19 results about "Grey box" patented technology

The invention relates to the technical field of program testing, in particular to a seed scheduling method and system for multi-objective optimization in grey box fuzzy testing, and the method comprises the steps: dividing related indexes of seeds into state related indexes and state independent indexes; modeling a priority ranking problem of the seeds into a multi-objective optimization problem according to state related indexes, calculating Pareto leading edges of the seeds based on a non-dominated ranking algorithm, and screening out the seeds with high priorities; selecting seeds according to a seed selection basic principle; determining the weight of each index by using an analytic hierarchy process according to the state-independent indexes, evaluating the quality of the seeds through an approximate ideal solution algorithm, and distributing reasonable energy for the seeds; and based on the priority and the energy distribution result, selecting the seeds for mutation to generate a new test case. According to the method, seed priority division and energy distribution are optimized, so that the seed scheduling efficiency in grey-box fuzzy testing is remarkably improved, and accurate testing is realized at a higher code coverage rate and a faster vulnerability discovery speed.

The invention discloses an intelligent automobile fault identification method based on a minimum grey box, and the method comprises the steps: obtaining intelligent automobile fault injection data with fault classification labels, constructing a Bayesian network through employing a structure learning method, forming a directed acyclic graph which reflects a variable causal dependence relation, mapping the directed acyclic graph into a fault tree model, and carrying out the recognition of the fault tree model. The method comprises the following steps: establishing a corresponding structure of a fault event and a logic relationship, determining a minimum cut subset through a fault tree model, screening in combination with a Markov blanket, extracting a minimum and sufficient key variable set, forming a minimum grey box, performing parameter learning by taking a directed acyclic graph as structural input and in combination with fault injection data, and constructing a dynamic Bayesian network. And inputting the minimum grey box variable set into the network to realize the detection of the fault state of the intelligent automobile and the judgment of the fault type. According to the method, through minimum grey box construction and dynamic Bayesian network reasoning, the problems of large data redundancy, strong black box performance and insufficient causal interpretation in the prior art are solved.

The invention discloses an array antenna pattern simulation method based on a U-Net network and a grey box model, and the method comprises the implementation steps: constructing a network based on a U-Net architecture, taking an ideal array pattern without considering the coupling influence between antenna units as the input, taking a real array pattern considering the coupling influence between the antenna units as the output, and carrying out the simulation of the array antenna pattern based on the U-Net network and the grey box model. And mapping from an ideal array directional diagram to a real array directional diagram is realized. Through fusion of a numerical method and deep learning, the defects that a traditional numerical method is low in simulation efficiency and a black box model based on a deep learning method is insufficient in simulation precision are overcome. According to the method, high precision can be kept in simulation of array antenna directional diagrams of different structural sizes, the simulation speed of the array directional diagrams is remarkably increased, the simulation memory requirement is reduced, the design period of the array antennas is greatly shortened, and the method can better meet the requirement for efficient iterative optimization in engineering application.

In exemplary embodiments, methods and systems are provided that include one or more vehicle sensors and a vehicle processor. The one or more sensors are configured to acquire sensor data about the vehicle's operation. The processor is coupled to the one or more sensors and is configured to enable at least the input of the sensor data into a grey-box model, including a neural network model for the forces acting on the vehicle and a physics-based model for the vehicle's body movements; the determination of a vehicle trajectory using the grey-box model; the determination of a control action for the vehicle based on the grey-box model using the sensor data; and the execution of the control action for the vehicle in accordance with the instructions provided by the processor.

The invention discloses a vulnerability feature-guided directional grey box fuzzy testing method and system. The method comprises the following steps of: extracting dangerous code blocks from input source codes of a tested program based on given vulnerability features by adopting a code similarity detection tool; extracting a function call graph, a control flow graph and data flow dependence from the input source code of the tested program by adopting a graph extractor by taking a dangerous code block as a target, and generating a combined graph from the function call graph, the control flow graph and the data flow dependence; carrying out distance calculation on the dangerous code blocks through a specified distance algorithm so as to find out related code blocks in the merged image; and performing selective instrumentation and fuzzy testing on the input source code of the tested program based on the obtained dangerous code block and the related code block. According to the method, the target site guiding advantage of the directional grey box fuzzy test is exerted, and meanwhile, the overhead of path exploration is remarkably reduced, so that more accurate and more efficient vulnerability detection is realized.

The invention discloses an intelligent system grey box fuzzy testing method based on state awareness. The method comprises the steps that S1, a state-memory data set is acquired; s2, constructing and training a feature representation network; s3, calculating a feature center and a judgment threshold value of each known state category; s4, collecting the memory when the intelligent system runs in real time, obtaining a to-be-tested memory snapshot, and extracting a feature vector of the to-be-tested memory snapshot by using the feature representation network; s5, performing similarity comparison on the feature vector of the memory snapshot to be detected and all feature centers in the state prototype library, and if the distance between the feature vector of the memory snapshot to be detected and any feature center does not exceed a corresponding judgment threshold, judging that the intelligent system is in the known state category; otherwise, deducing the state as a potential new state; and S6, verifying the potential new state, and if the potential new state is confirmed to be a new state, adding the feature vector of the potential new state into the state-memory data set, and dynamically updating the state prototype library. According to the invention, more accurate description and efficient testing of complex state logic of the intelligent system are realized.

Computer-implemented method for simulating module operation of a module of a modular industrial plant include providing an initial model for the module based on a simulation, e.g., a white-box simulation or a black-box simulation or a grey-box simulation; performing a knowledge-based enhancement step or a data-driven enhancement step comprising obtaining an enhanced model; and simulating module operation of the module by means of the enhanced model.

The application provides a kind of grey box fuzzing test method and system based on Intel-PT and Syzkaller, method includes: first fuzzing test engine starts test cycle, constantly test and synchronize test progress to manager;Manager sends the new use case found by second fuzzing test engine to first fuzzing test engine in response to the poll request of first fuzzing test engine, the effective queue of first fuzzing test engine is updated, start to generate new use case and test;First fuzzing test engine sends new use case to executor, executor prepares to execute test;Executor identifies kernel address range before executing test, collects static storage area of new use case, applies to start PT tracking function by super call;Intercept control register setting request, obtain the value of control register, and assign the value of control register to the parameter of super call, the value of control register is the page table base address of tracking system;Process super call to set the kernel address range of tracking, set page table base address, start PT tracking;Executor starts to parse and execute input use case, real-time call registered parsing function to complete parsing, fill in static storage area;After the execution of multiple rounds of system call of executor, updated coverage information is synchronized to first fuzzing test engine.The grey box fuzzing test method and system based on Intel-PT and Syzkaller provided by the application can be used for closed source version linux system binary grey box fuzzing test, to form effective feedback information closed loop.

The invention relates to a stateful protocol grey box fuzzy test method based on heuristic variation scheduling, which comprises the following steps: S1, initializing a test environment, and establishing a high-speed communication channel based on a shared memory, S2, dividing an initial seed set into a plurality of source seed systems according to sources; s3, selecting a target test state based on an inferred protocol state machine, and selecting a message sequence capable of reaching the state from a corpus, S4, selecting an enabled mutation operator to perform havoc mutation on candidate subsequences according to a current generation offspring mutation strategy of a current source seed system; s5, the test case is sent to the server through the shared memory channel, and the execution state is monitored; s6, generating a new generation of filial generation strategy after the first generation of filial generation strategy of the current source seed system is evaluated; and S7, repeating the steps S3 to S6 until a test termination condition is met. According to the method, global exploration and local development can be effectively balanced, the diversity and pertinence of test cases are improved, and the coverage capability of protocol states and code paths and the discovery probability of deep vulnerabilities are remarkably enhanced.

The invention relates to a complex control flow-oriented directional grey box fuzzy test method, which comprises the following steps of: S1, constructing a target call stack and performing program instrumentation, and collecting the target call stack and path feedback information during operation at an instrumentation point; s2, coarse-grained calling sequence similarity screening is carried out, and coarse-grained calling sequence similarity is obtained; s3, fine-grained semantic similarity calculation: performing weighted aggregation on semantic vectors of all function nodes in the target call stack to form stack-level semantic vectors, and calculating fine-grained semantic similarity by using cosine similarity; s4, carrying out comprehensive distance fusion to obtain a comprehensive distance for seed priority evaluation; according to the method, the path identification and approximation capability of the directional grey box fuzzy test in a complex control flow scene is optimized; the efficient and extensible directional fuzzy testing method is provided for modern software security testing, the analysis ability in a complex control flow scene is expanded, the software security is improved, and the vulnerability verification process is optimized.

This invention belongs to the field of software security testing technology, specifically a method for automatically generating vulnerability proof-of-concept (PoC) based on option-aware directed gray-box fuzzing. The invention includes: performing static analysis on the target program, extracting each configuration option and its directly affecting variables, analyzing the complete value space and control conditions of the variables, and inferring the constraint relationships between options; based on option knowledge, employing option validity inference technology based on taint analysis during the fuzzing phase to dynamically identify option combinations effective in reaching the target location; introducing an alternating guided fuzzing strategy, alternately executing option input fuzzing and file input fuzzing, systematically driving driver execution to approximate the target vulnerability location, and ultimately efficiently generating the target vulnerability PoC. This invention effectively solves the problems of traditional directed gray-box fuzzing tools ignoring the influence of program configuration options, the independence of options and file mutations, and the lack of coordination, significantly improving the efficiency and success rate of vulnerability PoC generation.

In exemplary embodiments, methods and systems are provided that include one or more sensors of a vehicle and a processor of the vehicle. The one or more sensors are configured to obtain sensor data as to operation of the vehicle. The processor is coupled to the one or more sensors, and is configured to at least facilitate inputting the sensor data into a grey-box model including a neural network model for the forces acting on the vehicle and a physics-based model for body motion of the vehicle; determining a trajectory of the vehicle, using the grey-box model; determining a control action for the vehicle, based on the grey-box model using the sensor data; and executing the control action for the vehicle, in accordance with instructions provided by the processor.

The invention discloses a fully mechanized coal mining face dynamic cutting simulation method based on a black box-grey box-white box coal seam model. Constructing a virtual geological simulation module, and generating a virtual geological model data file; on the basis of virtual geologic model data, a coal rock interface is constructed through simulated drilling sampling and a Kriging interpolation algorithm, an exploration error coefficient is introduced, and black box, grey box and white box coal seam error models are constructed; importing a three-dimensional model and an error model of the fully-mechanized coal mining equipment into the simulation platform, and constructing an equipment-coal seam coupling model; calling real-time pose change data of a coal mining machine roller, and adaptively generating a cutting top plate and a cutting bottom plate; and utilizing a pre-trained intelligent model to predict the cutting force in real time, dynamically updating the state of the virtual exploration coal seam error model based on a prediction result and a real-time cutting track, and autonomously regulating and controlling the operation parameters of the coal mining machine. According to the method, a three-level error model is constructed, a dynamic model system of the geological cognition process can be truly reflected, and deep fusion of exploration errors and cutting simulation is achieved.

The invention provides a fuzz test debugging method and device based on a recording and replaying technology, and relates to the technical field of computers.The method comprises the steps that fuzz testing is conducted on a target object, and first information in the testing process is recorded; the first information comprises interaction information between test strategies used in the fuzzy test process and metadata related to seeds in the test process; after the fuzz test is finished, obtaining first information, and replaying the recorded fuzz test behavior based on the first information to obtain second information; the second information comprises an execution path of the seed and information related to the test behavior. According to the fuzz test debugging method and device based on the recording and replaying technology, the execution condition of the key strategy in the fuzz test process is accurately recorded and replayed, so that the analysis and improvement process of the fuzz tester is simplified, and the fuzz test debugging method and device are particularly suitable for optimizing the grey box fuzz test guided by the coverage rate.

The application discloses a microcontroller unit (MCU) firmware grey box fuzzing test method. The method first establishes a high-fidelity peripheral simulator in a simulation execution environment, the peripheral simulator including peripheral registers, internal states and state machines, and determines the interrupt line enable state by jointly monitoring the global interrupt controller register and the peripheral interrupt control register to realize the controlled triggering of the peripheral interrupt. Secondly, the time when the firmware or DMA reads the peripheral data register is determined as an input consumption point during firmware execution, an incremental snapshot containing the processor register state, the peripheral state and the memory dirty page information is saved at the input consumption point, and only the input data that has not been read is mutated after the snapshot is restored. Finally, the test input is divided into multiple input streams corresponding to different peripherals or data registers, a probability distribution is constructed according to the access frequency of each input stream, and the input stream to be mutated is selected, so that the path exploration efficiency and the vulnerability discovery capability of the fuzzing test are improved.

The invention relates to an automatic SSTI grey box penetration test method based on reinforcement learning. Comprising the following steps: taking a Web application program source code as input, and outputting a risk point list through white-box risk prediction by adopting CodeQL and reinforcement learning; in a full black box mode, crawling Web application dynamic content to identify potential injection points, and forming an injection point set together with the risk point list; and constructing an escape agent and a killing-free agent, sharing a state space, testing the injection point set, and generating a corresponding attack load. Vulnerability information returned by CodeQL is associated with a URL, and is converted into an injection point which can be identified by a black box attack load generator, and the injection point is supplemented to a black box analysis process to reduce the missing report rate; a black box attack load generator is used for replacing manual feedback, a white box risk prediction result is automatically verified, the authenticity of vulnerabilities is verified through actual attack testing, and automatic optimization of the false alarm rate is achieved.

The invention relates to a waste paper pulp heavy low-concentration deslagging control method and system, and the method comprises the following steps: S1, collecting real-time process data from a DCS system through an OPC UA protocol, and forming an original data flow; s2, preprocessing the original data to obtain preprocessed data; s3, derivative features are calculated based on the preprocessed data, feature vectors are formed, and state representation containing historical information is constructed; s4, constructing a grey box model as a reinforcement learning environment in combination with a deslagging process physical mechanism and a data driving method; s5, based on the grey box model, constructing a structured state space, an action space and a reward function of reinforcement learning; and S6, according to the structured state space, the action space and the reward function, adopting an improved SAC algorithm to obtain an optimal control scheme. The waste paper pulp heavy low-concentration deslagging efficiency and reliability are effectively improved.

The application discloses a coverage-guided grey-box fuzzing method based on a large language model (LLM) agent collaboration. The method first constructs a test environment composed of a fuzzer, a mutation server, a source code coverage collector, and multiple large language model agents. The mutation server provides a hot-loadable mutation strategy module to the executor through an inter-process communication channel. The source code coverage collector independently runs a source code coverage instrumentation binary to collect source code coverage when each new test case is stored. Second, during the fuzzing execution process, Thompson Sampling is used to adaptively schedule between the traditional byte mutator and the large language model mutator, and Thompson Sampling dynamically allocates mutation opportunities based on the posterior probability of new coverage generated by each mutation path. Finally, during the test iteration process, the coverage analysis agent generates an exploration direction based on the source code coverage data, the mutator writing agent generates and hot-loads new mutation strategy modules based on the exploration direction, the test case generation agent generates test cases based on the exploration direction and imports them into the executor, and each agent retains experience between iterations through a persistent memory file to improve the coverage convergence speed and vulnerability discovery capability of the fuzzing.

The invention relates to the technical field of software security testing, in particular to an embedded network equipment closed source protocol grey box fuzzy testing method and system, and the method comprises the steps: extracting a condition jump instruction address in a target program, constructing a terminal debugging environment, and obtaining an initial population for generating a test case; reading all seeds of the seed path, establishing communication between the sub-process and the debugging port, and setting a breakpoint at a conditional jump instruction address of the target program; executing a target program in a terminal debugging environment, if a breakpoint signal is received, executing and recording instruction addresses before and after execution of flow jump when a program segment is executed to a breakpoint in a single step to generate an edge identifier, allocating a variation strategy to seeds in a seed path based on the edge identifier, and executing a seed variation operation, and taking the current fuzzy input as a vulnerability sample and storing the vulnerability sample until a crash signal is triggered during execution of the target program. According to the method, automatic vulnerability mining of a closed source protocol and an embedded system can be realized.