211 results about "Beam search" patented technology

The invention provides a personalized teaching method and system based on generative artificial intelligence. The method comprises the following steps: constructing a knowledge graph and a teaching strategy template library of a target course; constructing a feature enhancement matrix; calculating node activation intensity, generating a plurality of candidate learning paths through a beam search algorithm, and screening an optimal learning path in combination with a preset teaching strategy template; using the generative model to generate standardized teaching content adaptive to the teaching strategy; dynamically optimizing a teaching scheme through a neuroplasticity model; and carrying out teaching adjustment based on the optimized teaching scheme. Standardized teaching content is generated by utilizing the generative model, the teaching content quality is improved, the teaching content is closely matched with a learning path and a teaching strategy of a learner, meanwhile, a teaching scheme is optimized through the neuroplasticity model, teaching parameters and cognitive resource allocation are adjusted in time according to real-time feedback and cognitive states of the learner, and the teaching efficiency is improved. Limited cognitive resources are reasonably distributed to key parts of teaching contents, and the teaching effect is improved.

The invention discloses a non-point source pollution emergency bypass decision control method, particularly relates to the technical field of automation and process control, and aims to solve the problems that existing scheduling cannot align multi-source measurement and evaluate credibility in minutes under a unified time base, and is lack of risk measurement integrated with a tide level phase and a moisture regain path and same-layer gating. The bypass rhythm is easy to misjudge, and the oscillation is difficult to audit. A minute-level risk density is constructed through multi-source alignment and credibility grading under a unified time reference, injected tide level phases and moisture regaining paths, bypass rhythms are constrained by same-layer gating and tide windows, and beam search, boundary arbitration and fixed window closed-loop correction are matched, so that the risk density is improved. Therefore, under the superposition of rainstorm and tide, decision making and execution are stably limited in a non-backflow red line, bearing and equipment limit, misjudgment and oscillation are remarkably reduced, the problems of unavailability and difficult auditing caused by multi-source asynchronization and moisture regain coupling are solved, intelligent decision support is provided for non-point source pollution treatment, and it is ensured that the treatment technology is accurate and efficient.

The invention belongs to a decoding method, and provides a multi-modal brain-computer interface decoding method and a related device for solving the technical problems that an existing non-intrusive brain language decoding method is insufficient in adaptability in global context and weak in generalization performance in a cross-subject scene, and multi-modal neural feature collaborative enhancement is difficult to achieve. And determining the called execution module. The execution module comprises a feature extraction module, a cross-subject standardization module, a multi-mode semantic fusion module, a language recognition module and a semantic consistency module. By obtaining the unified semantic representation and combining the beam search algorithm, the fairness and universality in different language groups are remarkably improved, multiple modes can be supported, and the brain signal decoding precision and robustness are improved. In addition, cross-subject semantic representation generalization can be realized, and individual specificity is effectively reduced.

The invention discloses a method and a system for detecting a jailbreak attack of a large language model, which are applied to discovering potential security risks of the large language model, and the method comprises the following steps: constructing a jailbreak attack-malicious query data set of the large language model; an automatic antagonism suffix generation method fusing cluster search and a gradient descent algorithm is provided; the detection system of the prison break attack large language model is constructed by training, testing and evaluating the validity of the resistant suffixes based on a prison break attack data set, and comprises a malicious query data set display unit, a detection validity verification unit, a comparison and analysis unit before and after an attack and a model defense strategy unit. The method is helpful for discovering defects of a large language model alignment mechanism, so that a more powerful large language security model is designed and constructed.

The invention discloses a method for generating an image description text based on a large model, which relates to the technical field of image processing, and comprises the following steps: an image preprocessing step: dividing a target level through semantic segmentation and extracting key visual information by adopting hybrid denoising and self-adaptive normalization; a feature extraction step: fusing the multi-scale visual features and the semantic features, and generating a high-dimensional fusion feature vector through cross-modal alignment; a large model initialization and adaptation step: loading the pre-training model and performing incremental fine tuning, and dynamically adjusting the Prompt template; a text generation step: generating candidate texts through logic constraint and beam search; and a text optimization adjustment step of outputting a final description text based on multi-dimensional evaluation and user preference iterative correction. According to the method, the semantic matching degree, logic coherence and common sense accuracy of image description are improved, multi-element scenes are adapted through dynamic adaptation and iterative optimization, and high-quality text description support is provided for high-precision and diversified scenes.

The invention discloses a network attack path deduction method and device based on a large language model, equipment and a medium, and relates to the technical field of network security, and the method comprises the steps: generating each initial deduction result corresponding to a current abnormal network activity through employing a target large language model and a beam search algorithm; verifying each initial network attack deduction path by using a target knowledge graph to obtain a target confidence degree corresponding to each initial network attack deduction path; and determining a target network attack path with the highest target confidence from the initial network attack deduction paths, and determining an initial deduction result corresponding to the target network attack path as a target deduction result so as to display the target deduction result by using a time sequence animation. The initial network attack deduction path is generated by using the large language model, and the target network attack deduction path is determined based on the knowledge graph, so that the problem that complex and changeable attack techniques cannot be adapted due to rule engine stiffness is solved.

The invention discloses a wafer scheduling method and semiconductor process equipment. The method comprises the following steps: determining a first candidate action sequence according to an objective function value of each action sequence obtained by cluster search; determining a second candidate action sequence which reaches the process chamber fastest in all action sequences corresponding to the same new wafer; in response to the fact that the number of actions in the candidate action sequence reaches an action number threshold value, determining a target action sequence according to the target function value; in response to the fact that a first action of the target action sequence meets a preposition constraint, the first action of the target action sequence is executed in a simulation mode, and the preposition constraint is that new wafers executing different tasks are taken out from the loading and unloading positions according to the sequence of reaching the process chamber; and in response to completion of simulation scheduling of all wafers, outputting an action scheduling sequence including sequential simulation execution. According to the invention, the wafer can be processed as soon as possible. The development period is short, the maintenance cost is low, and the result is more accurate and reliable.

A computer-implemented method for optimizing a video encoder. The method includes the step of defining a plurality of CUs of the video encoder that correspond to a plurality of stages. The plurality of stages includes at least a first stage, and a second stage immediately after the first stage. The method further includes the steps of providing a decision space of encoding parameters of the video encoder, and defining, for the second stage, a subspace being a subset of the decision space. The subspace contains b1 optimal decision paths from the first stage to the second stage, wherein b1 is defined as a beam size for the first stage. The b1 optimal decision paths are the b1 decision paths that have lowest accumulated cost among all decision paths from the first stage to the second stage.

A method for updating a question-answer mechanism includes obtaining, by a data system implementing the question-response mechanism, an input query, in response to the input query: generating a token sequence associated with the input query by applying a beam search algorithm on generated bound knowledge graphs, applying the input query to a large language model (LLM) of the question-response mechanism to obtain first outputs, applying the generated token sequence to a graph neural network (GNN) to obtain second outputs, applying fusion layers on the first outputs and the second outputs to generate a response associated with the input query and generated token sequence, and performing a remediation using the generated response, wherein the remediation comprises updating the LLM and the GNN based on the generated response.

The invention discloses an intelligent fast beam training method for low-altitude millimeter wave communication, and belongs to the technical field of wireless communication. The method comprises the following steps: establishing a channel and transmission model of the low-altitude millimeter wave cellular-free large-scale MIMO system; in the channel estimation stage, a pilot signal is used for estimating a channel and preliminarily positioning the position of a user; in the beam training stage, firstly, customized three-dimensional beam codebook design is carried out based on obtained positioning information; and then, an enhanced multi-agent depth Q network algorithm is adopted to efficiently carry out beam search in the designed customized codebook, so that accurate three-dimensional beam alignment is realized. According to the method, the customized codebook and the intelligent learning algorithm are combined, the beam forming process is effectively optimized, the complexity and overhead of beam training are remarkably reduced, the convergence speed is increased, the system performance and reliability of downlink communication are improved, and the method is suitable for a high-maneuverability low-altitude unmanned aerial vehicle communication scene.

The invention discloses a robot control system and method based on AI voice, and belongs to the technical field of robot intelligent control. The system comprises a local voice processing module, a dialogue management and context understanding module, an instruction mapping and generating module and a robot communication and control module. The local voice processing module is integrated with a 16 kHz sampling ASR engine, voice-to-text conversion is realized through a 6-layer LSTM network and beam search decoding, and the TTS engine adopts a WaveNet architecture to synthesize voice. A 7 billion parameter large language model is arranged in the dialogue management module, a 32-layer Transform decoder architecture is adopted, 10 rounds of dialogue history is maintained, and intention recognition and entity extraction are achieved through a multi-head self-attention mechanism. And the instruction mapping module performs trajectory planning by adopting quintic polynomial interpolation and quaternion spherical linear interpolation. According to the invention, full-link localization processing is realized, the speech recognition delay is 180 milliseconds, the position error is less than 0.5 mm, and the problems of high delay, high privacy risk and weak complex instruction understanding ability in the prior art are effectively solved.

To provide native hardware support for complex operations such as a beam search operation, an integrated circuit device can be implemented with multiple computational circuit blocks coupled in series to form a pipeline. Each computational circuit block includes an arithmetic logic unit (ALU) circuit having a first numeric input, a second numeric input, a primary result output, and a secondary output. The ALU circuit is programmable to perform an arithmetic operation on the first numeric input and the second numeric input to generate the primary result output, and provide one of the first numeric input or the second numeric input to the secondary output, which is fed back to the ALU circuit.

The invention discloses a simulation method and system for students with different cognitive levels based on a large language model. According to the method, the cognitive prototype is constructed based on the past learning records of the students, and the mastering degrees of the students on different knowledge concepts are clearly expressed through the knowledge graph. The cognitive prototype can be used for accurately predicting the performance of the student in a new task, including whether the problem can be solved or not and specific errors possibly caused. Meanwhile, a self-evaluation-self-optimization loop process based on beam search is introduced, and a student answer consistent with a prediction behavior is iteratively generated, so that student simulation which is more real, natural and consistent in cognition is realized on the premise that additional training is not needed. Related results show that the staged and cognitive prototype-driven student simulation framework provided by the invention is superior to the existing mainstream method in multiple dimensions, and behavior characteristics of students with different cognitive levels can be reproduced more truly.

The invention is suitable for the technical field of gene sequencing, and provides a nanopore sequencing signal identification method and system based on Transform, and the method comprises the following steps: carrying out normalization processing on a nanopore sequencing signal, and segmenting the nanopore sequencing signal into signal windows with fixed lengths; signal features are extracted through a Transformer encoder comprising a convolutional layer and a multi-head self-attention module; generating a prediction probability of a current base through a cross attention mechanism by adopting an autoregressive Transform decoder and combining the signal features and context information of a decoded base sequence; iteratively generating an optimal base sequence based on a beam search algorithm, and optimizing decoding efficiency in combination with key value cache; and carrying out overlapping region fusion and quality control on a segmented decoding result, and outputting a final sequence. According to the method, the precision and reliability of nanopore sequencing in complex biological sequence analysis are remarkably improved, and the recognition accuracy on multiple gene sequencing data sets is higher than that of an existing CTC-based model.

The disclosure relates to a communication technique for a 5G communication system for supporting a higher data transfer rate after a 4G system, and a system therefor. The disclosure may be applied to an intelligent service (for example, a smart home, a smart building, a smart city, a smart car or connected car, healthcare, digital education, retail business, and a service relating to security and safety), based on a 5G communication technology and an IoT-related technology. The disclosure relates to a method and device for performing beam searching in a mobile communication system using beamforming.

The invention discloses a medical text condition generation method under rule constraints, and belongs to the technical field of natural language processing and medical information. According to the method, a medical rule knowledge graph is constructed, graph neural network coding is adopted, and medical business rules are converted into constraint representation; designing a rule-perceived attention mechanism, and fusing medical rule constraints in a generation process; constructing a constraint-aware Transform decoder, and ensuring the generation normalization by adopting an improved beam search algorithm; and establishing a multi-level quality control system, and performing three-dimensional evaluation from local consistency, global consistency and rule conformity to generate a result. According to the method, the technical problem that existing medical text generation cannot meet complex rule constraints while guaranteeing fluency is solved, the rule coincidence rate reaches 95% or above, the quality score is improved by 18-25% compared with an existing method, and the method is excellent in performance in application of multiple medical departments.

A method for performing time constrained queries for event sequence visualization may include receiving an event sequence dataset including event sequences. Each event sequence includes at least two sequentially occurring events. The method further includes generating a forward index having event pattern sets that include pattern occurrence data for each event sequence positioned according to a fixed positional order. The method includes generating a time constraint search structure, such as a B−tree, B+tree, or table, that includes keys and pointer values. The method also includes inserting, into the search structure, a sequence identifier corresponding to each event sequence based on the fixed positional order and time interval data for the at least two sequentially occurring events of each event sequence. The method may also include searching the search structure to identify key values satisfying a received time interval constraint, and outputting the sequence identifiers corresponding to the identified key values.

The invention discloses a text-to-instruction system based on fuzzy pinyin matching and parameter robust analysis. The text-to-instruction system comprises an input preprocessing module, a text-to-instruction conversion module and a text-to-instruction conversion module, the fuzzy pinyin matching module is used for uniformly transferring the standardized text and a pre-stored instruction template into a silent pinyin sequence, and constructing a weighted editing distance matrix based on a preset pinyin character replacement cost; the parameter alignment search module is used for executing beam search on the weighted editing distance matrix and outputting a plurality of candidate matching paths with the minimum cost and a corresponding parameter fragment set; the parameter robust distribution module is used for performing legality and consistency checking on parameters in the candidate matching paths, completing parameter standardization and performing error correction in combination with a parameter candidate dictionary; the instruction output module is used for encoding the template identifier and the parameter key value pair obtained through robust distribution into a structured instruction and outputting the structured instruction; according to the method, text-to-instruction conversion can be completed in real time.

The invention provides a cargo loading scheme determination method, electronic equipment and a storage medium, and relates to the technical field of logistics management. The method comprises the steps of obtaining loading space parameters and attribute information corresponding to each to-be-loaded cargo; sorting the to-be-loaded goods according to the attribute information corresponding to each to-be-loaded goods to obtain a goods sequence; according to the cargo sequence, the loading space parameters and the attribute information corresponding to each to-be-loaded cargo, adopting a beam search algorithm and a maximum spare space algorithm to obtain an initial loading scheme; according to the initial loading scheme, the loading space parameters and the attribute information corresponding to each to-be-loaded cargo, the initial loading scheme is destroyed and repaired through a self-adaptive large-field search algorithm, a final loading scheme is obtained, and the final loading scheme is used for indicating the final loading position of each to-be-loaded cargo in the loading space. Therefore, the loading rate of the loading scheme is improved.

The invention discloses a monomolecular structure design and generation method based on a large language model, and relates to the crossing field of monomolecular electronics and artificial intelligence. In order to solve the problems of low single molecule design efficiency and insufficient data, SingMolT5 is obtained through field fine tuning on the basis of MolT5, and generation from a natural language to a molecular SMILES is realized. According to the method, data are collected from literatures in the single molecule field and an existing disclosed molecule library, a fine tuning data set containing 329 high-quality instructions is constructed, and the fine tuning data set comprises three basic data types including a molecular skeleton, an anchoring group and molecules. The model is initialized by a MolT5 check point, and a cross entropy loss function is used in the fine tuning process; reasoning, generating and using a beam search strategy; the follow-up evaluation indexes comprise BLEU, Levenshtein, MACCS, RDK, Morga and effectiveness. The method provides an innovative technical path for'low experience and data dependent type single molecule design 'in the field of single molecule electronics, and is particularly suitable for aided design of a single molecule device core function unit and a single molecule sensing probe molecule, namely, a target molecular structure can be quickly converted through a natural language, the design threshold of researchers is reduced, and the design efficiency is improved. And a traceable technical basis can be provided for screening and iteration of a monomolecular structure based on a high-quality data set and rigorous evaluation logic.

A hierarchical beam-deployment method has the step of transmitting signal beams of multiple levels towards a service area, each signal beam comprising a synchronization signal and physical broadcast channel block (SSB); the footprint of a higher-level beam has a beam-width smaller than that of a lower-level beam, and is at least partially overlapping with the footprint of the lower-level beam. Accordingly, a beam-search method has the steps of detecting a first SSB of a first signal beam; determining, based on the first SSB, a presence of one or more second signal beams having one or more footprints at least partially overlapping with a footprint of the first signal beam; detecting one or more second SSBs of the one or more second signal beams; and selecting at least one of the first and second SSBs for performing an initial-access procedure based on the selected SSB.

The invention discloses a channel routing optimization system based on a graph neural network, and the system comprises the following modules: a dynamic network graph construction module which is used for collecting network topology and link state data, and constructing a dynamic network graph model; the node embedding generation module is used for inputting the data into an improved MTGNN model and generating a high-quality node embedding vector by introducing a contrast learning mechanism; the candidate path discovery module adopts a greedy neural network algorithm based on beam search, carries out parallel extension and retains high-score candidate paths according to node embedding vector scores, and generates a candidate path set; the optimal path selection module is used for calculating the comprehensive utility value of each path, sorting the comprehensive utility values and outputting an optimal path; and the routing configuration and optimization module is used for generating and issuing a routing configuration instruction according to the optimal path. According to the invention, the accuracy and efficiency of routing decision and the utilization rate of network resources are improved, and intelligent path optimization in a complex dynamic network environment is realized.

The invention provides a context adaptive speech recognition method integrating entity replication and vector retrieval, and belongs to the technical field of artificial intelligence. The method comprises the following steps: converting input voice into acoustic feature representation; converting the external entity dictionary into vector representation; screening out candidate entities through an index technology, and determining to generate output or copy entities from a standard word list according to a matching degree calculated by an attention mechanism; jointly optimizing the model by comprehensively optimizing the objective function, wherein acoustic modeling, language modeling, entity copying accuracy and entity distinguishing capability are included; continuously updating the index during training; in reasoning, in combination with beam search and confidence threshold mechanisms, selection is made between generation of outputs from a standard word list and copying of entities. In the aspect of recognition performance, a dynamic copy-generation fusion mechanism and a contrast learning strategy are adopted in the method, the accuracy of named entity recognition is effectively improved, particularly, the method is excellent in performance when proper nouns such as personal names and place names are processed, and meanwhile the distinguishing capacity for entities similar in pronunciation is enhanced.

In order to reduce load due to a beam search of a base station and a communication device, one aspect provides a base station characterized by comprising: a reception unit for receiving, from another base station, a signal including information about a reception beam that the other base station uses when receiving a signal from a first communication device; and a control unit that determines a candidate for a beam to be used for communication between the other base station and the second communication device according to information about a transmission beam used when transmitting a signal to the first communication device and the information about the reception beam.

This invention provides a near-field millimeter-wave codebook generation method and related components. The near-field millimeter-wave codebook generation method is applied to a near-field hierarchical communication beamcodebook system. The method includes: obtaining the level number of the codebook at the current level in the near-field hierarchical communication beamcodebook system; determining the codebook level of the current level according to the relationship between the level number and a preset level, wherein the codebook level includes an upper-level codebook and a lower-level codebook; and obtaining a near-field millimeter-wave communication codebook based on the codebook level of the current level. The near-field millimeter-wave codebook generation method provided by this invention considers the characteristics of the spherical wavefront in the near-field region, avoids performance degradation, and achieves coverage of the entire Fresnel region. It designs a lower-level codebook covering polar coordinates and an upper-level codebook designed to reduce beam search overhead.

The invention discloses a front door adjustment-based harmful modal cause detection depolarization method, which comprises the following steps of: constructing a structural causal model, creating a back door path between a modal cause and a harmful label by using unobserved confounding factors, and eliminating harmful modal cause detection through front door adjustment; the front door adjustment then blocks the back door path by introducing reasoning as an intermediary between the memetic and the tag. The method comprises the following steps: decomposing the causal effect of a medal factor on a harmful tag into the influence of the medal factor on an intermediary, quantifying by adopting a multi-modal large-scale language model with different beam searches to calculate the probability of each reasoning sequence and the influence of the intermediary on the tag, and performing effective approximation through a normalized weighted geometric averaging method; finally, post-interpretation is synthesized from the inference sequence and the predictive tag using a large language model. According to the scheme, the most advanced performance is achieved while information interpretation is provided.

The application provides a knowledge processing method and device for power inspection defect detection, and the method comprises the following steps: obtaining an optimal path to connect a starting point entity and an ending point entity in the power equipment knowledge graph to form a Markov decision process according to the starting point entity and the ending point entity; training the Markov decision process to obtain a trained Markov decision process; an agent in the trained Markov decision process selects output edges in turn by using a random beam search strategy, and traverses to a new entity; when the agent searches for an optimal path by using the random beam search strategy, the path automatically forms a predicted explainable source. The application not only makes the power equipment knowledge graph knowledge completion explainable, but also improves the ability to understand complex environments and infer high-level causal relationships.

The present invention discloses a method for optimizing FMS system scheduling and resource allocation based on Petri nets, belonging to the technical field of manufacturing system optimization and intelligent control. The method comprises the following steps: allocating resources of a flexible manufacturing system according to production requirements of the flexible manufacturing system, obtaining current processing information of the flexible manufacturing system, and establishing a Petri net model of the flexible manufacturing system based on the current processing information of the flexible manufacturing system; dividing the transition set of the Petri net into base partitions, and constructing a base reachable graph based on the base partitions; using an improved simulated annealing algorithm combined with a generative filter beam search algorithm to perform integrated optimization of resource allocation and scheduling in the base reachable graph to obtain an optimized resource allocation scheme and machine processing time; and, based on the optimized resource allocation scheme and machine processing time, sorting the operations of all workpieces on a time axis according to the process start time to generate a final scheduling result.

The present disclosure relates to a communication method and device in a beamforming-based wireless communication system supporting a plurality of frequency bands, the communication method for a terminal, according to an embodiment of the present disclosure, comprising the steps of: performing a beam search using a plurality of first reception beams in a first frequency band and receiving a synchronization signal block (SSB) comprising first information about a base station provided at the same location; performing, on the basis of the first information, a beam search for a plurality of second reception beams in a second frequency band belonging to a subset of the first reception beams that received the SSB in the first frequency band; and communicating with the base station in the second frequency band by selecting an optimal pair of transmission and reception beams as result of the beam search for the plurality of second reception beams.

The application relates to a bundle search processing method and device, computer equipment and a storage medium. The method comprises the following steps: performing bundle search on a word table based on a current to-be-matched object and a current bundle width to obtain a current search result, wherein the current search result comprises a plurality of target words which are matched successfully with the current to-be-matched object and the number of which matches the current bundle width; performing width attenuation on the current bundle width according to a bundle width attenuation mode matched by the number of executed bundle searches to obtain a bundle width for next bundle search; taking each target word as a to-be-matched object for next bundle search, and iteratively performing bundle search until the bundle width obtained by attenuation is equal to a bundle width threshold value, and taking the search result at this time as a target to-be-matched object, and performing bundle search on the word table based on the bundle width threshold value to obtain a target search result. By adopting the method, the search efficiency can be improved while ensuring the search effect in an ultra-long sequence scene.