2461 results about "Conversion method" patented technology

The invention discloses a query statement conversion method and device based on a large model, equipment and a medium, and relates to the field of computer science, and the method comprises the steps: obtaining a natural language query statement sent by a user side, comparing the natural language query statement with a high-frequency problem in a target knowledge base, and obtaining a query statement conversion result; whether a high-frequency problem with the similarity with the natural language query statement exceeding a preset similarity threshold value exists in the target knowledge base or not is judged; if the high-frequency problem that the similarity with the natural language query statement exceeds a preset similarity threshold does not exist, determining a target database table corresponding to the natural language query statement based on the target knowledge graph, and extracting a creation statement corresponding to the target database table, so as to determine target information corresponding to the target database table according to the creation statement; and analyzing the target information by using a preset large model to generate an SQL statement corresponding to the natural language query statement. According to the method and the device, the natural language query statement can be efficiently and accurately converted into the structured SQL statement.

In the field of document processing, format conversion is a common demand, for example, a PDF (Portable Document Format) or a scanning copy is converted into an editable format such as Word, Excel and the like, a traditional method generally adopts an OCR (Optical Character Recognition) technology to directly recognize document content and output a target format, and the method has the defects that the recognition accuracy is insufficient, errors can only be checked and corrected item by item after the target format is converted, and the efficiency is high. In order to solve the problems of lack of visual comparison with an original manuscript, easiness in omission, incapability of adjusting an identification result in real time in a format conversion process and high subsequent editing workload, the invention provides a brand-new document format conversion method, and provides a visual interactive interface by introducing a three-column visual interface and a real-time team correction mechanism. According to the method, the user can check and adjust the recognition result in real time in the conversion process, the accuracy of document conversion is remarkably improved, the flexibility of document conversion is achieved, the user is supported to carry out secondary editing on the content before formal output, visual comparison between the recognized content and an original manuscript is achieved, and the check difficulty is reduced.

The invention relates to the technical field of artificial intelligence, can be applied to the field of medical health, and discloses a timbre conversion method, device, equipment and medium based on cultural semanteme, the method comprises the following steps: constructing a cultural semantic timbre library comprising a semantic label and timbre characteristic parameter mapping relationship, the semantic label characterizing emotional semanteme of a target timbre, and the timbre characteristic parameter mapping relationship between the semantic label and the timbre characteristic parameter; the timbre characteristic parameters comprise a pitch range, rhythm rhythm and a harmonic structure; performing feature extraction based on text, image and audio multi-mode information to obtain semantic keywords, visual emotion features and audio acoustic features; performing attention weight fusion on the features through a multi-modal fusion deep learning model, and dynamically adjusting model parameters in combination with a semantic timbre library to generate a target timbre; and finally, intelligent conversion from the multi-mode information to the adaptive tone is realized. Through semantic-driven multi-modal feature collaborative optimization, the defect that timbre conversion machinery is stiff and lacks emotional expression is overcome, and the integrating degree of timbre expression and semantic scenes is improved.

The invention discloses a conversion method and system based on SysML and Modelica model semantic mapping. The method comprises the following steps that S1, a Modelica standard model library is imported into a SysML modeling tool; s2, completing the definition of behavior elements, and forming an SysML model instance; s3, exporting a Modelica model which can be executed by the Modelica simulator; s4, establishing a mapping relation table between SysML model elements and Modelica model elements; s5, when a change event occurs in the SysML model, updating the content of the corresponding Modelica model; s6, the Modelica model is imported into the SysML model, and a corresponding SysML model instance is generated; s7, simulation result mapping and behavior modeling generation are executed, a state transition diagram structure is formed, and structure updating of internal block diagrams in the system model is completed. According to the method, bidirectional conversion and joint simulation optimization between models are realized, and the method is suitable for cross-platform modeling, collaborative design and functional verification scenes of a complex system.

The invention relates to the technical field of Internet of Things, in particular to a multi-protocol conversion Internet of Things terminal communication method. Comprising the following steps: monitoring a communication data packet sent by Internet of Things terminal equipment in real time through a protocol sniffing module, and extracting a protocol feature field; matching the protocol feature field based on a dynamic protocol library, and identifying a native protocol type of the equipment; generating a protocol conversion strategy according to a target platform protocol demand, and calling a corresponding protocol analysis engine to perform semantic analysis on the data packet; the analyzed data content is reconstructed according to a target protocol format, and a unified communication interface packaging layer is added; and an optimal communication link is selected through an adaptive routing algorithm, and cross-protocol data transparent transmission is completed. The invention provides an integrated communication method capable of realizing efficient protocol dynamic identification, high-reliability semantic conversion, intelligent link optimization and safe packaging, so as to break through the technical barrier of multi-protocol Internet of Things terminal interconnection.

The invention discloses an intelligent data acquisition and conversion method based on a visual rule model, which comprises the following steps: S1, converting a configured rule chain into a data body in a JSON format by a visual rule modeling layer, analyzing the data body into an AST abstract syntax tree, and performing field mapping on the AST abstract syntax tree, the mapping fields are converted into executable codes through an intelligent conversion engine, and an execution plan is generated; s2, the intelligent acquisition layer pulls data from the heterogeneous data source, cleans the data and then allocates the data to an intelligent conversion engine for parallel processing; and S3, the intelligent conversion engine monitors the data consanguinity map and the field-level conversion path in real time, and sends a signal to the visual rule modeling layer or the intelligent acquisition layer according to the abnormal condition type so as to realize rule chain reconfiguration or data reacquisition. The method has the advantages that non-coding docking and real-time supervision of heterogeneous system data are realized, and the problems of dynamic adaptation, semantic understanding and data blood relationship interpretability in financial fund management are solved.

The invention provides a file conversion method and device, equipment and a program product, and the method comprises the steps: obtaining a to-be-converted file which is an image file or a portable document format file, i.e., a PDF file; identifying elements in the to-be-converted file, determining attributes of the identified elements based on an identification result, and performing logic layering on the identified elements; the elements comprise texts; generating a structured document based on the logical layering result and the attributes of the identification elements; and converting the structured document or the edited structured document into an editable file in a portable document format, so that the PDF can be edited. According to the method and the device, the automatic conversion from the image or the PDF to the editable PDF is realized, and in the conversion process, through the structured document generated in the middle, while the logic level of the document is improved, a user is supported to carry out instant editing, and the editing efficiency and convenience are improved.

The invention discloses an MRI-PET mode conversion method and system based on a generative adversarial network, and belongs to the technical field of artificial intelligence medical image generation. And the multi-scale structure representation injection module injects multi-scale anatomical prior information at different stages of the encoder, and overcomes the limitations of insufficient utilization of prior information and single injection scale. And the adaptive semantic residual fusion module adopts semantic attention guidance and double-branch attention weighting, adaptively fuses fine-grained local features and global context information, harmonizes the difference between the fine-grained local features and the global context information in an abstract level and a semantic category, and solves the problems of feature conflict and semantic fuzziness in a bottleneck region. The direction sensing space-frequency discriminator realizes multi-dimensional and fine-grained adversarial supervision through a space, frequency and local image block multi-branch collaborative discrimination mechanism, and improves the structural fidelity and spectrum authenticity of a synthetic image. And the generated image is superior to the existing method in indexes such as structural similarity and peak signal-to-noise ratio, and has higher clinical practical value.

The invention relates to the technical field of vehicle-mounted communication, and discloses an intelligent identification and conversion method for a heterogeneous vehicle-mounted connector protocol. According to the method, after a vehicle-mounted connector is connected, preset disturbance signal sequences are sequentially injected, response behaviors of the preset disturbance signal sequences in a signal channel are recorded, and a corresponding response sample set is constructed by analyzing response level, time delay and feedback data. The system further archives and encodes the sample set, generates a disturbance response behavior track containing state jump characteristics, and reasones a communication protocol type and a corresponding state driving factor through track comparison. On the basis, a field stable region and a boundary change mode are extracted, a protocol adaptive sampling rule is generated, then real-time segmentation and extraction are carried out on subsequent communication data, and a standardized data frame is constructed. Finally, data format reconstruction and semantic conversion among heterogeneous vehicle-mounted protocols are realized, and communication compatibility and data intercommunication capability among various connectors are improved.

The invention provides a method for converting an SAR image into an optical image based on a multi-mode conditional diffusion model. The method comprises the following steps: firstly, collecting SAR image and optical image data with a pairing relationship, and ensuring that a sample has diversity in the aspects of scene types, target structures and image styles; secondly, an image-language multi-modal model is adopted to generate text descriptions of SAR and optical images respectively, unified language descriptions are constructed through semantic analysis and fusion, and a light-SAR-text three-modal training sample set containing structure, semantic and style information is formed; according to the method, a denoising diffusion model guided by a multi-modal condition is designed and trained, a noise adding-denoising reconstruction process of an original optical image is taken as an optimization target, an SAR image, language description and a style image are introduced as multi-modal prompt conditions, and the expression of the generated image in the aspects of structure reduction, semantic alignment and style presentation is comprehensively guided.

The invention discloses a Tibetan language multi-dialect real-time semantic conversion method based on a cross-language BERT model, and the method comprises the following steps: S1, collecting original text corpora of each dialect of the Tibetan language, and constructing a standardized training corpus set; s2, performing parameter initialization on the mBERT model, and preliminarily training the mBERT model; s3, constructing a semantic modeling model, and performing fine adjustment on the semantic modeling model; s4, receiving to-be-converted text input, and encoding; s5, obtaining an intermediate semantic representation vector of the text through a semantic coding sub-module; s6, inputting the intermediate semantic representation vector into a semantic generation sub-module, and generating target text output; and S7, executing syntactic consistency correction and language fluency correction. According to the method, mBERT modeling and an adversarial optimization mechanism are fused, real-time semantic consistency conversion of multiple dialects of the Tibetan language is achieved, and the method has the advantages of being high in accuracy, high in robustness and low in response delay.

The invention relates to the technical field of model deployment, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a model service cross-framework request conversion method, device, equipment and medium, and the method comprises the steps: receiving a reasoning request based on a first model service framework, analyzing input data and a request path, and extracting input parameters and model identification information; determining a target request path of the second model service framework according to the model identification information, and converting the input parameters into a target input format; sending the converted input data to a target reasoning service, receiving a reasoning result and extracting an output parameter; and converting the output parameter into an output format of the first model service framework and returning the output format to the client. According to the method, an adaptation mechanism is introduced between the client and the target model service, so that calling logic which originally depends on a specific reasoning framework can be accessed to heterogeneous reasoning service without changing, and structural compatibility and communication unification of cross-platform deployment and calling of the model service are realized.

The invention discloses an industrial communication protocol conversion system and conversion method. The method comprises the following steps: acquiring data of an industrial communication source protocol; protocol conversion: converting the source protocol data into target protocol data according to a sequence of protocol analysis, standardized modeling and protocol repackaging; according to protocol analysis, original data transmitted from different hardware interfaces are analyzed according to corresponding industrial protocols; the standardized modeling parses the data into a data model in a unified format; protocol repackaging: recoding the data model into a data frame conforming to a communication specification according to a target protocol rule; and data output: pushing the obtained target protocol data to a third-party server through an MQTT protocol or directly storing the target protocol data in a database for a third-party system to access. Through three core technologies of protocol analysis, unified data modeling and protocol packaging, lossless conversion among different industrial protocols is realized, and an efficient and reliable data communication infrastructure is provided for an industrial automation system.

The invention belongs to the technical field of multi-modal intelligent data management, and discloses a multi-modal business data co-processing and intelligent conversion method, which comprises the steps of fusing multi-source data, coupling cross-modal time sequence association, detecting image-text semantic conflicts, and generating an RAG semantic unit and a space-time alignment semantic packet. Constructing a four-dimensional priority label to divide resource pool categories, dynamically allocating resource pool computing resources, generating a low-redundancy feature set, triggering a forced preemption strategy during resource competition, and synchronously generating a resource arbitration log; injecting a dynamic rule node according to a data type, executing confidence arbitration in combination with a conflict arbitration identifier, generating business decision content and generating a rule correction decision packet; constructing a full-link semantic thinking chain, synchronously capturing a system operation track, and aligning to generate a dual-channel tracing report; identifying residual sensitive fields, triggering hot update to load corresponding rules, quantizing effectiveness to generate compliance reports, reversely updating an industry rule base, and optimizing a resource allocation strategy.

The invention discloses a memory partition-based dynamic instruction conversion memory access conflict optimization method, which comprises the following steps of: allocating memory sub-regions for each computing core according to a load condition in an ARM (Advanced RISC Machines) many-core system to establish a first mapping relation table; when the executable file is loaded and executed in a dynamic instruction conversion mode, obtaining a second mapping relation table between the thread and the calculation core, and obtaining a thread and a memory address corresponding to a to-be-converted instruction of a memory access instruction class, obtaining a converted first ARM virtual address and a related first address space according to the first mapping relation table and the second mapping relation table, and determining an adjustment mode of the converted instruction according to the relation between the first ARM virtual address and the first address space to complete instruction conversion. The memory access conflict in the dynamic conversion process from the x86 instruction to the ARM instruction on an ARM many-core system is effectively reduced, and the execution performance of a program after dynamic instruction conversion is improved.

The invention provides an engineering drawing cross-format intelligent conversion method and system, and relates to the technical field of drawing format intelligent conversion, the method comprises the following steps: loading a first format engineering drawing, activating a multi-modal fusion structure identification engine, and obtaining auxiliary input information of a user; performing feature extraction according to the auxiliary input information, and establishing an identification attention constraint; after a multi-modal fusion structure recognition engine is initialized through recognition attention constraints, multi-modal content extraction of the first-format engineering drawing is executed, and a primitive extraction result is established; analyzing the auxiliary input information to obtain a second format; performing multi-dimensional format difference modeling on the first format and the second format; and performing format adaptation optimization on the primitive extraction result by using a difference compensation network, performing optimization verification by using the bound semantic identifier, and establishing a second format engineering drawing. Through the method and the device, the technical effect of improving the drawing structure reduction degree, the semantic recognition accuracy and the visual style matching degree can be achieved.

The invention relates to the technical field of computer programming, and discloses a Go language structural body automatic conversion method and system. The method comprises the following steps: firstly, receiving definitions of a source structure body and a target structure body as input, automatically matching fields of the two structure bodies through a field mapping rule engine according to a preset rule, and generating a field mapping relationship; then, based on the field mapping relation, a code generator is used for automatically generating a conversion code containing field assignment and type conversion logic; finally, the generated conversion codes are compiled into an executable conversion function, and the function can convert data in the source structure instance according to the field mapping relation during program running and fill the data into the target structure instance. According to the method, automation of structural body conversion is achieved, the problem that errors are prone to occurring when conversion codes are written manually is solved, and the maintainability of the codes is improved; the mode of reflection during operation is replaced by code generation, so that the conversion efficiency is remarkably improved; meanwhile, the method supports batch conversion and concurrent processing, and can meet the performance requirements of large-scale data processing.

The invention relates to a target detection method and system based on improved YOLOv11, and aims to improve the grabbing precision and operation efficiency of a mechanical arm in a complex environment. The method combines an augmented reality technology, a target detection algorithm, a depth sensing technology and an advanced motion control system, and specifically comprises the following steps that an RGB-D camera is used for collecting data; establishing a three-dimensional space model, and performing accurate alignment processing on the depth image and the color image; based on a mercuric chloride Atlas 200I DK A2 development board, in combination with an improved YOLOv11 model, efficient detection and identification of a target are realized; a target is converted from a pixel coordinate system to a camera coordinate system through a coordinate conversion method, and then the target is further mapped to the three-dimensional space position of a mechanical arm base coordinate system; finally, grabbing operation is completed through the elephant mechanical arm MyCobot 280M5. By combining RGB-D information, a deep learning detection algorithm and a three-dimensional coordinate conversion method, efficient and accurate target grabbing of the mechanical arm under the complex environment is achieved through the system, and the system has the advantages of being high in precision and real-time performance, easy to maintain, easy to integrate and deploy and the like.

The invention relates to a spacecraft FMEA application method based on a SysML model and a knowledge graph. The spacecraft FMEA application method comprises the steps that logic analysis is conducted on functional faults of a spacecraft, and a functional fault logic model and a fault mode basic element semantic model are obtained; according to the function fault logic model and the fault mode basic element semantic model, utilizing an SysML model tool to construct an FMEA digital model of the spacecraft; an XML metadata conversion method is adopted, and the FMEA digital model is converted into FMEA structured data; mapping concepts in the fault mode basic element semantic model and subclasses and relationships thereof into concepts, subclasses and relationships of an FMEA knowledge graph ontology model to obtain the FMEA knowledge graph ontology model; fMEA structural data and historical fault information are obtained through FMEA digital model conversion, an FMEA knowledge graph ontology model is instantiated, and an FMEA knowledge graph is established; based on the FMEA knowledge graph, spacecraft FMEA analysis is carried out, automatic statistical calculation of fault mode occurrence frequency and fault mode reduction measure recommendation are achieved, and the aim of reducing spacecraft design risks and on-orbit operation and maintenance risks is achieved.

A secret calculation device includes a memory and one or more processors coupled to the memory. The one or more processors are configured to: perform a first secret calculation operation on a first ciphertext obtained by encrypting input data according to a first homomorphic encryption method by using first homomorphic computation based on the first homomorphic encryption method to calculate a second ciphertext; convert, by using a ciphertext conversion method defined between the first homomorphic encryption method and a second homomorphic encryption method, the second ciphertext into a third ciphertext encrypted by the second homomorphic encryption method without decrypting the second ciphertext; and perform a second secret calculation operation on the third ciphertext by using second homomorphic computation based on the second homomorphic encryption method to calculate a fourth ciphertext.

The invention belongs to the technical field of industrial automation control, and discloses a PLC-based fault diagnosis method and device. The method comprises the following steps: acquiring data of target equipment acquired by a sensor through a data acquisition point; according to a preset conversion method, determining the data as a physical quantity numerical value of a corresponding physical quantity type, and storing the physical quantity numerical value to a circular buffer area with a preset time window length; determining a period according to a preset time length, and determining a plurality of characteristic values of the physical quantity values in the circular buffer area based on the data acquisition points; inputting the plurality of feature values into a pre-trained decision tree fault recognition model, and outputting a fault recognition result of the target equipment, and according to the target fault type label of the fault identification result, triggering an alarm signal, and recording a fault log of the target equipment. Whether the equipment breaks down or not can be identified based on the data during the operation of the equipment, and the fault type can be determined.

The invention discloses a skin image super-resolution conversion system and a skin image super-resolution conversion method based on deep learning, and relates to the technical field of image super-resolution conversion, and the system evaluates whether illumination is qualified or not by collecting skin near-illumination and analyzing illumination conditions of the skin near-illumination, establishing a light environment analysis model and calculating a light influence coefficient. The method comprises the following steps: preprocessing qualified images, including color space conversion, histogram equalization, edge detection and denoising; thirdly, evaluating the quality of image pixels, calculating a plurality of pixel features, judging whether the pixel features conform to the standard or not, and performing optimization through strategies such as sharpening and de-noising; a multi-scale feature extraction network and a self-attention mechanism are adopted to extract image features, a cross-modal mapping model is constructed, super-resolution conversion is performed through a generative adversarial network, and image details and definition are improved. And finally, monitoring the quality of the super-resolution image, evaluating the conversion effect, and adjusting the processing strategy according to the quality coefficient to ensure that the output image meets the medical diagnosis requirement.

The present application relates to a waveform file conversion method and apparatus for chip testing, and a computer readable storage medium. The method comprises: performing content analysis on a waveform file to be converted, to acquire period information, pin information, and state information of pins in each period of said waveform file; generating a PAT-format digital signal waveform file; and on the basis of a mapping relationship between pins and ATE channels, converting the PAT-format digital signal waveform file into a binary vector file for ATE testing. A waveform file in any format is converted into a PAT-format digital signal waveform file which uses periods as rows and pins as columns, so that each row vector not only contains current time information but also contains a logical state of each pin; in addition, to facilitate understanding, the PAT-format digital signal waveform file is further converted into a binary vector for ATE testing, so that efficient processing of waveform files in a plurality of different formats can be realized, thereby improving the universality of conversion.

The invention discloses a matrix acceleration and conversion method and system supporting variable blocks, and relates to the technical field of big data processing and GPU parallel computing. The method comprises the following steps: acquiring multi-source heterogeneous data, and generating a numerical matrix with uniform row and column dimensions through data type identification, field semantic analysis and cross-format standardized conversion; dynamically calculating an optimal partitioning threshold value based on matrix sparseness, row and column sizes and GPU video memory bandwidth parameters, adopting a full partitioning strategy for the dense region according to a row and column ratio and entropy analysis, and implementing a compression storage partitioning strategy for the sparse region; a GPU multi-level cache system is used for distributing heterogeneous blocks, a CUDA kernel function is automatically selected in combination with block density, and parallel computing is achieved through a thread bundle dynamic scheduling mechanism. According to the method, the problems of low heterogeneous data fusion efficiency, video memory bandwidth waste and unbalanced calculation load are solved, and the matrix operation speed is remarkably improved.

The invention discloses an AI scene conversion method and system based on visitor behaviors. According to the method, visitor behavior data are collected in real time through a multi-mode sensor network, and a UWB positioning base station, a millimeter wave radar module, a high-definition camera, a microphone array and a pressure sensor matrix are included. The system adopts an edge computing architecture, is equipped with an NVIDIAJetsonAGXXavier processor and a 512-core VoltaGPU (Graphics Processing Unit), and operates three core AI models to realize intelligent analysis. The LSTM residence time prediction model adopts a three-layer bidirectional architecture, each layer is configured with 256 hidden units, and visitor behavior sequences within 30 minutes are analyzed to predict the residence time. The Transform interaction intention recognition model comprises six layers of encoders and eight multi-head attention modules, and visual, voice and behavior three-mode characteristics are fused to judge visitor interaction requirements. According to the technical scheme, self-adaptive adjustment of the exhibition hall environment is achieved, the visitor satisfaction degree is improved by 28%, the operation cost is reduced by 15%, and important technical support is provided for intelligent transformation of modern exhibition places.

The invention discloses a remote sensing image ground object sample conversion method based on a diffusion model, and the method comprises the following steps: obtaining remote sensing data, and carrying out the preprocessing; constructing a neural network; the neural network takes a stable diffusion model as a basic network, and comprises a picture perception compression module, a potential diffusion model module and a seasonal invariant feature learning module. Training the neural network in stages to obtain a final neural network; and outputting a season conversion result. According to the method, a stable diffusion model is used as a basic generation model, and remote sensing images in different seasons are generated by using the strong generation capability of the diffusion model. According to the method, the seasonal invariant feature extraction network is constructed, the features of related ground features in different seasons are fully considered in the seasonal conversion process, and enough remote sensing ground feature information is ensured. According to the method, on the basis of a stable diffusion model, a domain-adaptive classifier is additionally added in a transfer learning mode to restrain the quality of a generated image, so that a remote sensing image sample with better quality is generated.

The invention provides a structured data representation method, a structured data conversion method, a structured data scheduling method, a structured data conversion device, a structured data scheduling device, a structured data conversion system, a medium and electronic equipment. The representation method comprises the following steps: responding to an operation of dragging a module from a module library to a visual editing canvas by a user to form a node corresponding to the dragged module; responding to the operation that a user establishes visual connection between input ports and output ports of different nodes on the canvas to form edges, wherein the edges define a data flow and calculation dependency relationship between the nodes; a structured data representation of a machine-readable AI model structure is generated based on all the nodes on the canvas and user-configured parameters thereof, and a set of edges defining the edge interconnection relationship. According to the representation method, conversion from graphical representation to structured data of the artificial intelligence model is completed.

The invention relates to the field of underground pipe gallery space detection, in particular to an underground pipe gallery detection method based on unmanned aerial vehicle point cloud scanning, and the method comprises the steps: 1, arranging a reflector plate, and collecting the three-dimensional coordinates of the reflector plate in an absolute coordinate system; 2, performing three-dimensional laser scanning when the unmanned aerial vehicle flies, performing point cloud acquisition at the entrance of the underground pipe gallery, acquiring spatial point cloud data by a laser radar carried by the unmanned aerial vehicle, and recording environment image information by a visual sensor; 3, establishing a relative coordinate system associated with spatial features, and carrying out point cloud SLAM calculation; step 4, calculating overground and underground continuous spatial point clouds based on an SLAM algorithm, and extracting three-dimensional coordinates in the scanning point clouds; 5, converting the point cloud in the relative coordinate system to an absolute coordinate system by using a Rodrigue matrix conversion method; and step 6, acquiring real spatial position information of the underground pipe gallery. The real spatial position of the underground pipe gallery in an absolute coordinate system is obtained through continuous scanning of the unmanned aerial vehicle in combination with ground GNSS measurement.

The invention discloses a PDF document format conversion method, a storage medium and computer equipment. The format conversion method of the PDF document comprises the following steps: analyzing an original PDF document based on a preset document analysis algorithm to extract a plurality of original content blocks of the original PDF document; based on the structure type of each original content block, extracting document content and layout features included in the corresponding original content block; and performing semantic conversion on the document content and the layout features based on a pre-trained large model to obtain a plurality of target content blocks in a preset target format, and arranging each target content block in the target format document based on the position information of the original content block corresponding to each target content block. By means of the method, the content of the PDF document can be efficiently and accurately analyzed and converted into the corresponding target format document, the target format document can keep the document content and layout characteristics of the original PDF document, and the integrity and readability of the converted document are effectively improved.

The invention provides an execution feedback and iterative optimization-based natural language to SQL (Structured Query Language) conversion method. The method comprises the following steps of: generating a plurality of candidate SQL statements; performing parallel execution on the candidate SQL statements, and capturing grammar error information in execution results and line number classification numbers of different candidate SQL return results; generating an error pattern abstract based on the grammar error information, converting the error pattern abstract into a natural language constraint instruction, and injecting the natural language constraint instruction into a prompt word generated in the next round; when the line number classification number is smaller than or equal to a preset threshold value, a self-consistency voting mechanism is adopted, a candidate SQL set with the largest number of returned same lines is selected, and the SQL with the shortest execution time is selected from the set to serve as final output. According to the method, the accuracy, robustness and execution efficiency of conversion from the natural language to the SQL can be remarkably improved by executing a feedback-driven multi-round iterative optimization mechanism.