing learning, returning to the Euclidean space to complete self-supervised reconstruction; and based on the encoder, fine-tuning a mechanical arm operation task: rendering the scene observation point cloud into an RGB image, processing the image, distributing the image through two channels, fusing information in a strategy network to complete fine-tuning, and predicting the mechanical arm action. The hyperbolic geometric representation is introduced into the 3D multi-perspective visual self-supervised pre-training, and the 3D multi-perspective representation based on the mixed geometric space is pre-trained for the mechanical arm operation, so that the next operation action can be effectively predicted, and the generalization operation capability of the mechanical arm in different scenes is greatly enhanced.

The invention relates to the technical field of power grid fault classification diagnosis, in particular to a power grid fault diagnosis method and system based on text vectorization, and the method comprises the steps: carrying out the word segmentation processing and embedded conversion processing of a power grid alarm information text, obtaining an initial vector, extracting a keyword, and carrying out the vectorization processing of the keyword, and obtaining a label vector; encoding the initial vector for several times to generate a text vector, then splicing the text vector and a tag vector to obtain a spliced vector, carrying out feature extraction for several times to generate a high-dimensional feature vector, and carrying out full-connection forward processing on the high-dimensional feature vector to obtain a power grid fault diagnosis result. Deep semantic and logical relationships in the text can be mined, fault type samples with similar features are distinguished, and the accuracy of power grid fault diagnosis is improved.