54 results about "Data stream analysis" patented technology

The invention provides a sparse and dense state calculation method, device and equipment and a readable storage medium, and the method comprises the steps: receiving an input data stream, analyzing the feature information of the input data stream, and marking a sensitivity level for the data stream; according to the sensitivity level of the data stream, a corresponding calculation strategy is called, and the calculation strategy at least comprises secret state calculation, hybrid calculation and plaintext calculation; and executing a calculation strategy for calculation associated with the data stream according to the calculation strategy corresponding to the data stream in AI training and / or AI reasoning. Through the technical scheme of the specification, the sensitivity level is marked for the data stream, and the secret state, mixing or plaintext calculation strategy is dynamically matched, so that the overall calculation overhead is remarkably reduced while the high-sensitivity data security is ensured, and the AI training and reasoning efficiency is improved.

The invention discloses an engineering construction digital intelligent supervision management and control system, and relates to the technical field of engineering construction supervision. The system comprises an information acquisition module which is used for acquiring equipment state, image and environment data; the data fusion layer module is used for carrying out cleaning, alignment and Kalman filtering fusion on the multi-source data to generate a multi-modal construction data flow; a progress calculation and prediction unit, a deviation judgment unit and a digital modeling unit are arranged in the analysis monitoring module, and the analysis monitoring module is used for updating a progress state based on an XGBoost model, analyzing a node dependency relationship by using a graph neural network, calculating a deviation propagation risk coefficient and generating a hierarchical deviation signal in combination with three-level judgment conditions; and the management and control layer module executes hierarchical response according to the level of the deviation signal, and feeds back execution effect data to the front-end module to realize model optimization and closed-loop iteration. According to the invention, intelligent perception, risk prediction and adaptive management and control of the construction process are realized, and the real-time performance, accuracy and autonomy of project supervision are improved.

The invention is suitable for the field of vehicle-road cooperation, and provides a real-time intelligent data stream analysis and management method and system, and the method comprises the steps: obtaining multi-type data streams transmitted by a vehicle terminal, roadside equipment and a cloud platform in a vehicle-road cooperation scene; assigning priority levels to the multi-type data streams; carrying out parallel processing on the data streams with different priorities by adopting a self-adaptive analysis algorithm, and extracting environment perception features, vehicle state features and traffic event features; performing integrity and accuracy verification on the analyzed feature data through a real-time verification model, marking abnormal data and triggering a correction mechanism; according to the priority level and a data quality evaluation result, dynamically distributing, calculating and transmitting targeted resources, and pushing effective feature data to a corresponding decision node to support real-time response; priority transmission of key data is guaranteed through dynamic priority scheduling, abnormal data is verified and corrected in real time, and the problems of poor multi-source data fusion, key data delay and the like are effectively solved.

The invention provides an automatic object-level authorization vulnerability detection method based on a large language model, which comprises the following steps of: firstly, identifying a sensitive resource table through SQL (Structured Query Language) grammar analysis, and analyzing an operation path from a tracking request parameter to the sensitive resource table by utilizing a forward data stream; then, performing context sensing analysis on the database operation by adopting LLM to infer object-level sensitive operation; thirdly, recognizing program logic and a conditional protection mechanism in database query through static analysis, and recognizing custom verification logic such as method-level annotation and administrator permission in combination with an LLM and character string matching method; and finally, judging whether the sensitive operation lacks user-defined verification, ownership SQL constraint and access control conditions at the same time or not by integrating the mechanism, and generating a vulnerability report according to the judgment result. The invention further provides an automatic object-level authorization vulnerability detection system based on the large language model. Therefore, the detection precision can be effectively improved, and false alarm and missing alarm are remarkably reduced.

The invention relates to an industrial control system PLC control logic decompilation method and application thereof, and the method comprises the following steps: carrying out simulation execution on a PLC control application program, realizing memory relocation and variable initialization processing on the control application program, and extracting an executable instruction and corresponding data; converting the extracted executable instruction and data into a unified intermediate representation to obtain an IR instruction; on the basis of a heuristic data flow analysis algorithm, variables in the IR instruction sequence are identified, forward backtracking analysis is carried out starting from the output variables, the accurate dependency relationship between the variables is recorded and constructed step by step, and a complete variable dependency graph is formed; and based on the variable dependency graph, combining branch and loop structure information in a control program, and reconstructing a complete PLC control logic in a structured text language. Compared with the prior art, the method has the advantages that the positioning precision and the recovery efficiency of PLC control logic tampering are improved, and the safety and the maintenance efficiency of an industrial control system are remarkably improved.

This invention relates to the field of distributed streaming data processing and dynamic reconfiguration of cross-domain processing logic, and particularly to a real-time data stream analysis method and system for the international automotive market. The method includes: acquiring pure physical kinematic state data of the vehicle and pre-authorized original environmental data from a pre-defined system; calculating and generating a transient symmetric key; encrypting the original environmental data into ciphertext data and writing it into a memory buffer; inputting the kinematic data into a gradient descent optimization model, fine-tuning the feature transformation computation graph weight matrix and bias terms, and outputting a new rule domain matching operator path; reading the hardware clock cycle value and injecting it into the header of the ciphertext data tensor to generate a routing marker, guiding the data flow to the new operator path; after confirming the path switch, sending an overwrite command to replace the transient symmetric key with an all-zero physical state, blocking the decryption and restoration path of the remaining ciphertext data. This invention solves the problem of feature data leakage or computational interruption caused by differences in feature transformation standards during cross-regional driving.

Techniques for data flow analysis of cloud-based software applications are described. A first portion of source code of a software application is determined to obtain data from a data source identified by a first resource identifier of a cloud provider network, the determination based on a mapping of a first statement in the first portion of the source code to an application programming interface (API) call. A trace of a data flow from the first portion of the source code to a second portion of the source code is obtained. A data sink identified by a second resource identifier of the cloud provider network is determined based on a mapping of a second statement in the second portion of the source code to another API call. A result that includes an identification of a data flow from the data source identified by the first resource identifier to the data sink identified by the second resource identifier is generated.

The invention relates to the field of intelligent analysis, and particularly discloses a binary code-oriented static taint analysis system and method, which introduces an on-demand micro-symbol execution mechanism on the basis of basic data flow analysis through binary lifting and control flow diagram construction, that is, aiming at memory ambiguous nodes in a data dependence diagram, a micro-symbol execution mechanism is introduced on the basis of the basic data flow analysis; refined alias analysis is achieved through reverse instruction slicing and constraint solution, then an enhanced data dependency graph is generated, and path-sensitive taint propagation iteration is conducted on the basis to generate a final taint state table. In this way, complex dynamic memory addressing can be effectively recognized, taint chain breakage and misinformation explosion caused by ambiguity ignoring or excessive approximation are avoided, and resource allocation of an analysis engine is remarkably optimized.

The invention relates to the technical field of distributed streaming data processing and cross-domain processing logic dynamic reconstruction, in particular to a real-time data stream analysis method and system oriented to an automobile internationalization market, and the method comprises the steps: obtaining vehicle pure physical kinematics state data and preset system authorization original environment data; calculating and generating a transient symmetric key, encrypting original environment data into ciphertext data, and writing the ciphertext data into a memory buffer space; inputting kinematics data into a gradient descent optimization model, finely adjusting a feature conversion calculation graph weight matrix and an offset item, and outputting a new rule field matching operator path; reading a hardware clock period numerical value, injecting a ciphertext data tensor head to generate a routing mark, and guiding data to flow to a new operator path; and after path switching is confirmed, an overwriting instruction is sent to replace the transient symmetric key to be in an all-zero physical state, and a decryption restoration path of the residual ciphertext data is blocked. According to the method, the problem of feature data leakage or calculation interruption caused by feature conversion standard difference in cross-regional driving is solved.

The invention relates to the technical field of data stream processing, and discloses an intelligent decision command system based on data stream analysis, comprising: a corpus collection unit used for deploying a plurality of corpus collection agents used for collecting text data on each process node; the corpus database is used for storing the corpora in the database in a classified manner according to the correlation between the corpora and the decision-making scene; the decision-making scene unit is used for collecting decision-making demand text data input by a decision maker, the intelligent analysis unit takes the decision maker demand text data and all data related to the decision-making scene in the corpus database as input data of a decision-making model, and the decision-making output unit optimizes an anomaly detection model through incremental learning. Multi-dimensional information visualization display of first-line decision and central command is realized; the technical problems that in the prior art, the data island phenomenon is serious, and intelligent data analysis is lacked are solved.

The application discloses a high-position remote landslide disaster early warning method and system based on multi-source data fusion, and relates to the technical field of early warning. First, a hardware risk perception unit is deployed at the data acquisition front end of an early warning server, running data streams of each sensor data channel are obtained through the hardware risk perception unit, and multi-dimensional labels of each sensor data channel are synchronously acquired, then according to the running data streams, interruption risk index values of each sensor data channel are analyzed and obtained, the multi-dimensional labels are analyzed in real time, early warning results of each sensor data channel are marked, and then based on the interruption risk index values and the early warning results of each sensor data channel, the running mode of an early warning dispatcher of each sensor data channel is adjusted, based on the adjusted running mode of the early warning dispatcher, scheduling execution adjustment of each sensor data channel is carried out, and high-position remote landslide disaster early warning is carried out after the scheduling execution adjustment, thereby improving the reliability and timeliness of high-position remote landslide disaster early warning.

The invention relates to a misuse detection method and device of an encrypted API and terminal equipment. The method comprises the following steps: constructing an encrypted API knowledge graph; extracting an encrypted API file from the to-be-detected item file; respectively performing control flow analysis and data flow analysis on the encrypted API file to obtain function call information and static variable information of the encrypted API file; and inputting the encrypted API knowledge graph, the encrypted API file, the function call information and the static variable information into an encrypted API agent for encrypted API misuse detection to obtain a detection report of the to-be-detected item file. Through application of the method and the device, the problems of relatively high false alarm rate and missing report rate of an encrypted API misuse detection method depending on rules in related technologies are solved, and the technical effect of improving misuse detection accuracy of the encrypted API is realized.

A system and method for encrypted data compression with hardware management combines data compression techniques with real-time hardware optimization. The system receives an input data stream comprising data blocks, analyzes frequency distributions within the stream, and generates a conditioned data stream by identifying portions that deviate from target frequency distributions and applying conditioning rules. A hardware management layer continuously monitors the system resource utilization including processing latency, memory consumption, and field-programmable gate array (FPGA) resource usage during data processing operations. When resource utilization exceeds performance thresholds, the system generates hardware optimization plans and implements configuration changes including memory reallocation, processor adjustments, and FPGA resource management. The system applies Burrow's-Wheeler transforms (BWT) to improve data compressibility and creates error streams through XOR operations between original and conditioned data. The hardware management layer enables dynamic adaptation to changing processing demands while maintaining optimal system performance for data compression operations.

The invention discloses a data-driven extensible communication protocol analysis system, method and device and a medium, and relates to the technical field of communication protocol processing and software automation, and the system comprises a log preprocessing module which is used for forming a binary protocol frame to be analyzed; the frame analysis module is used for carrying out structured analysis on the binary protocol frame and comprises a protocol atom library used for storing and analyzing an atom analysis function of a basic data field; the atomic type mapping table is used for associating an atomic type identifier declared in a protocol description file with the atomic analytic function; and the data stream analysis and visualization module is used for generating visual output based on the data stream context formed in the frame analysis process. According to the invention, the analysis of the newly added protocol field only needs to register a new atomic function in the protocol atomic library without modifying the core architecture of the system, and the newly added protocol type only needs to establish a new frame structure mapping relation, so that the system can quickly adapt to the evolution requirements of various communication protocols.

The invention discloses a privacy routing forwarding method, device and system, and a storage medium, and introduces a homomorphic tree HTree structure to realize homomorphic routing query, so that a router carries out routing matching in a ciphertext space under the condition of not knowing a plaintext destination address, data flow analysis of a malicious router is prevented, and privacy forwarding of a user data packet is ensured. By adopting the technical scheme of the invention, the method is applied to the field of Internet, provides an encrypted routing query and forwarding function, and ensures the privacy of a data packet.

The invention discloses a multi-source time series data stream concept drift detection method based on dual inspection, and belongs to the field of data stream mining. The method comprises the following steps: acquiring a prediction result of a prediction model and a real observation value of multi-source time series data, and dividing a time window; judging whether the current window error obviously changes or not through first hypothesis testing; matching a window most similar to the current data from the historical data, calculating the error change degree between the two windows, and judging whether the prediction error of the current window obviously exceeds the prediction error of the similar historical window through secondary inspection; and if the two tests are both significant, determining that the concept drift occurs. Through double inspection and similar window comparison, the problem that the misjudgment rate is high in multi-source data flow concept drift detection in a traditional method is solved, the accuracy and robustness of drift detection are improved, and the method is suitable for dynamic data flow analysis in the fields of finance, industrial monitoring and the like.

The invention discloses a telemetering data stream analysis method and system based on a dynamic decoding rule base, and relates to the telemetering and data processing technology, and the method comprises the steps: constructing a dynamic decoding rule base and a tentative instruction set; according to a preset strategy or an adaptive sequence, sending a corresponding tentative instruction to the target data source, and collecting telemetering data; receiving and caching an original byte stream from a corresponding data source; for any sent heuristic instruction, in the original byte stream, according to a sending timestamp and a preset response delay window of the heuristic instruction, positioning a candidate response data block; based on the dynamic decoding rule base, generating a corresponding candidate decoding rule for any candidate response data block, and performing trial decoding by using the candidate decoding rule; and carrying out rationality evaluation so as to determine an applicable decoding rule of the telemetry data stream according to an evaluation result. The invention provides an effective mechanism for actively and systematically guiding and automatically generating trusted decoding rules.

The invention relates to the technical field of power system relay protection, in particular to a protection setting value digital checking method based on print data stream analysis, which comprises the following steps: collecting multi-source print data, and preprocessing the collected multi-source print data; an improved BERT-BiLSTM-CRF model is used for conducting intelligent analysis on the preprocessed multi-source printing data, and structured information of fixed value items, numerical values and units is extracted; on the basis of a pre-constructed checking rule base, multi-dimensional checking including numerical range checking, logic consistency checking, historical consistency checking and device matching checking is carried out on the analyzed constant value data, and checking grade outputting and deviation analysis are carried out; an adjustment suggestion is generated for the check alarm constant value through a reinforcement learning DQN algorithm, and check rules and model parameters are optimized in combination with a feedback mechanism; according to the method, the checking efficiency is improved by integrating multi-source data software acquisition, deep semantic analysis, multi-dimensional collaborative checking and closed-loop iterative optimization.

The application discloses a kind of fire-fighting Internet of Things remote monitoring method and system, by collecting environmental parameter to construct real-time monitoring sequence, using data stream analysis method accurately identifies abnormal point, and when threshold is triggered in abnormal, fusion mechanism is triggered, multidimensional index is integrated into vector, risk signal is extracted and grouped similar mode to judge hazard type, and then matching historical response record generates optimized response chain.The action sequence is determined by priority evaluation algorithm, and the disposal instruction is generated by combining system linkage judgment, and the execution effect is evaluated by simulating the application scenario, and finally the record and template are updated to form a closed-loop response mechanism.The application realizes intelligent management of the whole process from anomaly detection to accurate disposal through multi-index fusion and dynamic adjustment, significantly improves the accuracy and efficiency of environmental risk early warning and emergency response.

The invention relates to a method for analysing a data stream of data packets received via a communications network, a protocol analysis (DAP) according to the following steps:analysing the first packet (P0) of a batch to check whether the first packet (P0) announces the protocol of the next data packet (P1);analysing the next protocol or protocols Pi+n with n≥2 by repeating the steps of explicit detection classification (S10), session detection classification (S20), and deep packet inspection classification (S40) until at least one protocol, the identity of which is certain, is identified; and—associating a label with the data packets; and its implementation system.

The application provides a kind of multi-data stream processing method and device with equivalent index.The method comprises: determining corresponding equivalent indexable expression chain according to association analysis expression;Equivalent indexable expression chain is used to construct equivalent index expression graph;Equivalent index expression graph includes multiple graph nodes and the edge between graph nodes;Graph node includes the equivalent indexable expression corresponding to flow table, and the edge includes the association relationship of the equivalent indexable expression corresponding to the two graph nodes connected;Graph node in equivalent index expression graph is traversed, and the equivalent expression path of equivalent index expression graph is determined;Equivalent expression path is used as the index corresponding to flow table, and data stream is analyzed and processed based on the index of flow table.The method provided by the application can analyze the equivalent index of multi-data stream association analysis expression, to quickly realize the association analysis of multi-data stream, so as to improve the efficiency of multi-data stream analysis processing.

The application provides a data center machine room operation and maintenance anomaly detection method based on deep learning, comprising: for a low-dimensional feature set, using a long short-term memory network to analyze time series and capture subtle patterns in low-dimensional dynamic characteristics to obtain a behavior pattern sequence; extracting packet structure and time delay changes from an abnormal candidate set, using an isolation forest algorithm for secondary screening to determine a hidden abnormal behavior set; according to the abnormal pattern set, using a sliding window mechanism to analyze real-time data streams to determine whether abnormal patterns continue to appear to obtain an abnormal persistence score; according to the final abnormal detection result, updating the parameters of the behavior benchmark model, using an online learning mechanism to optimize the model to obtain an updated behavior benchmark model; for the updated behavior benchmark model, cyclically executing real-time data stream analysis to continuously capture subtle patterns to obtain continuous abnormal detection output.

The invention relates to the field of program dynamic analysis, and provides a data flow analysis device and equipment, and the device comprises a data receiving module which is used for obtaining instruction data and taint source information; the queue management module is used for storing the instruction data into an instruction queue; the instruction decoding module is used for acquiring one instruction from the instruction queue as a current instruction, decoding the current instruction and determining code semantics of the current instruction; the taint calculation module is used for determining the taint propagation rule of the corresponding execution circuit, and performing taint propagation calculation on the current instruction based on the execution circuit to obtain taint propagation data; the memory access interface is used for writing the taint propagation data into the taint record mapped by the memory; and the exception output module is used for outputting stain source information of the exception occurrence position. The problems that in the prior art, hardware-based taint propagation implementation only supports single-source taint analysis and is low in efficiency are solved, and high-performance multi-source dynamic taint propagation analysis based on a special hardware module is achieved.

The invention relates to the technical field of mobile operating system data processing, and discloses a self-adaptive SSE data stream analysis method and device based on a swan operating system, computer equipment and a storage medium. According to the method, a dynamic complexity score is generated by monitoring a system resource state in real time and analyzing SSE data flow protocol feature complexity; dynamically selecting a single-state machine mode, a dual-state machine cooperation mode or a multi-state machine assembly line mode based on a preset complexity threshold value and a memory utilization rate threshold value; and configuring a corresponding processing assembly line for analysis by utilizing a (at) ohos.buffer module of the swan mongolian system. According to the method, the problems of low memory efficiency, poor fault tolerance and insufficient analysis performance in the prior art are solved, and the data analysis efficiency and the system stability are remarkably improved.

The application discloses a gantry crane safety monitoring management system based on artificial intelligence, and belongs to the technical field of crane safety monitoring. The system comprises a data acquisition module, a video acquisition module, a wireless transmission module, an artificial intelligence analysis module and a remote monitoring terminal. The data acquisition module acquires crane operation data and standardizes and encapsulates the crane operation data, and the video acquisition module completes global video acquisition of work. Various types of data are encrypted and transmitted to the artificial intelligence analysis module through an industrial-grade wireless network bridge. The module relies on an edge intelligent gateway, calls a cloud multi-modal large model through a standard interface, carries out data flow analysis and video flow analysis, and outputs safety early warning and linkage control instructions in stages. The remote monitoring terminal realizes remote centralized management and control, data query and report generation. The application provides a more perfect and intelligent gantry crane safety monitoring management system based on artificial intelligence.

The application discloses a kind of satellite firmware program reverse analysis methods based on signature mechanism and taint analysis.The method is by identifying encryption function, decrypting firmware content and tracking data flow, analysis program logic and dependency, to clarify the implementation of authentication and communication mechanism, and restore normal code logic.Specifically including:1) by constructing encryption function signature database, using static analysis tool and signature library matching, quickly identify encryption or decryption function in firmware;2) using Bi-LSTM neural network, similarity calculation is carried out to target code and vulnerability template, so as to track data flow and analyze program logic;3) comprehensive encryption function identification and data flow tracking result, parse communication protocol and authentication mechanism, restore its implementation logic.The application overcomes the problem of low efficiency and insufficient automation of traditional reverse analysis method, can significantly improve the precision and efficiency of firmware reverse analysis.In particular, when dealing with complex data flow and custom function, the data flow in business logic can be effectively clarified, thereby providing security for satellite ground station equipment.

A compute system includes: a control circuit configured to: store a sensor data stream of a region of interest provided by optical sensors, analyze an overlap region of the sensor data stream to identify a feature line that is viewed by a first optical sensor and a second optical sensor of the optical sensors, calculate a correction distance, measured in pixels, for the first optical sensor by a position correction of the feature line in the overlap region, correct an optical distortion of the first optical sensor in the sensor data stream for the region of interest by applying the correction distance, and calculate GPS coordinates of the feature line for controlling a device.

This invention relates to the field of power battery monitoring technology, and specifically to a real-time streaming analysis method for power battery data based on Kalman filtering. The process includes the following steps: S1. Acquire real-time operating data of the battery, including cell voltage, current, temperature, and timestamp; S2. Construct the state-space equation of the battery based on a dual RC equivalent circuit model, and use the variable forgetting factor recursive least squares method to identify the model parameters online and update the coefficient matrix of the state-space equation in real time; S3. Set the initial parameters of the sunflower optimization algorithm, using the mean square error between the SOC estimate and the true value as the fitness function, and use the sunflower optimization algorithm to perform offline global optimization on the process noise covariance matrix Q and measurement noise covariance matrix R of the extended Kalman filter to obtain the optimal Q matrix and R matrix; S4. Extend the optimal Q matrix and R matrix obtained from the offline optimization into the Kalman filter, and perform SOC prediction based on the current current and historical state during vehicle operation, and perform Kalman filter correction in combination with the measured terminal voltage to output the real-time SOC estimate; S5. Based on the real-time SOC estimate and operating data, generate analysis results for range prediction, thermal management, and safety warning.

Systems and methods are described for determining task specific data streams for controlling a repositionable structure. The system may include one or more repositionable structures operatively coupled to one or more instruments, and a control system operably coupled to the one or more repositionable structures, the control system configured to (i) receive a plurality of data streams from one or more data sources; (ii) analyze the data streams to identify tasks to be performed by the repositionable structures, instruments, or one or more auxiliaries; (iii) determine, based on the tasks to be performed and via a modality selection machine learning model, a set of task-specific data streams; (iv) generate, based on the set of task-specific data streams and via a robotic action machine learning model, action tokens for controlling the one or more repositionable structures; and (v) control the repositionable structures based upon the one or more action tokens.

The invention relates to the technical field of structure safety monitoring, and discloses an anti-seismic detection method based on an intelligent retractable canopy. The method comprises the following steps: establishing a detection task for a target contraction canopy, wherein the task comprises unique identification information; extracting a historical environment data set of the target canopy according to the identification information; generating a dynamic environment simulation sequence based on the data set to simulate future environment state change; constructing a canopy structure point cloud model according to the sequence, wherein the model reflects physical structure forms of the canopy in different environment states; applying a simulated seismic load to the point cloud model to obtain a structural response data flow of the point cloud model; analyzing the data stream, and identifying a key deformation area and a stress concentration point; and calculating an anti-seismic stability index of the canopy structure, and comparing the anti-seismic stability index with a preset threshold value to generate a detection report. According to the invention, through dynamic environment simulation and variable form point cloud modeling, the accuracy and practicability of anti-seismic detection are improved.