19results about How to "Efficient and accurate extraction" patented technology

The invention relates to a power transmission line multi-defect detection method and system based on improved YOLOv8s, and the method comprises the steps: collecting insulator defects and bird nest images on a power transmission line, and obtaining an initial data set; adopting an SAM large model to preprocess the initial data set to obtain a target data set; dividing the target data set into a training set, a verification set and a test set; taking the YOLOv8s network model as a reference model to construct an improved YOLOv8s network model as a power transmission line multi-defect detection network model; inputting the training set into a model for model training; inputting a verification set into the trained power transmission line multi-defect detection network model for model evaluation; and inputting the test set into the model to obtain a power transmission line defect detection result, and evaluating the model. According to the improved YOLOv8s network model, the problems of missing detection and false detection of defects can be avoided, the detection precision of the model is improved, and the light weight of the model is considered at the same time.

The invention relates to a portrait generation and feature similarity recommendation method and system based on a large language model, and the system comprises a data collection module, a feature extraction module, a similarity calculation module and a recommendation module, the data collection module collects multi-modal data from a social media platform, and covers audio, video, text and interactive data information; the feature extraction module performs deep feature analysis and processing based on the collected multi-modal data, and extracts features in multiple dimensions; the similarity calculation module generates a training sample based on the feature matrix generated by the multi-dimensional feature extraction module in combination with a large language model, constructs a similarity matrix of audio, video, text and interactive data features, and calculates similarity scores to determine candidate objects; the recommendation module provides personalized recommendation based on the comprehensive similarity score generated by the similarity determination module. Through combination of a large language model and multi-modal features, the system can accurately analyze behaviors and expressions and generate personalized recommendation.

The invention discloses a raw material sampling device for photoresist production, which relates to the field of photoresist production and comprises a bottom plate and a storage box fixedly mounted on one side of the top of the bottom plate. In the sampling process, a driving motor drives a first crank through a belt transmission piece to promote a first push rod to perform circular motion so as to drive a first driving block to slide, and the mechanism ensures the swing of a swing rod, so that the reverse acting force of a second driving block is triggered, and a second crank also moves along with the second driving block; the rotating rod transmits force to the sliding block to trigger swinging of the movable rod, the action directly controls movement of the second fixed rod and the sampling piece, it is ensured that the sampling pipe can accurately cut into the sample storage box, stable and consistent sampling is achieved, sampling is conducted under the same standard and condition each time, and therefore the consistency of the sampling amount is ensured, and the sampling efficiency is improved. In addition, due to the automatic sampling process in the closed environment, the external pollution risk is reduced to the maximum extent.

The invention provides a form highlighting method and device based on a large model, computer equipment and a storage medium. According to the method, a to-be-recognized form screenshot is input into an image recognition model to obtain an image recognition result, then a unique identifier is added, a first mapping relation is determined, and a text character string and an identifier character string are further constructed. And then instructing the large language model to generate result information required by a highlighting function based on the information, and finally, responding to a highlighting request of a user based on each piece of result information to carry out rapid positioning and highlighting. According to the scheme, a complete process from the form image to structural information extraction and visual display is realized, and the efficiency, accuracy and convenience of form highlighting processing are improved by effectively depending on a large language model.

The utility model provides a disc type multichannel ultrasonic sample processing device which comprises a shell, a circular groove is formed in the upper surface of the shell, a reagent tube mold is fixedly installed in the circular groove and is in a circular ring shape, and a reagent tube is slidably connected into the reagent tube mold. A radial vibration disc is slidably connected to the circular ring center of the reagent tube mold, and a longitudinal vibration transducer is fixedly mounted at the bottom of the radial vibration disc. By designing the reagent tube mold, the longitudinal vibration energy converter and the radial vibration disc, when the reagent tube mold is used, the motor is started to drive the longitudinal vibration energy converter and the radial vibration disc to move upwards in the reagent tube mold so as to relieve extrusion on the inner wall of the reagent tube mold, a reagent tube can be conveniently inserted into the reagent tube mold, and then the longitudinal vibration energy converter is started to drive the longitudinal vibration energy converter and the radial vibration disc to move upwards. And the radial vibration disc and the reagent tube are driven to shake, so that samples are uniformly dispersed and mixed, non-contact and non-pollution of the samples are realized, and the samples can be extracted more efficiently and accurately.

The invention discloses an innovative process for extracting effective components from tylosin waste liquid for liquid fertilizer, and relates to the technical field of fertilizer preparation, and the innovative process comprises the steps of tylosin pretreatment, extraction separation, concentration purification and nutrition proportioning. On one hand, the ultrasonic cavitation effect can effectively break agglomerated particles possibly existing in the tylosin waste liquid, so that filtering is more sufficient and efficient, on the other hand, the polydopamine coating modified ceramic membrane can better intercept tiny impurity particles, meanwhile, the adsorption and blocking phenomena of impurities on the surface of the membrane are relieved, and the filtering efficiency is improved. Compared with a traditional pure physical filtering mode, the tylosin waste liquid treatment device has the advantages that suspended particles, colloids, part of macromolecular impurities and the like in the waste liquid can be more effectively removed, and the tylosin waste liquid is magnetized, so that the treatment efficiency is improved, and the treatment cost is reduced. The interaction between effective components of the tylosin and other subsequently available components and substances such as an extracting agent can be improved.

The invention discloses a multi-modal cognitive screening method and device based on a brain-computer interface and a galvanic skin bracelet. The device comprises a brain-computer interface module, a galvanic skin bracelet, a signal feature extraction module and a cognitive evaluation module. The brain-computer interface module is used for acquiring an electroencephalogram signal sequence when a user performs a cognitive test; the skin electric bracelet is used for acquiring a skin electric signal sequence when a user performs a cognitive test; the signal feature extraction module is used for performing feature extraction processing on the acquired electroencephalogram signal sequence and the acquired skin electrical signal sequence to obtain an electroencephalogram feature sequence and a skin electrical feature sequence; the cognitive assessment module is used for performing cognitive impairment assessment processing on the electroencephalogram feature sequence, the galvanic skin feature sequence, the skin electric signal sequence and the electroencephalogram signal sequence to obtain a cognitive impairment assessment result.

This invention discloses a method for constructing a parameter extraction model for active ingredients in traditional Chinese and Tibetan medicines. The method involves acquiring multi-source sample data of traditional Chinese and Tibetan medicines and constructing key derived features from the preprocessed multi-source sample data. A random forest model is used to generate a corresponding importance score for each key derived feature, forming a feature importance vector. The feature importance vector and the key derived features are weighted and summed to obtain a fusion feature, and the variance and correlation coefficient matrices corresponding to the fusion feature are calculated. The average values ​​of the variance and correlation coefficient matrices are weighted and summed to obtain a comprehensive score. The depth and width of the parameter extraction model to be trained are determined based on the comprehensive score. The number of fully connected modules and neurons in the parameter extraction model is determined by the depth and width to obtain the corresponding target architecture. The parameter extraction model under the target architecture is trained using the fusion features and extracted parameter data to obtain a trained parameter extraction model.

The invention discloses a harmonic state evaluation method for an AC / DC hybrid power system, and the method comprises the steps: collecting the voltage and current waveforms of the AC / DC side of the hybrid power system, carrying out the frequency domain analysis, extracting a harmonic component, building an LCC-HVDC power transmission AC system measurement equation, and a converter transformer branch positive and negative sequence harmonic model, and obtaining an LCC-HVDC power transmission system measurement equation; establishing a linear harmonic state estimation model based on a power network, an element, a measurement, a bus voltage measurement equation, a branch current measurement equation and a power transmission system measurement equation; based on the correlation between the measured value and the state quantity, considering the N-1 criterion, and establishing observability constraint of the AC / DC hybrid power system; and based on the observability constraint and the linear harmonic state estimation model, obtaining a stability index of the AC / DC hybrid power system. According to the invention, harmonic state evaluation and risk analysis can be carried out on the to-be-evaluated system according to the harmonic state stability index, and a harmonic treatment optimization suggestion is given.

The invention relates to a control method of a microwave power amplifier of an unmanned aerial vehicle, and the method comprises the steps: sequentially transmitting signals of different types and intensities to the unmanned aerial vehicle, and enabling an unmanned aerial vehicle signal detection module to recognize the types of received signals, detect the intensities, and give different gains; the microwave power amplifier carries out signal amplification, and the execution system executes corresponding actions according to received signals; acquiring various data of the test flight time period of the unmanned aerial vehicle, and extracting corresponding gain data in which the execution action state of the unmanned aerial vehicle meets specified requirements; constructing a training data sample and inputting the training data sample into the gain prediction model for training; collecting a received signal in real time and inputting the received signal to a gain prediction model for gain prediction; performing corresponding gain amplification processing on the received signal; and monitoring the action execution state of the unmanned aerial vehicle in real time, feeding back the action execution state to the gain prediction model, and optimizing parameters of the gain prediction model. Different amplification processing of signals of different types and intensities is realized, the signal quality is enhanced, and the stability and reliability of the signals are improved.

The application provides a kind of multimodal synergistic industrial voice control method.Application to intelligent manufacturing technology field, the method includes collecting industrial field original voice stream, generates noise spectrum entropy value by frequency domain feature analysis, according to the noise interference level determined by dynamic matching array beam forming weight and adaptive filter coefficient, and gets pure voice instruction by noise reduction;Extract its voiceprint features and semantic text, generate instruction semantic structure tree combined with industrial knowledge graph, calculate identity matching degree and semantic analysis confidence respectively, and get instruction comprehensive effectiveness coefficient by weighting;Satisfy the benchmark time to map it to sub-instruction, evaluate process conflict and equipment risk through digital twin pre-play, determine execution strategy;Real-time acquisition of running data in execution, whether to trigger abnormal rollback is determined according to trajectory deviation rate.This improves the efficiency of industrial voice control.

The invention aims to provide a rolling bearing weak fault feature enhancement extraction method, and belongs to the field of rolling bearing fault diagnosis. Firstly, a 1 / 3 binary tree filter bank is adopted to divide an original signal into a plurality of frequency bands, then spectrum amplitude modulation (SAM) is carried out on each filtered signal, a discrete normalized envelope spectrum peak factor (DNESP) index is established, the most MO range is selected in a self-adaptive mode, and then a normalized square envelope spectrum is reconstructed. And finally, constructing a normalized envelope spectrum peak factor (NESP) as an evaluation index, quantizing and reconstructing the richness of fault information in the normalized square envelope spectrum, and adaptively selecting a demodulation band with significant fault information according to the richness, thereby realizing enhanced extraction of fault features. The method provided by the invention can effectively extract weak fault features of the rolling bearing, and has good robustness for various interferences and noises.

The application discloses a vegetation evolution three-dimensional dynamic feature extraction and visualization method based on remote sensing data, and belongs to the technical field of remote sensing information processing and ecological environment monitoring. The method comprises the following steps: step 1, basic data acquisition and processing; step 2, horizontal gradient vegetation coverage spatial centroid migration calculation; step 3, vertical gradient vegetation evolution feature identification; and step 4, vegetation evolution three-dimensional dynamic feature extraction and visualization. The method disclosed by the application is based on multi-dimensional space fusion analysis of remote sensing vegetation index and digital elevation model data, can identify the horizontal migration characteristics and vertical gradient evolution characteristics of vegetation coverage, comprehensively consider the horizontal and vertical vegetation dynamic change characteristics, construct a vegetation three-dimensional curved surface model, efficiently and accurately extract and visualize the spatial structure and dynamic change trend of large-scale vegetation, and provide high-precision spatial analysis technical support for ecological environment monitoring, surface vegetation dynamic research and ecological restoration evaluation.

The invention discloses a zero abnormal sample fault diagnosis method and system for a low-frequency power transformer, and belongs to the technical field of low-frequency power transformer operation state detection. According to the zero abnormal sample fault diagnosis method for the low-frequency power transformer, original voiceprint data are preprocessed by constructing a voiceprint fragment generation model, a voiceprint feature extraction model, a sample fault diagnosis model and a zero abnormal sample fault diagnosis model; dimensionality reduction compression and accurate and efficient feature extraction of the original voiceprint data are realized; meanwhile, the voiceprint multi-dimensional characteristic parameter matrix is used as model input, a sample fault diagnosis model is trained and tested, a zero abnormal sample fault diagnosis model is formed, efficient and accurate recognition of abnormal voiceprint samples of the low-frequency transformer can be achieved, therefore, under the condition of zero abnormal samples, sound samples which are not seen can be accurately diagnosed, and the fault diagnosis accuracy of the low-frequency transformer is improved. And the generalization capability is high, so that the accurate detection and evaluation of the operation state of the low-frequency power transformer can be realized.

This invention belongs to the field of signal analysis and processing technology, and provides a fault diagnosis method and system based on broadband data compression processing. The method includes: acquiring the fault current signal of the optical-pile-load system; decomposing the fault current signal using wavelet transform to obtain the wavelet packet coefficients and decomposition levels for each frequency band of the fault current; calculating the corresponding wavelet energy entropy based on the wavelet packet decomposition results of the fault current signal under various combinations of wavelet functions and decomposition levels, thereby determining the optimal combination of wavelet mother function and decomposition level; performing wavelet packet decomposition on the normal current signal and the fault current signal based on the optimal combination of wavelet mother function and decomposition level, reconstructing the current signals of each frequency band before and after the fault, and calculating the Tsallis entropy ratio of the current signals of each frequency band before and after the fault to determine the characteristic frequency bands of the fault current signal. This invention can accurately extract the fault characteristic frequency bands from the MHz-level broadband electrical state data of the optical-pile-load system.

The present invention relates to the field of medical big data technology in artificial intelligence, and in particular to a tumor knowledge base construction system and method based on a large model and prompt word optimization, the system comprising: a data acquisition module, a prompt word generation and optimization module, an entity recognition and relationship extraction module, a knowledge graph construction module, a clinical decision support module, and a knowledge base update and maintenance module. The present invention uses prompt words that are initially designed and continuously optimized to ensure that the subsequent large language model LLM accurately and efficiently extracts required information, and uses the large language model LLM to quickly and accurately extract related entities and the association relationship between related entities from target text information, so as to construct an accurate hierarchical tumor knowledge graph database. When the hierarchical tumor knowledge graph database is subsequently applied, a personalized clinical report on recommended treatment plans can be generated by inputting the clinical data of the target patient, thereby providing better decision support for medical personnel.

The application provides a dynamic gesture recognition method based on light-weight multivariate space-time convolution, which comprises the following steps: constructing a space feature extraction module based on a pseudo 3D gated attention fusion network; extracting multi-scale space features by using the space feature extraction module; injecting a guided heat map by using a gated attention fusion module to enhance key region features and suppress background interference, so as to obtain a space refinement feature sequence; decomposing the space refinement feature sequence into multiple sub-variables; capturing long-range and local time dependence in parallel by using a modern convolution module; and modeling the intra-variable and inter-variable relationships respectively by using a decoupling feature interaction network, so as to obtain a dynamic gesture recognition result. By combining the multivariate feature decomposition strategy with the modern convolution, the double-branch design advantage of the modern convolution is to balance long-range dependence and local details, solve the limitations of traditional convolution in capturing long-range time dependence, and model complex dynamics.

The application provides a steel coil number recognition method and system based on DBNet and CRNN. The steel coil number recognition method based on DBNet and CRNN comprises the following steps: acquiring a steel coil image; using a DBNet model to extract a target frame region of a steel coil number in the steel coil image, and returning coordinate information corresponding to the target frame region; cutting the steel coil image according to the coordinate information to obtain a regional image of each steel coil number; taking each regional image obtained by cutting as input, and using a CRNN network model to recognize the character sequence of the number text. The steel coil number recognition method provided in the embodiment of the application can realize automatic, efficient and accurate steel coil number extraction and recognition, and is suitable for steel coil related automatic processing and management scenarios.

The invention provides a data extraction method and device, equipment and a medium. The method comprises the steps that data to be extracted are obtained from a change database, change data obtained through change detection are stored in the change database, and the change data comprise at least two of text data, image data and video data; encoding the to-be-extracted data to obtain feature data of the to-be-extracted data; and performing data extraction on the feature data of the to-be-extracted data according to a preset requirement to obtain extracted data. According to the technical scheme, the data information extraction efficiency is improved.