532results about How to "Implement extraction" patented technology

The invention relates to the technical field of data processing, and discloses a real-time data integration method and system based on ground observation data. The method comprises the following steps: performing multi-dimensional preprocessing on original marine observation data to obtain time-space standardized observation data; performing multi-parameter cross comparison analysis on the time-space standardized observation data to obtain stratification deviation characteristic data; constructing an adaptive multi-factor correction rule based on the layered deviation characteristic data to obtain a dynamic correction parameter; based on the dynamic correction parameters, performing nonlinear correction on the real-time temperature profile data, and performing intelligent fusion with a preset reference data set to obtain preliminary integrated observation data; performing multi-level quality control and probability uncertainty evaluation on the preliminarily integrated observation data to obtain refined observation data with a credibility index; and performing closed-loop feedback processing on the refined observation data to obtain target integration correction observation data. According to the invention, the efficiency and precision of real-time data integration based on ground observation data are improved.

The invention discloses a patient gait abnormity evaluation auxiliary system based on human body posture detection, and relates to the technical field of intelligent medical treatment, and the system comprises the steps: collecting patient gait video data, preprocessing the patient gait video data, and marking patient skeleton key points; based on the patient skeleton data set, calculating position coordinates of patient skeleton key points, and according to the position coordinates of the patient skeleton key points, identifying motion tracks of the patient skeleton key points; on the basis of the patient skeleton key point motion trail, patient gait features are extracted, and a comprehensive gait feature vector is formed; training a gait prediction model by using the patient skeleton data set, inputting the comprehensive gait feature vector into the gait prediction model, and predicting a gait anomaly index of the patient; and evaluating the prediction result, and generating an evaluation report. The gait abnormity index of the patient is predicted by recognizing the motion trail of the skeleton key points of the patient, comprehensive evaluation of the gait state of the patient is achieved, and clinical diagnosis accuracy and efficiency are improved.

The invention discloses an anomaly detection method and system based on a dynamic traceability graph, and belongs to the field of network anomaly detection.The anomaly detection method comprises the steps that the traceability graph is constructed, and node features of all nodes and features of edges connected with the nodes are spliced to serve as state feature vectors of the corresponding nodes; continuously monitoring system traceability information and dynamically updating the traceability graph; when a new edge et is generated, executing the following steps: inputting the et, a state vector st-of a graph structure surrounding the et at a t-moment and a generation moment t of the et into the graph attention network to obtain an embedded vector z of the et, and updating a source node and a destination node of the et; inputting z into a decoder, and predicting the type of et to obtain a prediction vector P (et); the decoder comprises a long short-term memory network and a plurality of full connection layers which are connected in sequence; calculating a reconstruction error RE between the actual edge type vector of the P (et) and the actual edge type vector of the et; if Th is greater than Th, judging that abnormity exists, otherwise, judging that no abnormity exists. According to the invention, the detection precision and real-time performance of anomaly detection can be improved.

The invention discloses an unmanned aerial vehicle infrastructure construction supervision method and system based on 5G, and belongs to the field of unmanned aerial vehicle monitoring. The unmanned aerial vehicle infrastructure construction supervision system comprises an unmanned aerial vehicle module, a 5G communication module, a data processing module, a cloud storage module, a mobile terminal module, a system management module and a visual management module; the unmanned aerial vehicle module is provided with a high-definition camera, an infrared sensor and a laser range finder, and high-quality construction site data is obtained; the data processing module processes and analyzes the data acquired by the unmanned aerial vehicle; the visual management module compares data acquired by the unmanned aerial vehicle with a construction plan, can generate a real-time visual chart of the construction progress, and displays the construction progress to a supervisor in a visual manner; real-time image transmission is achieved through the system, supervisors can know the progress condition of a project in real time, discover problems in time and take corresponding measures, construction efficiency and quality are improved, and cost and risks are reduced.

The invention discloses an oral cancer multi-task diagnosis system based on a Raman spectrum, and belongs to the technical field of biomedical detection. Comprising an acquisition module and an oral cancer multi-task diagnosis module, and the acquisition module is used for acquiring Raman spectrum data; the oral cancer multi-task diagnosis module is used for processing the Raman spectrum data through a trained oral cancer multi-task diagnosis model so as to generate and output a tumor staging result, a lymph node staging result and a histological staging result; the oral cancer multi-task diagnosis model comprises a bidirectional Mama branch and a residual attention fusion branch, local feature information in the Raman spectrum data is extracted by using the residual attention fusion branch, and multi-scale features are generated by fusing global feature information extracted by the bidirectional Mama branch. Feature information in Raman spectrum data can be effectively extracted, the accuracy of oral cancer diagnosis is improved, and multi-task diagnosis of oral cancer is realized; the problems that multi-task common diagnosis cannot be realized and the diagnosis accuracy needs to be improved in the prior art are solved.

The invention discloses an image semantic segmentation method based on cross-dimension multispectral edge fusion. The method comprises the following steps: firstly, constructing a bilateral multi-modal fusion encoder based on hierarchical attention, and extracting low-level and high-level features in an RGB image and a thermal infrared image; secondly, introducing an edge detection branch (namely a jumping edge guide structure), and enhancing the utilization of global semantic information; and finally, designing a deep supervision mechanism oriented to a plurality of loss functions, and supervising the network with the plurality of loss functions. According to the method provided by the invention, the RGB image and the thermal infrared image are used as input, various attributes and interaction between low-level and high-level features in a dynamic traffic scene are comprehensively considered, and the problems of information loss and the like are solved by fusing semantic and edge features, so that accurate and robust semantic segmentation of the RGB-thermal infrared image under a complex illumination condition is realized.

The invention provides a DEM super-resolution reconstruction method and system fusing gradient feature Transform, and belongs to the technical field of image processing, the reconstruction method is composed of a plurality of residual feature fusion blocks, a multi-scale pyramid segmentation Transform layer in the residual feature fusion blocks fuses multi-head self-attention and pyramid segmentation attention, and the multi-scale pyramid segmentation Transform layer fuses the multi-head self-attention and the pyramid segmentation attention. In order to make the reconstructed DEM closer to reality, a double-filtering convolution block is used, high and low frequency features of the DEM are fully extracted through a parallel structure of two different filtering blocks, and finally super-resolution reconstruction is carried out under double constraints of a gradient domain and a height domain. The method adapts to topographic features of different scales and complexities, and solves the problem that DEM details cannot be accurately recovered by a traditional algorithm.

The invention provides an industrial robot safety monitoring method and system, and relates to the technical field of safety monitoring, and the method comprises the steps: arranging a multi-mode sensor array to collect the state information of a robot; performing empirical mode decomposition on the vibration signal to obtain an intrinsic mode function reflecting the abrasion degree of the robot transmission system; performing wavelet transform on the temperature signal to obtain a temperature feature set reflecting the heat dissipation condition of the robot; through knowledge extraction, knowledge representation and knowledge fusion technologies, semantic representation and linking are carried out on various knowledge in robot components, performance parameters, degradation modes and maintenance measures, and a management knowledge graph corresponding to the robot is constructed; and performing reasoning mining on the management knowledge graph based on a rule mining algorithm of path sorting, extracting degradation characteristics corresponding to each component of the robot in the management knowledge graph by mining a reasoning path formed by frequently co-occurring entities and relationships, and predicting a health state trend of the robot in combination with a pre-constructed time sequence prediction model.

The invention provides a large-scale satellite task dynamic constraint conflict resolution method and system based on heterogeneous graph learning, and the method comprises the steps: carrying out the mathematical description of a task planning problem, and designing an optimization objective function; carrying out characteristic analysis and mathematical expression on the constraint; constructing a satellite task static constraint conflict graph model for the static constraint; constructing a satellite task dynamic constraint conflict graph model for the dynamic constraint; solving the static constraint conflict graph model; solving the dynamic constraint conflict graph model to obtain a satellite task selection probability; and performing decision optimization on the satellite task to finally obtain a satellite task planning result. According to the method, efficient scheduling and allocation decision-making of the multi-satellite large-scale intensive tasks are realized, the problem of efficient calculation of constraint investigation and conflict resolution in the multi-satellite task planning process under the intensive tasks is solved, and the conflict resolution frequency in the multi-satellite task planning solving process is greatly reduced.

The invention discloses an electromagnetic field observation data-based sea wave characteristic analysis method, which comprises the following steps of: processing electromagnetic field data by adopting a time-frequency spectrum analysis method, identifying distribution characteristics of electromagnetic field signals caused by sea waves at different time and frequencies through time-frequency decomposition, and extracting periods and time-varying characteristics of the sea waves; preprocessing the electromagnetic field data by using dominant frequency analysis and band-pass filtering technologies to obtain sea wave induced magnetic field time domain data; time domain and frequency domain induction electromagnetic field polarization direction analysis technologies are combined, and the movement direction of sea waves and the change characteristics of the sea waves along with time are inferred by measuring the change of the electromagnetic field polarization direction; the quantitative relation between the sea wave induction electromagnetic field intensity and the sea wave height is established, and the sea wave height change of the place is effectively estimated through the sea wave induction electromagnetic field intensity. By comprehensively analyzing the electromagnetic field data, the sea wave period, the time-varying characteristics, the motion direction and the sea wave height are evaluated, and a new way is provided for marine environment monitoring.

The invention discloses a coal moisture content and spatial distribution analysis method and system based on an image spectrum, and belongs to the technical field of coal moisture content analysis, and the method comprises the following steps: obtaining a visible light image and a hyperspectral image of to-be-detected coal; semantic segmentation is performed on the visible light image, and the visible light image is divided into a dry area and a water-containing area; the method comprises the following steps: extracting hyperspectral data of each pixel point in a hyperspectral image of a water-containing area, extracting spectral features of a plurality of wavebands by using a sliding window, and dynamically adjusting through an attention mechanism to extract spectral features related to moisture; constructing a sample sequence by taking the spectral feature and the spatial feature of each pixel point of the water-containing area as samples, inputting the sample sequence into a Transform model, and outputting a feature matrix; pooling the feature matrix, inputting the pooled feature matrix into a least square SVM model, and outputting the water content corresponding to each pixel point of the water-containing area; and performing graded developing treatment according to the water contents of different pixel points to obtain accurate coal water and spatial distribution conditions thereof.

The invention discloses a capacitive touch screen touch control adjusting method and device and electronic equipment, and relates to the field of display screens. The method comprises the following steps: acquiring environmental parameters of a plurality of touch points of the capacitive touch screen; determining a calibration touch threshold value corresponding to each touch point according to the environmental parameters; obtaining the touch signal intensity of each touch point; the touch signal intensity of each touch point is compared with a corresponding calibration touch threshold value, the touch state of each touch point is obtained, and the touch state comprises an untouched state and a touched state; generating a first compensation touch point based on the touch state of each touch point, and generating a second compensation touch point based on the touch state of each touch point; and based on the touch state of each touch point and the compensation touch point, generating a touch event corresponding to the touch operation, and sending the touch event to the target program to update a user interface corresponding to the target program. By implementing the technical scheme provided by the invention, the accuracy of touch recognition is improved.

The invention provides an extraction device containing a penthorum chinense pursh extract and a beverage, and relates to the field of food processing. According to the extraction device containing the penthorum chinense pursh extract, through the design of a closed-loop transmission system, continuous feeding, extraction and discharging of penthorum chinense pursh materials are achieved, a driver provides a stable power source through a bevel gear transmission mechanism, it is ensured that the materials continuously move and are stirred in an extraction pipe, and the extraction efficiency is improved. Penthorum chinense pursh materials can be continuously fed and extracted without shutdown treatment. Compared with the prior art, continuous extraction is achieved, the production efficiency is improved, the operation process is simplified, human intervention is reduced, the ultrasonic extraction effect is enhanced, the extraction efficiency is improved, real-time monitoring and feedback are achieved, and it is ensured that the extraction process is stable.

The invention relates to the technical field of intelligent route generation, in particular to a multi-modal large model-based route generation method and system and a storage medium, and the multi-modal large model-based route generation method provided by the invention comprises the following steps: extracting target information corresponding to an obtained webpage link; a target webpage corresponding to the webpage link comprises tourism information of at least two geographic points; determining description information and geographic information corresponding to all target geographic points in the target information based on a multi-modal large model; performing structured processing on the description information and the geographic information of all the target geographic points to obtain structured information; and automatically generating an editable route according to the structured information based on the multi-modal large model, and generating a route text representation and / or a route audio representation of the editable route. The webpage content corresponding to the webpage link is extracted and processed through the multi-modal large model, so that the tourism route of the place where the user wants to go is made by directly utilizing the tourism strategy on the platform.

The invention relates to the technical field of three-dimensional reconstruction, and provides a data synthesis method and a training method for monocular three-dimensional reconstruction based on synthesized data, and the method comprises the steps: 1, rendering a picture containing a specific object into a plurality of pictures I corresponding to the object at different angles; step 2, obtaining a mask m of the object based on the picture I, and 3, guiding the picture generation model to generate a new picture which conforms to the text description and retains the structural information of the specific object by utilizing the picture I by utilizing the mask m and the additionally input text description. According to the method, a large number of 3D models and picture pairs are generated through different viewpoints, styles and background conditions, a rich training data set is provided, and the generalization ability of the monocular three-dimensional reconstruction model can be remarkably improved through the diversity. By randomly applying disturbance within a certain angle range of a plurality of viewpoints (R, T), the adaptive capacity of the model to different camera poses is improved, and the robustness of monocular three-dimensional reconstruction is enhanced.

The invention relates to a railway tunnel portal safety state detection method, system, device, medium and program, and the method comprises the steps: carrying out the two-dimensional data collection of a railway tunnel portal to obtain a tunnel portal image, and carrying out the three-dimensional data collection of the railway tunnel portal to obtain a tunnel portal point cloud; performing multi-layer target feature detection and two-dimensional target labeling on the tunnel portal image to obtain a two-dimensional detection result; performing point cloud denoising and spatial voting clustering on the tunnel portal point cloud to obtain a primary point cloud class group; performing bounding box detection and three-dimensional target labeling on the tunnel portal point cloud according to the primary point cloud class group to obtain a three-dimensional detection result; and performing detection target alignment and scene modeling on the two-dimensional detection result and the three-dimensional detection result to obtain a standard detection result, and performing safety analysis on the standard detection result to obtain the tunnel portal safety state, thereby improving the efficiency of the tunnel portal safety state.

The invention belongs to the technical field of oil-gas exploration and development, and relates to a transient electromagnetic wave through casing resistivity measurement method, which comprises the following steps of: establishing a cased well stratum model by considering a borehole, a casing and a stratum and combining transient electromagnetic wave logging characteristics; in the cased well stratum model, resistivity is measured by adopting an apparent resistivity extraction method based on a late signal, or an anomaly-background method based on a middle signal, or a compensation differential coil structure; the compensation differential type coil structure comprises a transmitting coil and at least one compensation coil system, the compensation coil system at least comprises a first receiving coil and a second receiving coil which are sequentially arranged, the transmitting coil and the second receiving coil are wound in the forward direction, and the first receiving coil serves as the compensation coil and is wound in the reverse direction; and the transmitting coil and the compensating coil system are sequentially arranged from front to back. According to the method, the transient electromagnetic wave through casing logging response precision can be improved, and formation resistivity extraction under the influence of a high-conductivity casing is achieved.

The invention discloses a die bonder thermal error prediction method, device and equipment and a storage medium, and relates to the technical field of data prediction, and the method comprises the following steps: collecting multiple groups of temperature data of a die bonder, and obtaining thermal errors corresponding to the multiple groups of temperature data; adding a plurality of LSTM layers and attention layers into the CNN model to obtain an initial thermal error prediction model; training the initial thermal error prediction model through the multiple groups of temperature data and the corresponding thermal errors to obtain a thermal error prediction model; inputting to-be-predicted temperature data into the thermal error prediction model, and performing feature extraction on the to-be-predicted temperature data through the CNN layer to obtain temperature features; performing time sequence feature extraction on the temperature features through a plurality of LSTM layers to obtain global features; and performing weighted summation and mapping on the global features through an attention layer to obtain a thermal error. According to the method, the CNN, the LSTM and the attention layer are combined, so that the prediction accuracy of the thermal error is greatly improved.

The invention relates to an intelligent building extraction method based on a remote sensing image. The method comprises the following steps: S1, obtaining a high-resolution data set; s2, designing a remote sensing image building semantic segmentation network based on a coding-decoding structure, wherein the remote sensing image building semantic segmentation network comprises a feature extraction skeleton network and a semantic segmentation decoder; s3, a remote sensing image building DSM estimation network based on the generative adversarial network is designed and comprises a DSM generator and a DSM discriminator, and the DSM generator comprises a DSM generator encoder and a DSM generator decoder; s4, designing a feature fusion and reinforcement module; s5, designing a loss function; s6, training and optimizing a remote sensing image building intelligent extraction network according to the high-resolution data set and the loss function; and S7, performing building intelligent extraction based on the remote sensing image through the trained remote sensing image building intelligent extraction network. According to the method, the problem of high DSM acquisition cost is solved, and the performance of extracting the building in the remote sensing image is effectively improved.

The invention provides a rolling bearing fault diagnosis method based on a Swinin-TCN integrated model, and aims to improve the accuracy of fault diagnosis. According to the method, firstly, time-frequency domain features of one-dimensional vibration signals are extracted through fast Fourier transform (FFT) and variational mode decomposition (VMD), and then fault classification is carried out by using a Swinin-TCN integrated model. The Swin-TCN model combines an SE-TCN branch network and a Swin Transform branch network, the SE-TCN branch network extracts global time domain features by using a channel attention mechanism, and the Swin Transform branch network extracts local features by using a window attention mechanism. Model training uses a training set, verification is performed through a verification set to prevent overfitting, and the test set is used for evaluating model performance. Compared with other models, the model provided by the invention is better in accuracy, precision, recall rate and F1 score, shows stronger feature extraction capability, and improves the fault recognition accuracy.

The invention discloses a method for extracting curcumin from turmeric, which comprises the following steps: adding a deep eutectic solvent solution into turmeric powder, adding pectinase, uniformly mixing, adding terpilenol, performing magnetic stirring extraction and centrifugation, and taking an upper solution, thereby obtaining the curcumin. According to the method, a solvent system combining a deep eutectic solvent (DES), pectinase and terpilenol is adopted, extraction and enrichment of curcumin in turmeric can be achieved through a simple one-step method, the process is greatly simplified, recycling and reutilization of reagents are achieved, and compared with an existing extraction method, the extraction method has the advantages of being easy to operate, green, environmentally friendly, high in yield and the like. The method is short in extraction time, low in production cost and suitable for large-scale industrial production.

The invention discloses an instrument electric signal detection system, which comprises a signal acquisition module, a signal processing module, a data display module and a data storage library, and relates to the technical field of electric signal detection. According to the instrument electric signal detection system, the voltage waveform curve graph is established according to the converted digital signals, the change trend of the corresponding digital signals is determined according to the curve processing algorithm, a plurality of zero crossing points in the change process are searched, then curve segmentation operation is carried out with each zero crossing point as a node, and the segmented curves are covered. According to the method, the signal change period is determined, and the corresponding calculation period is compared with the set period of the output source, so that the electric signal detection efficiency is improved, meanwhile, the accuracy of a data result is verified, the detection operation can be quickly distinguished aiming at waveform change data, and actual application is facilitated.

The invention discloses a digital audio watermark processing method and device, equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: receiving a watermark adding instruction, and carrying out Fourier transform on original host voice to obtain an audio signal; identifying a watermark information adding bit in the audio signal through the audio watermark model, and embedding a watermark message in the watermark information adding bit; the interfering audio signal and the audio signal carrying the watermark message are combined. Receiving the message extraction signal, and separating an interference audio signal in the combined audio signal; a watermark information adding bit in the audio signal is identified, and a watermark message is extracted from the watermark information adding bit. The invention also relates to the technical field of block chains, and the host voice is stored on a block chain network. According to the method, the watermark information adding position is automatically identified through the audio watermark model, so that the watermark information is added and extracted, the dependence on expert knowledge is remarkably reduced, and the watermark processing speed and efficiency are improved.

The invention discloses a user identity recognition system and method and a related device, and relates to the field of image recognition, and the method comprises the steps that a data collection module obtains a monitoring image including hand actions of a target user and bank card information of the target user, extracts pre-stored hand action features and pre-stored hand biological features corresponding to the bank card information, and transmits the pre-stored hand action features and the pre-stored hand biological features to a server; outputting an image segmentation feature by using a preset residual module in a preset prediction branch, outputting a hand image feature map based on the image segmentation feature by using a preset feature fusion module, and processing the hand image feature map by using a preset attention module to obtain an identity recognition associated feature map; and the identity recognition module outputs an identity recognition result based on a pre-stored hand action feature, a pre-stored hand biological feature, and a consistency comparison result of the hand action feature and the hand biological feature, and extracts a hand action feature and a hand biological feature of the target user from the pre-stored hand action feature and the pre-stored hand biological feature. According to the invention, the ATM identifies the user identity based on the image identification technology.

The invention relates to the technical field of hydrological monitoring, and discloses a regional groundwater reserve change dynamic monitoring method based on gravity satellite time series data analysis. According to the method, the multi-source data scale effect and observation noise are effectively eliminated, the physical reliability and resolution of groundwater signal extraction are improved, and the recognition precision and early warning perspectiveness of the abnormal loss trend are enhanced. A full-chain monitoring closed loop of multi-source satellite observation, high-precision physical inversion and dynamic statistical benchmark is constructed, multi-dimensional environment data is aligned, a component separation strategy based on multi-element physical mechanism constraint is established based on a water balance principle, and anomaly diagnosis is performed in combination with a self-adaptively updated long-time-sequence historical database.

The invention belongs to the technical field of smart urban rails and computer vision, and particularly provides an image recognition-based method for extracting and recognizing a rail-mounted area in a complex turnout area, and the method comprises the steps: processing an input image through an instance segmentation model to obtain an n-dimensional matrix mask; processing the matrix mask through a rail-mounted area processing module to obtain a matrix mask MaskRe; analyzing the number of rail-mounted areas of MaskRe, and intercepting a turnout area image through a turnout area image acquisition module; obtaining the positioning and reversing states of the turnout, namely the travel path direction of the train, through the target positioning model, so as to obtain the unique rail travel area of the train; and carrying out visualization processing, and outputting an RGB image result. According to the method, the scene image of the complex turnout area is processed through the instance segmentation model, the switch tip state is positioned and identified through the target positioning model, and the current running rail running area of the train in the complex turnout area image obtained at different scenes and shooting angles can be accurately identified and extracted.

The utility model discloses a separation and concentration device for intestinal flora, which belongs to the technical field of biological extraction and comprises a flora extraction device and a negative pressure drainage apparatus, the flora extraction device is sequentially connected with a plurality of filters from top to bottom, and every two adjacent filters are detachably connected. The mesh size of the screen of each stage of filter is gradually reduced from the uppermost stage to the lower stage, the lowermost stage of filter is used for enriching and concentrating the intestinal flora suspension separated by the previous stage of filter, and the negative pressure drainage device is connected with the lowermost stage of filter and is used for collecting liquid generated in the enriching and concentrating process of the lowermost stage of filter; according to the separation and concentration device, the homogenizing device, the multi-stage filtering device and the enrichment and concentration device are integrated into a whole, the flora separation and concentration process is simplified, and the extraction efficiency of intestinal flora is improved.

The invention relates to a script language based on a natural language and a method and system using the script language. The method comprises the steps that a preset configuration file is read and initialized, and a script execution environment is obtained; the input script content is analyzed, a unified command queue and a label mapping table are obtained, and the unified command queue comprises an internal command, an external command and a natural language command; directly executing the internal command to obtain an internal command execution result; executing the external command on the host system to obtain an external command execution result; inputting the natural language command into the large language model for processing to obtain a natural language processing result; compared with an existing script language, the script language is more visual and easy to use, and programming can be conducted through a natural language. Meanwhile, through interaction with the large language model, the execution result of the external command can be interpreted and processed, and the intelligent degree of the system is improved.

The invention discloses a visibility prediction method and system, and the method employs a dynamic window division and abnormal value processing technology based on data variability, solves a problem that meteorological data abnormal values are difficult to accurately recognize, achieves the precise recognition and processing of the abnormal values, and improves the data quality. By adopting the missing value interpolation method based on the weighted distance, the problem that the missing value of the meteorological data is difficult to fill is solved, the accurate interpolation of the missing value is realized, and the integrity and the reliability of the data are enhanced. According to the method, a dynamic normalization processing method is adopted, the problems of different scales of meteorological factor data and local characteristic differences are solved, adaptive normalization of the meteorological factor data is achieved, important fluctuation information is reserved, and the sensitivity of a model to data details is improved.

The invention belongs to the technical field of electric digital data processing, and particularly relates to a large model optimization method for multi-matrix aggregation multi-dimensional features. The method comprises the following steps: firstly, designing a basic optimization operator FoRA: Fusion of Rank; secondly, the basic optimization operator designed in the first step is adapted to a whole TransformerLayer layer of a Transformer large model based on federated learning; and finally, quantifying the Transform large model which completes optimization operator adaptation in the previous step so as to reduce the demand of training the large model for resources, then locally finely adjusting the quantized large model based on federated learning and a client, and converging trainable low-rank parameters after fine adjustment to a global model. The problems that an existing large language model needs a large amount of high-quality data and computing resource training, and downstream tasks cannot be well processed are solved.