15 results about "Memory behavior" patented technology

The invention belongs to the field of industrial control safety, provides a PLC (programmable logic controller) memory behavior analysis-based control system configuration logic anomaly detection method, and solves the problems of logic attack detection blind areas, incomplete detection coverage, process semantic deficiency and the like caused by hidden attacks (modifying PLC control logic through a legal programming interface). The method comprises the following steps: S1, collecting a PLC register with a timestamp and a coil state sequence; s2, identifying a key memory address, constructing an address cluster set, and supporting event completion and process mining; s3, generating an XES standard event log by adopting a clustering and symbol back-pushing method; s4, improving a Heuristics Miner algorithm, automatically identifying a starting point and an end point of a control flow, and constructing a dependency graph with an identifier; s5, converting the control logic into a formalized model, and defining a legal state transition sequence; and S6, monitoring a real-time event stream on line, mapping the real-time event stream to the model and dynamically playing back the real-time event stream, and realizing anomaly detection by combining track alignment and comparison. The method is complete in coverage, clear in semantics, capable of tracing attack paths and capable of improving the defense capability of a logic layer.

The invention provides a maze robot brain-like cognitive system based on an emotion and memory mechanism, and the system comprises a perception and emotion generation module which obtains the environment state of a robot, and calculates the reflective fear value and cognitive fear value of the robot; the memory module is used for recording short-time memory, working memory and long-time memory of the robot in combination with an anti-thought memory mechanism; the behavior decision-making module is used for comparing the evaluation function based on the cognitive fear value with the energy margin to obtain a behavior decision-making result; the learning module is used for exploring and learning a target node and an energy node to obtain an updated target guide table and an updated energy guide table; and the action selection module is used for updating the target guide table and the energy guide table according to the behavior decision result, and outputting the next action. According to the invention, the survival rate of the robot in the environment learning process can be improved, and long-time learning is realized; and the robot can be prevented from repeatedly exploring in an invalid area, the learning efficiency is improved, the energy consumption is reduced, and self-survival is ensured.

The invention discloses a self-adaptive memory allocation optimization method and device, terminal equipment and a storage medium, and belongs to the field of memory management. The method comprises the steps of monitoring thread data and system state data in real time, extracting thread memory behavior characteristics and resource state characteristics when one variable quantity exceeds a safety threshold value, and performing weighting calculation according to a preset weight coefficient to obtain a target strategy score; when the strategy score difference value of the target strategy score and the current strategy score is smaller than a preset difference value threshold value, the current memory allocation strategy is kept; when the strategy score difference value is not smaller than a difference value threshold value, a target memory allocation strategy is matched in a preset strategy library according to the thread memory behavior and the resource state characteristics, a current memory allocation strategy is updated, memory allocation parameters are adjusted, and a current strategy score is updated; when the current memory allocation strategy and the current strategy score are compared for the first time, the current memory allocation strategy and the current strategy score are the initial memory allocation strategy and the corresponding initial score, and the problem that efficiency and compatibility cannot be considered at the same time in the prior art is solved.

The application discloses a memory abnormality early warning method and device based on data mining and a medium, relates to the technical field of memory abnormality early warning, and comprises the following steps: performing semantic element mapping on memory running semantic data, and outputting a semantic state time sequence vector; performing cross-domain time sequence alignment and associated feature extraction on the semantic state time sequence vector and memory performance index data, and outputting memory associated rules; performing cause-effect analysis on the memory performance index data, constructing a graph structure on a cause-effect edge set through a PC algorithm, and outputting a memory index cause-effect graph; performing memory behavior deduction on real-time memory running semantic data according to the memory associated rules, outputting memory abnormality early warning data, performing cause-effect tracing and report integration on the memory abnormality early warning data through the memory index cause-effect graph, and outputting a memory abnormality early warning report. The application realizes interpretable early warning and root cause positioning of memory abnormality through the cooperation of semantic state time sequence vector construction and memory index cause-effect graph generation.

This invention discloses a memory-based behavioral adversarial interactive identity verification method and apparatus based on a metaverse. The method involves: acquiring first behavioral information from user feedback based on received interactive information during the interaction process; determining a difficulty weight index for each interactive message received by the user in historical interactions and its historical average response time; constructing a memory tree for each user based on the difficulty weight index; building an adversarial model; obtaining second behavioral information through the adversarial model's feedback of interactive information; performing conditional loop-based adversarial verification based on the first and second behavioral information to obtain a verification result; determining the order of interactive information by combining the memory tree and the previous verification result during the adversarial verification process; obtaining a comprehensive verification score based on all verification results and the difficulty weight index; and determining the user's identity verification result based on the comprehensive verification score, thus making the identity verification result more accurate and objective.

The invention discloses a memory abnormity early warning method and device based on data mining and a medium, and relates to the technical field of memory abnormity early warning, and the method comprises the following steps: carrying out semantic element mapping on memory operation semantic data, and outputting a semantic state time sequence vector; performing cross-domain time sequence alignment and associated feature extraction on the semantic state time sequence vector and the memory performance index data, and outputting a memory association rule; performing causal analysis on the memory performance index data, performing graph structure construction on the causal edge set through a PC algorithm, and outputting a memory index causal graph; and performing memory behavior deduction on the real-time memory operation semantic data according to a memory association rule, outputting memory abnormality early warning data, performing causal tracing and report integration on the memory abnormality early warning data through a memory index causal graph, and outputting a memory abnormality early warning report. According to the method, interpretable early warning and root cause positioning of memory anomalies are realized through dual-mechanism cooperation of semantic state time sequence vector construction and memory index cause and effect graph generation.

The invention relates to the technical field of memory management, in particular to a memory management optimization method based on resource environment limitation and related equipment. The method comprises the following steps: acquiring high-precision metadata in a memory operation event by hijacking a memory allocator interface; performing statistical evaluation according to the historical event data in combination with the high-precision metadata to obtain a memory behavior pattern feature vector; predicting the memory behavior pattern feature vector through a lightweight prediction model to obtain prediction request data; comparing the predicted request data with the fluctuation amplitude of the historical request to obtain a memory management strategy; and the memory management operation is optimized by monitoring the delay improvement rate and the fragment rate. The memory demand can be dynamically predicted, the allocation strategy can be adaptively adjusted, fragment accumulation can be inhibited, and real-time performance, low fragment and high memory utilization rate can be realized.

The invention belongs to the technical field of data prediction, and relates to a memory state prediction model based on an irregular memory behavior sequence and application thereof. The model comprises an embedded layer, a Transform layer, a fusion layer based on attention and an MLP layer. According to the method, the time dependency relationship in the memory behavior is captured by utilizing a self-attention mechanism, and the problem of irregular time intervals is solved through time embedding representation. Then, aggregating memory state vectors by using an attention mechanism; an attention-based fusion layer is designed to better consider the influence of each historical memory behavior on the current memory state. The model provided by the invention can be applied to a personalized learning system and an online education platform, provides a personalized learning plan for learners, and optimizes course design; or the memory state of the learner is monitored, and a prompt is pushed when the learner is about to forget.

The invention relates to a generative dialogue recommendation system and method based on a large language model, the system comprises a three-tower neural network model formed by jointly training three independent encoders, and the three-tower neural network model comprises a user memory behavior tower, a current context tower, a user portrait tower and a feature fusion and decoder; the training module is configured to construct training data, calculate a loss function and complete training to obtain a generative dialogue recommendation model; and reasoning and recommendation, which is configured to output the generated Top-K most possible problem recommendation sequences based on the generative dialogue recommendation model and the input of the triple of the real user. According to the method, user intention mining, semantic matching recommendation and recommendation problem generation can be carried out.

One or more embodiments of the present disclosure provide a method, device, electronic device, and storage medium for dynamic allocation of memory resources. The method includes: obtaining memory behavior characteristic data, the memory behavior characteristic data reflecting the behavioral characteristics of the system or process in terms of memory usage; encoding the memory behavior characteristic data to obtain a state code for the memory behavior characteristic data; dynamically selecting a target prediction model based on actual memory usage operations at historical moments, the actual memory usage operations representing information reflecting the policy effect and system status generated after the execution of the memory allocation strategy at the historical moment; obtaining a memory allocation strategy based on the state code and the target prediction model; and dynamically allocating memory resources based on the memory allocation strategy. Through the present disclosure, memory can be reasonably allocated, resource utilization can be effectively improved, and cost and energy consumption can be reduced.

The invention provides a file-free attack detection method, device and equipment and a computer readable medium, and belongs to the field of network and information security. According to the scheme, in the process of monitoring the key process, the calling behavior fragments related to the memory can be continuously collected, and the calling behavior sequence is obtained; performing dynamic feature extraction on the calling behavior sequence to obtain memory behavior features of the key process, such as memory behavior fingerprints, execution chain node representation and the like; and executing a corresponding anomaly detection method based on the memory behavior characteristics, such as a fingerprint matching method corresponding to a memory behavior fingerprint, an abnormal path scoring method corresponding to execution chain node representation and the like, and obtaining an anomaly detection result without file attack. According to the scheme, the calling behavior sequence of the memory related to the key process is collected for dynamic feature extraction, single-point, continuous and multi-section behavior anomalies can be comprehensively recognized, potential attack behaviors can be mined, linkage analysis is carried out on the link level, continuous and anti-interference detection is achieved, and high elasticity and flexible configuration are achieved.

The embodiment of the invention provides an in-vehicle service pushing method and device of an automobile, the automobile and a storage medium, and the method comprises the steps: generating short-term memory data based on the operation data of the automobile in a target time period, generating long-term memory data based on the historical operation data of the automobile, and generating scene memory data based on the current scene of the automobile; and according to the real-time driving scene of the vehicle, matching at least one corresponding memory behavior from the short-term memory data, the long-term memory data and the scene memory data, so as to push at least one vehicle machine service to the user. According to the method, multi-dimensional data collaborative analysis can be realized, the data is divided into long / short-term memory and scene memory, intelligent adaptation of the driving scene is realized, reinforcement learning can be continuously carried out through user feedback, and the continuous optimization and adaptive capability of the method is improved.

The invention discloses a data processing method, device and equipment and a readable storage medium, and the method comprises the steps: obtaining game progress information corresponding to an object virtual character controlled by an object in a game world, and if the game progress information meets a state triggering condition, executing the state triggering; if yes, obtaining fact information associated with the game progress information in a fact knowledge base of the game world, and generating role state information according to the fact information and the game progress information; inputting the role state information into a retrieval model, and generating target role configuration data through the retrieval model; acquiring historical memory behavior data, and generating latest interaction data according to the target role configuration data and the historical memory behavior data; and generating a target fixed virtual character according to the target character configuration data and the latest interaction data, and putting the target fixed virtual character into a game environment associated with the game progress information. By adopting the method and the device, efficient updating and personalized interaction service for the interaction content of the fixed virtual character can be realized.

The invention discloses a method and system for automatically positioning windows cloud desktop memory leak, and belongs to the technical field of cloud computing, and the method is implemented by the following steps: establishing a driver-level memory monitoring mechanism, tracking the change conditions of a page storage pool and a non-page storage pool in real time through a PoolMon tool, and storing the change conditions of the page storage pool and the non-page storage pool in real time; collecting memory allocation statistics of each pool label, handle counting of a driving object and a memory occupation trend, and emphatically monitoring a driving module related to the virtualization component; an application-level memory monitoring mechanism is established, private working sets, handle counts and GDI object use conditions of all processes are collected regularly through an expanded WMI query interface, a process-level memory use file is generated, and a memory behavior mode in the life cycle of the process-level memory use file is recorded; and starting a verification process when the potential memory leak is detected. According to the invention, automatic distinguishing and identification of drive-level and application-level memory leaks can be realized; and based on an intelligent judgment mechanism of a threshold value, false alarms are reduced.

The invention discloses a phototransistor for simulating neural synapses, a preparation method and a multispectral weight updating circuit, and belongs to the technical field of photoelectric synapse devices. The photoelectric transistor is composed of a gate bottom electrode, a gate dielectric layer, a metal oxide film layer, a quantum dot-PMMA blending layer and source and drain electrodes, an active layer of the device is reconstructed into an oxide / quantum dot / PMMA heterojunction system, the light response of the device is enhanced through the programmable photoelectric characteristic of quantum dots, and the spectral response range of the device is widened; based on a heterojunction charge capture effect and a double-electric-layer ion-electron coupling effect, the long-term synaptic plasticity of the device is greatly improved, so that the device can accurately simulate various memory behaviors related to biomimetic synapses, including generation of current after excitatory synapses, conversion from short-term plasticity to long-term plasticity and the like; a multispectral weight updating circuit built based on the device can realize dynamic weight regulation and control by using all-optical signals, simulates a self-organization reconstruction process of a neural circuit in a human neural network, and has important application prospects in the fields of bionic visual perception and neural morphology calculation.