351results about How to "Ensure completeness" patented technology

The invention discloses a heat supply network pipeline hydraulic and thermal working condition abnormal fluctuation intelligent identification and fault early warning method, and relates to the technical field of computer processing, and the method comprises the following steps: S01, collecting hydraulic and thermal parameters of a heat supply network pipeline in real time; s02, acquiring abnormal data of the time sequence sample set based on the thermodynamic mechanism model, detecting and repairing the abnormal data; s03, extracting characteristic parameters capable of reflecting abnormal fluctuation of the hydraulic and thermal working conditions of the heat supply network pipeline from the repaired time sequence sample set; s04, establishing an abnormal fluctuation identification model by utilizing an intelligent algorithm, and judging whether the hydraulic and thermal working conditions of the heat supply network pipeline have abnormal fluctuation or not; and S05, when the abnormal fluctuation is identified, giving a grading early warning mechanism and risk prevention and control suggestions of the damage degree grade of the associated pipeline. According to the invention, by fusing multi-source data and an intelligent algorithm, full-chain intelligent management of the hydraulic and thermal working conditions of the heat supply network pipeline from abnormal accurate sensing, fault intelligent diagnosis to risk active early warning is realized.

The invention discloses a connecting piece welding quality detection method and system, a medium and electronic equipment, and relates to the technical field of industrial automation and machine vision. The method specifically comprises the following steps: controlling a moving mechanism carrying an imaging device to move to a detection position, and obtaining a to-be-detected image containing a connecting sheet; inputting the to-be-detected image into a pre-trained target detection model, and obtaining a prediction bounding box of a connection sheet region; performing size expansion based on the prediction bounding box, and generating a region of interest covering the deformation region of the connecting sheet; in the region of interest, extracting a target feature point coordinate of a connecting sheet and a centroid coordinate of a welding spot by using an image processing algorithm; and calculating a distance value between the coordinates of the target feature point and the coordinates of the mass center of the welding spot, and judging the welding quality of the connecting piece based on the distance value. The invention aims to realize high-precision and high-efficiency welding quality detection of the connecting piece under the conditions of complex background interference and uncertain target form.

The invention relates to the technical field of computer vision, and discloses a camouflage target detection method based on cascade frequency domain perception and refinement feature guidance, comprising the following steps: a computer device obtains a camouflage image and extracts a multi-scale feature map; performing frequency domain separation by using a frequency domain sensing module, generating low-frequency and high-frequency characteristic sub-bands, and performing cross-hierarchy aggregation from bottom to top through a cascaded frequency domain sensor to obtain frequency fusion characteristics; receiving the frequency fusion features from top to bottom by using an aggregation guide decoding unit, performing guide refinement in combination with the deep decoding features, adapting to irregular edges by using a variable kernel convolution strategy in the refinement process, and generating a prediction feature map in combination with a partial convolution strategy; and outputting the predicted feature map as a detection result. According to the invention, through frequency domain information complementation and dynamic convolution refinement, the problem of difficult detection caused by highly similar texture of the camouflage target and the background is solved, and the positioning accuracy and edge segmentation integrity of the camouflage target are improved.

The invention relates to the technical field of batteries, and discloses a battery self-adaptive safety protection method and system based on data driving, and the method comprises the steps: extracting parameter features of a transient segment, a steady segment and a fluctuation segment after segmentation, constructing a comprehensive feature matrix based on the parameter features, all environmental factors of the battery are screened based on the environmental influence model, the working condition of the battery is determined based on the working condition identification model, the environmental influence factors and the state feature sequence, the working condition is verified based on historical data, and a corresponding monitoring strategy is selected based on the verified working condition or the re-determined working condition. And the acquisition interval is adjusted based on the monitoring strategy, the risk direction of the battery is determined based on the relationship between the standard state feature sequence and the state feature sequence, and whether power-off protection is performed on the battery is judged based on the association relationship of all the risk directions. According to the invention, through data driving of historical data, the flexibility and reliability of battery safety protection are ensured.

The invention discloses a discrete manufacturing multi-view BOM dynamic derivation method and system based on a tree-shaped network structure, and belongs to the technical field of intelligent manufacturing and product data management. The method comprises the following steps: constructing a tree body as a single product data source, and storing a hierarchical relationship, attribute definition and engineering semantics of a product in a tree structure; based on the tree body, dynamically matching derivation rules through a mesh mapping relation library; executing rule matching, semantic verification, structure reconstruction and dictionary verification by utilizing a dynamic view derivation engine to generate a target BOM view; semantic constraint and verification are carried out on the derivation process through a BOM data dictionary module; and realizing access control and audit based on role, scene and data sensitivity through the security and authority control module. According to the method, a new generation of xBOM dynamic derivation architecture which takes a tree form as an ontology, takes a net form as mapping, is configurable in rule and is strong in data specification is constructed; the fundamental problem of discrete manufacturing multi-view BOM management is solved, and engineering rationality and service flexibility are both achieved.

The invention discloses a container storage processing method and system based on TIPU, a terminal and a medium, and relates to the technical field of storage virtualization, and the technical scheme is characterized in that storage configuration information of NFS shared file information is encrypted; the method comprises the following steps of: creating a Configumap resource; a persistent storage declaration PVC is created, and a TIPU gateway is triggered to dynamically create virtiofs equipment; a POD is created, and virtiofs equipment is mounted to a container in the POD; the Configumap resource is analyzed, so that a first secret key fragment is injected into the virtiofs file system; and combining to form a complete secret key so as to decrypt the file information. According to the invention, the POD in the Kubernetes environment can be ensured to perceive and directly access the storage resources managed by the TIPU in real time, and efficient and dynamic unloading of the storage load is realized through seamless integration with the TIPU platform.

The invention discloses a causal knowledge graph construction and question-answering system based on retrieval enhancement generation and a large language model. The system comprises a multi-source heterogeneous knowledge base construction module, a retrieval enhancement generation module, a named entity recognition and causal triple extraction module and a knowledge fusion reasoning module, according to the method, a high-quality named entity annotation data set and a causal data set are constructed, and normalization and integrity of input knowledge are guaranteed; a mixed retrieval strategy (keyword + vector + sparse embedding) is provided, and the evidence coverage rate and recall precision are improved. Under RAG driving, LLM is combined, two rounds of causal triple extraction are achieved, and the causal relationship coverage degree and the direction judgment confidence degree are improved; a conflict detection and atlas fusion mechanism is designed to ensure the unity and consistency of new and old causal knowledge; a question answering system based on a causal atlas is also established, multi-hop causal reasoning is supported, and answers with controllable credibility and explainable are output.

The invention relates to the technical field of tobacco plant identification and counting, and discloses an unmanned aerial vehicle tobacco plant identification and counting method and system based on improved deep learning, which combines technologies of image classification, image segmentation, feature point extraction, machine learning, target detection and the like with refined intelligent identification of tobacco. According to the method, high-precision image recognition and analysis are realized by utilizing the efficient feature extraction and calculation capability of the method, point plant counting can be accurately performed, the efficiency and precision of agricultural production are greatly improved, the problem of calculation complexity of a traditional method in large-scale high-resolution image processing is solved, and powerful technical support is provided for precision agriculture. Therefore, the invention provides an unmanned aerial vehicle tobacco plant identification and counting method based on improved deep learning target detection and multi-dimensional post-processing, and high-precision and robust detection and counting of tobacco plants in large-scale tobacco field orthoimages are realized by constructing a three-stage assembly line of image preprocessing, improved target detection and multi-dimensional post-processing.

The invention discloses a multi-modal fusion fact verification method and system for reducing large model medical illusion. The method comprises the following steps: constructing a high-quality multi-modal medical data set; a fusion feature is obtained on the basis of a multi-mode Transform architecture; constructing a fact verification module, executing four verification steps, including semantic layer verification, logic layer verification, evidence layer verification and time sequence layer verification, on the basis of the fact verification module and the fusion features, endowing the preprocessed multi-modal medical data and each verification step with a dynamic confidence score, and determining whether the multi-modal medical data and the verification steps are matched or not on the basis of the confidence score and a preset threshold value. And medical illusion is corrected. According to the invention, cross validation and fact verification of multi-modal data are carried out, and a multi-level and all-around hallucination detection and correction system is constructed, so that the hallucination rate of a large model in the medical field is reduced fundamentally.

This invention belongs to the field of integrated circuit design technology, specifically relating to a method for optimizing integrated circuit design data. This method first constructs a "tool-layer-semantic" data acquisition architecture, integrating multi-channel parallel acquisition units to collect heterogeneous data from multiple sources. Compliance pre-verification is performed through hardware logic circuits and a verification rule base, extracting key data identifiers. A three-level incremental compression strategy based on key data priority, combined with an intelligent index integrating multi-dimensional indexes, is used for compression and storage. The stored data undergoes multi-dimensional optimization, including semantic deduplication, priority optimization, format conversion, and source quality backtracking, forming a unified format optimized semantic. Through an interactive interface, it is adapted and interacted with electronic design automation tools. After consistency verification and performance testing feedback loops, the final design data and full-process traceability information are obtained and stored. This invention achieves unified management, efficient compression, intelligent optimization, and full-process traceability of design data.

The application discloses a kind of digital culture intelligent resource library construction method and system, belong to digital cultural resource management technical field.The method includes: multi-source heterogeneous layered collection (six categories of resources are synchronously collected by multimodal terminal, high-precision detail capture), adaptive heterogeneous resource standardization processing (personalized algorithm, independent DCIR format), double-branch attention intelligent classification association (accurate classification, mining internal correlation), distributed encryption storage and intelligent indexing (safe and traceable, fast retrieval), dynamic self-optimization and intelligent updating (model incremental updating, resource real-time replenishment).The system, device and medium correspond to realize the above method.The present application completely avoids prior art, realizes the high-precision, intelligent, safe and dynamic construction of digital cultural resources, and improves the efficiency of resource protection and utilization.

This invention discloses a persistent data storage system, method, terminal, and medium. The system includes a secure virtual block device, an IDE module, and a secure storage operation module. The secure virtual block device resides in a trusted execution environment (TEE), providing a standardized block storage interface for the upper-layer file system and is compatible with the formatting and mounting of general file systems. The secure virtual block device stores its own secure metadata in a secure superblock. The IDE module resides in the TEE and is connected to the secure virtual block device, used to protect the confidentiality and integrity of data through encryption algorithms. The secure storage operation module resides in the rich execution environment and is connected to the IDE module, used to implement specific storage operations. This invention enables persistent data storage, and its combination with a TEE environment ensures the confidentiality, integrity, crash consistency, and freshness of data stored persistently in the TEE.

The embodiment of the specification discloses a data transfer quota checking method and device, a storage medium and a terminal. In response to a transfer quota checking request for a target data transfer task, a plurality of resource transfer data corresponding to the target data transfer task are collected. A target transfer quota checking strategy is determined according to the total number of all resource transfer data. The target transfer quota checking strategy is used to calculate each resource transfer data, and a transfer quota difference value between inflow data and outflow data in each resource transfer data is obtained. The transfer quota difference value is used to determine a transfer quota checking result of the target data transfer task. The related resource transfer data is collected with the data transfer task as an anchor point, the quota checking is avoided from the perspective of the main body, and the abnormal positioning accuracy is improved. The dynamic selection of the calculation mode based on the transfer quota checking strategy can adapt to data transfer tasks of different sizes and avoid the performance bottleneck caused by the fixed strategy.

The invention provides a wall recognition method and a storage medium, and the method comprises the steps: dividing an indoor building region and a peripheral region, and processing the proportional scale information and wall feature information, so as to achieve the precise separation of the indoor building region and a peripheral irrelevant region, and prevent the interference information of the peripheral region from affecting the wall recognition process; automatic and accurate acquisition of the scale is realized by using natural reference scale features generally contained in a peripheral area, and extra manual labeling or intervention is not needed; vectorization data and thickness parameters of the wall are extracted in a targeted mode, and the integrity and accuracy of wall feature information are ensured; according to the method, the identification result of the wall is jointly determined based on the proportional scale information, the vectorized data and the thickness parameters, so that fusion application of multi-dimensional key information is realized, dimensions, forms and physical attributes required by wall identification are comprehensively covered, and the identification reliability in a complex environment is improved.

This invention discloses an automatic verification and risk warning method for international trade information based on smart ports. It collects multi-source document data, including customs declarations, bills of lading, transportation routes, payment vouchers, and certificates of origin, and converts them into a standardized format through field mapping rules. A cross-document verification rule base is established, and rule matching is performed on key fields such as invoice amount and declared price, and bill of lading quantity and packing list, using a weighted scoring mechanism to calculate consistency scores. A spatiotemporal graph of trade behavior is constructed, and risk characteristics are statistically analyzed using an improved random walk algorithm, outputting a dynamic risk feature matrix. This matrix is ​​matched with a historical risk case database, and the probability of risks such as smuggling, false declarations, and tax evasion is calculated in conjunction with regional risk benchmarks. Risk thresholds are dynamically adjusted based on port clearance flow, triggering adaptive warnings and generating warning information including warning level, risk type, and inspection measures, thereby improving the intelligence level of port supervision.

The present application relates to data push method technical field, especially a kind of data push method, by mirror index tree engine, the engine adopts hierarchical index tree to carry out hierarchical management to rule, ensure that even under high data volume, it can realize fast, accurate rule matching, improve the efficiency of data push.For the potential problems caused by the dynamic modification of rule, mirror index tree engine also contains shadow hierarchical index tree, when rule is added, deleted or modified, reduce the interference to the ongoing rule matching task, ensure the stability and response speed of system, effectively reduce the negative influence of traditional rule engine on system performance when the number of rules grows and frequently changes.

The utility model provides a wind tunnel probe test coordinate frame device, which belongs to the field of wind tunnel tests and comprises two fixing plates, an adjusting block is provided with a coordinate displacement assembly capable of flexibly changing the coordinate position of a wind tunnel probe, and the adjusting block is provided with an adjusting block. The coordinate displacement assembly is provided with a rotating assembly which is used for rotating the wind tunnel probe so that the wind tunnel probe can adapt to the measurement requirement of the non-direct wind direction. According to the utility model, through the coordinate displacement assembly and the rotating assembly, during use, the position of the probe adjusting block is adjusted through rotation of the second screw rod, so that the probe adjusting block slides and displaces in the second sliding adjusting block, thereby enlarging the detectable range of the wind tunnel probe; the probe can be flexibly positioned to any position of the wind tunnel by adjusting the probe, and the angle of the probe clamp can be adjusted through the hinge design, so that the angle of the wind tunnel probe is changed.

The application discloses a mural reinforcing protective agent and a preparation method and application thereof, belongs to the technical field of cultural relic restoration, and can solve the technical problems that existing mural reinforcing materials cannot reinforce the lower layer structure of ancient murals and are prone to being attached to the surface of murals to cause the whitening of the surface pigments. The mural reinforcing protective agent comprises a bacterial cellulose carrier and Ba(OH)2 nanoparticles loaded on the bacterial cellulose carrier in situ. The application plays a reinforcing role on the lower layer structure of murals, realizes deep reinforcement of ancient murals, improves the protection effect, stability and durability, and the surface of the mural will not be whitened.

The invention relates to the technical field of pet medical information, and discloses a pet cloud medical record management method and system, and the method comprises the steps: collecting pet health data through a user terminal, and the pet health data comprises video and text description; uploading the video to a video processing module, automatically editing the video through the video processing module, and extracting key segments; the text description and the key segments are integrated, and a structured pet electronic medical record is generated through a medical record generation module; storing the pet electronic medical record to a cloud storage server, and storing the hash value of the pet electronic medical record to a block chain network; and according to a user instruction, safe sharing and multi-terminal synchronization of the pet electronic medical record are realized through the safety control module. According to the method, the problem that the pet cannot know the pain point of the illness state by himself / herself is solved, the behavior performance of the pet can be effectively recorded and traced, the burden of a pet owner for sorting data is relieved, and meanwhile, a more intuitive and complete diagnosis and treatment basis is provided for veterinarians.

The invention discloses a supply chain full-process traceability tracking method based on a block chain, and relates to the technical field of block chain and supply chain traceability, and the method comprises the specific steps: collecting supply chain node identity qualification basic information, and carrying out the structural processing; generating a unique initial dynamic chain code by using an encryption algorithm, and storing the unique initial dynamic chain code in a main chain after binding the unique initial dynamic chain code with qualification information hash; integrating the collected data into traceability data and checking node identities and chain codes; after the verification is passed, the traceability data is bound with a chain code for the second time, and a timestamp is added for chaining; updating the chain code according to the service state, adding an identifier to the old chain code and permanently storing the identifier; according to the invention, a rigorous node identity authentication system is constructed, so that the credibility of a traceability subject is guaranteed, and identity and qualification data can be traced and verified in the whole process; multiple terminals collect data cooperatively, completeness and timeliness are ensured through multiple associations, chain code dynamic update adapts to service changes, and data association continuity is ensured; the whole-process mechanism effectively solves the problem of traditional traceability data splitting.

The application discloses a kind of deep feature bayesian verification electronic evidence comparison method, it is related to electronic data forensics technical field, including, obtaining multi-source electronic evidence data, carries out structured reorganization and feature enhancement to multi-source electronic evidence data, generates enhanced evidence data;Enhanced evidence data is extracted using deep neural network to multi-dimensional deep feature, forms high-dimensional evidence feature vector;By bayesian inference method, the difference probability comparison of high-dimensional evidence feature vector is carried out, and the similarity confidence value is output;Similarity confidence value, deep evidence feature and high-dimensional evidence feature vector are block chain and traceability anchor are stored, and the output tamper-proof verification voucher;Based on tamper-proof verification voucher, similarity confidence value is carried out and is traced back to the grade mapping and verification of acceptance, and generates electronic evidence comparison report.The application is extracted by multidimensional deep feature and bayesian verification double mechanism cooperation, realizes the high-precision comparison of electronic evidence and uncertainty quantification.

The application discloses a business management and control system for the tobacco industry, and relates to the technical field of information management of the tobacco industry.The system comprises the following: a distributed event coordination bus adopts a multi-level priority queue of a publish-subscribe mode and a differentiated retry and time sequence order preserving mechanism to deliver events; a data acquisition and standardization module switches data processing modes according to network states, completes multi-source business data conflict arbitration and standardization conversion; a data state processing module checks events and updates business states based on a unified metadata model; a streaming monitoring and decision module identifies abnormalities through sliding window aggregation and dynamic threshold identification, generates abnormal instructions and business alarm events; a metadata synchronization module synchronizes metadata changes through hot loading; and a unified operation and maintenance response module executes automatic response arrangement based on events.The application realizes accurate, real-time and active management and control of the whole life cycle of the business of the tobacco industry, and improves the robustness, business efficiency, compliance level and risk response capability of the system.

The invention relates to the field of carbon emission prediction, in particular to a subway station construction stage carbon emission prediction method and system. Measuring and calculating carbon emission of the subway integrated station at different stages according to a carbon emission coefficient method, and performing data preprocessing to obtain carbon emission monitoring data; performing feature and target variable separation on the carbon emission monitoring data to obtain a feature matrix and a target variable vector; constructing a carbon emission prediction model to perform feature processing on the carbon emission monitoring data to obtain carbon emission feature data; the method comprises the following steps: constructing a carbon emission prediction integrated model based on LightGBM and XGBoost, identifying carbon emission characteristic data through the carbon emission prediction integrated model, and optimizing parameters in the LightGBM model and the XGBoost model by using grid search to obtain a carbon emission prediction amount in a subway station construction stage.

According to the test method for the high-low priority scheduling mechanism of the A664 switch, a minimum test environment is established, an A664 tester with four physical ports is used, and verification cases of high-priority data frame priority forwarding, priority data integrity and the number of high-low priority buffer areas in a congestion scene are designed. The integrity and accuracy of high-low priority transmission function verification are ensured, and the problem of verification test of the high-low priority working mechanism of the A664 network in the prior art is effectively solved. The method is high in universality and can be suitable for various A664 network devices and application scenes.

The application provides an air traffic control primary radar signal processing automatic test method and system, relates to the radar test field, and solves the technical problems of fragmentation of the prior art process, single evaluation dimension and poor traceability. The method specifically comprises: obtaining radar test data and structured test configuration; recursively analyzing the test configuration to generate an instruction set; loading data and executing signal replay according to the instructions to drive the software; synchronously collecting video data and processing results of each stage to form a multi-dimensional feature sequence; correlating and analyzing the two and extracting features to generate a holographic feature vector sequence; progressively comparing the sequence with a baseline determination model in layers to obtain a test determination result; integrating test item identifiers, video data, multi-dimensional feature sequences, holographic feature vector sequences and test determination results to record and generate a final test log report. An end-to-end automatic, multi-dimensional deep analysis and full-link evidence tracing radar automatic test method is realized.

The invention discloses an online reasoning-oriented verifiable data processing method, device and system, and the method comprises the steps that an evidence obtaining bypass module, an evidence bus and an evidence bin are configured in advance, the evidence obtaining bypass module triggers an abnormal event based on the evidence prompt in the reasoning process, and the evidence obtaining bypass module determines a time window including the triggering moment, an event header is created, an event abstract is generated and transmitted to an evidence bin through an evidence bus, and the event abstract comprises a decision track section and a calculation specification section at the triggering moment; collecting an evidence full text in the time window, dividing the evidence full text into a plurality of evidence blocks and configuring corresponding block indexes; sequentially transmitting the block index and the evidence block to an evidence bin through an evidence bus, generating an integrity verification field, and sending a sealing identifier; and the evidence bin performs integrity verification on the evidence packet after receiving the storage identifier, and if the verification is passed, the evidence packet is modified to be in a storage state. According to the method, verifiable evidence recording, chained storage and playback processing can be carried out on data of abnormal events in the online reasoning process.

The invention discloses an automatic part processing method based on CATIA and Excel. The method comprises the steps that a target part list is determined; a VBA program is utilized in Excel, part identifiers in the list are converted into combined search character strings capable of being recognized by CATIA, and special characters are automatically processed in the process; opening a general assembly in the CATIA, entering an advanced search interface, and importing the combined search character string; searching is executed to select 3D digital models corresponding to all target parts in batches at a time; and carrying out subsequent batch processing operation on the selected part set, wherein the subsequent batch processing operation comprises batch activation, color change, subset copying and creating or exporting. According to the method, whole-process automation from the list to three-dimensional model batch processing is achieved, manual operation of several hours is shortened to several minutes, the accuracy rate is 100%, the design efficiency and the data quality are greatly improved, and meanwhile based on a universal software platform, the implementation cost is low.

The utility model relates to LED screen production technical field, concretely disclose a kind of glue filling device for LED screen production, including fuselage, the inside of one side of fuselage is equipped with working chamber, still include: clamping buffer part and adjustable glue filling part, clamping buffer part is arranged in working chamber, the upper portion of adjustable glue filling part is installed on fuselage, lower portion extends to working chamber and is located above clamping buffer part;Clamping buffer part includes clamping assembly and buffer assembly, clamping assembly is arranged in the working chamber on fuselage, buffer assembly is provided with two, two buffer assemblies are mirror image and installed on clamping assembly.The utility model sets up clamping buffer part, solved the existing glue filling device when using, it is inconvenient to buffer the force of clamping fixed LED screen, the clamping moment of time clamping plate can exert greater force on screen, this can lead to LED screen distortion, appear scratch, even damage, affect the stable performance of subsequent glue filling work problem.

The invention discloses a preparation technology of a standard fructus perillae decoction, and relates to the technical field of traditional Chinese medicine decoction treatment.The preparation technology comprises the steps that fructus perillae decoction pieces are weighed and placed in a marmite to be decocted twice; comprising the following steps: Step 1: adding water which is 8 times of fructus perillae decoction pieces into a marmite for the first time, soaking for 30 minutes, boiling with strong fire, and decocting with slow fire for 30 minutes; step 2, adding water which is 6 times of the fructus perillae decoction pieces into the marmite for the second time, boiling with strong fire, and decocting with slow fire for 20 minutes; 3, filtering the liquid medicine with 200 meshes, and cooling the filtrate to room temperature in a cold water bath; step 4, combining the two filtrates, and carrying out vacuum concentration at the temperature of 50 DEG C; and Step 5, concentrating until the weight ratio of the medicinal material to the fluid extract is 1: 1, collecting a concentrated solution, and freeze-drying to obtain a sample. According to the method, the component transfer condition of the fructus perillae decoction pieces to the decoction is measured through objective quality standards, and the quality degree and effectiveness of the decoction are ensured.

The invention discloses a cloud server cipher machine-oriented virtual cipher machine fault drift method and system. The method comprises the following steps of: managing a VSM resource pool and issuing a platform drift key pair to each virtual cipher machine VSM; continuously monitoring the running state of each VSM, and when a fault is detected, packaging a migration packet of core data carried by the fault VSM; generating a symmetric key, packaging the migration packet into a digital envelope through a platform drift key, and generating a digital signature to obtain a key data drift packet; a target VSM is dynamically selected, a key data drift packet is received, a digital envelope is decrypted to obtain a migration packet, a digital signature is verified, the migration packet is transmitted to the target VSM, key assets and configuration information are thermally loaded, and data recovery is completed; unbinding the IP address of the source fault VSM and configuring the IP address to the target VSM in real time; performing secure erasure and initialization operation on the faulted VSM, and re-incorporating the faulted VSM into the resource pool; the system comprises a management center, an equipment management service and a virtual cipher machine.