678 results about "Target Response" patented technology

The invention discloses a knowledge base enhancement generation method and system based on hybrid retrieval and fact verification. The method comprises the following steps: receiving an original query of a user; performing intention analysis and rewriting on the query, identifying an intention type and generating a sub-query adapted to retrieval; executing mixed retrieval of vector retrieval, keyword retrieval and selectable knowledge graph retrieval based on the sub-query, and recalling related knowledge fragments; performing deduplication clustering, fact conflict recognition processing and correlation reordering on the knowledge fragments; according to the intention type and the reordered knowledge fragment, dynamically selecting a prompt template to construct an enhanced prompt; and sending the enhanced prompt into the large language model, and generating a target response with the reference source. According to the method, knowledge recall comprehensiveness is improved through mixed retrieval, knowledge reliability is ensured through fact verification, generation logicality is enhanced in combination with dynamic prompt construction, and response accuracy and credibility are improved.

The invention provides a multi-heat-source multi-material additive manufacturing optimization method and system based on machine learning, and relates to the field of additive manufacturing, and the method comprises the following steps: collecting process parameters, material physical properties and target response data in real time, and constructing a multi-source data set; on the basis, a multi-task learning model is trained, common features are extracted, target responses are synchronously predicted, a proxy target function is constructed through predicted values, if errors exceed a threshold value, experimental verification is conducted, and new data is fed back to update the model. And applying the weight coefficient to a multi-objective evolutionary algorithm, iteratively screening a Pareto candidate solution set, verifying a high-uncertainty solution, and updating the model to error convergence. And selecting and deploying an optimal parameter combination by using an entropy weight-TOPSIS method, and embedding an online learning mechanism to form closed-loop optimization. According to the method, multi-task learning and a self-adaptive evolutionary algorithm are combined, and efficient optimization of multi-heat-source multi-material additive manufacturing process parameters can be achieved.

The invention relates to the technical field of artificial intelligence, in particular to an agent-based question and answer method and device, equipment and a medium. The main agent obtains to-be-processed data input by a user for the to-be-executed task in the first dialogue round; determining a target intention of the to-be-processed data, determining a target response agent corresponding to the target intention in a plurality of response agents according to a pre-stored corresponding relationship between different intentions and the response agents, and sending the to-be-processed data to the target response agent; the target response agent processes the to-be-processed data based on a pre-configured large model of the target response agent, obtains a processing result and feeds back the processing result to the main agent; and the main agent receives and outputs the processing result, and the data to be processed are processed by using the large model, so that the question and answer accuracy is improved.

An interactive method based on a large model, a training method, and an intelligent agent, which relate to fields of artificial intelligence, speech recognition, speech interaction, deep learning, large models, and application scenarios of knowledge search, autonomous driving, intelligent customer service, intelligent speech control, smart e-commerce, AI healthcare. The interactive method includes: acquiring a request speech; performing a speech recognition on the request speech to obtain a speech recognition feature representing a request semantics; and processing the speech recognition feature using the large model to obtain a response text, where the response text includes response words arranged in sequence, a target response word among the response words is determined by processing the speech recognition feature and an associated response word feature using an attention fusion layer of the large model, and the associated response word feature is related to an associated response word arranged before the target response word.

The invention relates to the technical field of building construction monitoring, and discloses a construction progress real-time monitoring system based on BIM and Internet of Things, and the system comprises an Internet of Things data collection module which obtains real-time construction data through a sensor network and divides the real-time construction data into a spatial position flow and an equipment state flow; the BIM data integration module processes the two types of data streams to generate a first progress index and a second progress index; the progress deviation analysis module is combined with a preset progress reference model, and the progress indexes are fused and then mapped into a progress deviation value; the response decision module calls a preset scheme based on the deviation value to generate a target response strategy; the central processing unit is responsible for data transmission and logic verification. In addition, the system is further provided with a construction linkage module, an instruction output module and a progress storage module which are respectively used for optimizing resource scheduling, sending control commands and archiving data. According to the system, real-time monitoring, deviation analysis and dynamic adjustment of the construction progress are realized, and the intelligence and accuracy of construction management are improved.

The invention provides a cue word attack detection method and system of a large language model and electronic equipment, and relates to the technical field of text process.The method comprises the steps that a to-be-processed text block in an original text is determined; performing heuristic semantic analysis on the text block to obtain a first suspicious text block and a first risk score thereof; performing semantic analysis on the first suspicious text block and the context information thereof through a first language model to obtain a second suspicious text block and a second risk score thereof; performing structured verification of the attack intention on the second suspicious text block according to a preset meta prompt word through a second language model to obtain an attack intention verification result; the attack intention verification result comprises a third suspicious text block and a third risk score thereof; and executing a target response action on the original text based on the first risk score, the second risk score and the third risk score. According to the method, the efficient processing requirement of the long text can be considered, and the detection accuracy of complex attacks can be improved.

A method of performing a task based on a large model and an electronic device are provided, which relate to artificial intelligence technology, and in particular to fields of voice interaction, deep learning, large model, etc. The method includes: acquiring a demand feature characterizing a demand intention; performing a task by using the large model according to the demand feature, to obtain a response text, in which a target response word is determined based on: determining a query feature for each attention subtask in the task based on an associated response word feature; and performing, based on the demand feature read from a storage unit as a value feature and a key feature shared by the plurality of attention subtasks, the plurality of attention subtasks by using a computing unit according to a plurality of query.

The embodiment of the invention provides a distributed data service cluster access method and device, electronic equipment and a storage medium, and belongs to the technical field of finance and the field of medical services, and the method comprises the steps: obtaining a target business request from a target business scene and cluster configuration information of a distributed data service cluster; determining a target service cluster in a plurality of distributed data service clusters based on the target service request; searching a corresponding cluster access instance in a historical instance library according to the cluster configuration information; if the matched cluster access instance is found, determining the found cluster access instance as a target cluster access instance; if the cluster access instance is not found, creating a target cluster access instance according to the cluster configuration information; and performing data access on the target service cluster through the target cluster access instance to obtain target response data. According to the embodiment of the invention, the development complexity caused by environment change of the distributed data cluster can be reduced, so that the stability of data access is improved.

The invention relates to a multi-modal multifunctional composite metasurface design method and device and a medium, and the method comprises the steps: carrying out the parameterized sampling of the key size of a unit structure through Latin hypercube sampling, and generating a training sample set; electromagnetic simulation is carried out on the sampled structure to obtain two-state electromagnetic response data, and a mixed variable database is obtained; according to the target response, generating a plurality of groups of candidate structures through a generator in the improved conditional variation auto-encoder generative adversarial network, screening the candidate structures through a predictor, calculating a prediction comprehensive error, and selecting a preset number of candidate structures with the minimum prediction comprehensive error to perform full-wave simulation verification; and obtaining a simulation double-state response, calculating a simulation comprehensive error, and judging the current candidate structure according to a preset comprehensive error threshold value. Compared with the prior art, the method has the advantages of multi-state collaboration, low calculation cost, high degree of freedom and the like.

The invention relates to the technical field of electric power communication security, in particular to a power grid data secure transmission system and method based on quantum cryptography and a medium, comprising: a power grid data access module extracts identities of a power grid data request end and a target response end; the quantum key distribution module obtains an original shared key pre-stored by the two communication parties, and generates a quantum key according to the one-time session password and the original shared key; the communication security management module analyzes the shared key information from the quantum key received by the target response end to obtain analyzed shared key information, and performs communication security verification on the original shared key and the analyzed shared key information; and the power grid security communication module encrypts the to-be-transmitted power grid data when the communication security verification result is that the verification is passed, and transmits the power grid encrypted data to the target response end. Through cooperation mechanisms such as quantum key dynamic distribution and security verification, security protection of power grid data transmission is realized, and power grid data communication security is improved.

The invention discloses an interaction method and device, electronic equipment and a computer readable storage medium, and the method comprises the steps: carrying out the intention recognition of target demand information inputted by a user, and obtaining a target intention task; splitting the target intention task to obtain at least two sub-intention tasks; executing the corresponding sub-intention task by using the first target agent corresponding to each sub-intention task to obtain a sub-execution result corresponding to each sub-intention task; and generating target reply information fed back to the user based on the sub-execution results corresponding to the sub-intention tasks. In this way, the requirements of the user can be efficiently and accurately met.

The invention relates to the technical field of artificial intelligence, can be applied to the medical field and the financial science and technology field, and discloses a dialogue method and device based on a memory scheduling framework, equipment and a medium, which are applied to a bank virtual customer service scene or an intelligent triage and pre-inquiry scene. Obtaining a target dialogue page from all dialogue pages of the current dialogue chain in the short-term memory queue; a target topic segment and a target dialogue page are retrieved from the medium-term memory module, and target feature information is retrieved from the long-term personalized memory module; constructing a structured cue word, and generating a target response based on the structured cue word; storing the newly constructed dialogue page in a short-term memory queue; if the short-term memory queue is full, removing the dialogue page from the head of the queue; updating the topic segment in the medium-term memory module based on the removed dialogue page; and updating the long-term personalized memory module according to the popularity of the topic segment in the updated middle-term memory module. The dialogue response accuracy is improved.

The invention provides a scientific research question answering and reasoning all-in-one machine based on multi-agent cooperation and a control method, and relates to the technical field of artificial intelligence and multi-agent reasoning, and the all-in-one machine comprises a first agent which is used for carrying out the semantic understanding of a scientific research problem inputted by a user, and obtaining a semantic understanding result; performing semantic extension on the semantic understanding result according to a configured extension rule to obtain a first query request corresponding to the scientific research problem; the second intelligent agent is used for screening response information associated with the query request sent by the first intelligent agent from a pre-constructed scientific research knowledge base and sending the corresponding response information to the first intelligent agent; and the third agent is used for evaluating whether the target response information, the corresponding semantic understanding result, the first query request and the second query request meet the configured evaluation rule or not to obtain an evaluation result. The scientific research problem of the user can be accurately, comprehensively and reliably responded.

Disclosed in the present invention are a method and apparatus for synthesizing spatially cross-correlated multi-point ground motions fitting a response spectrum. The method comprises: using an influence matrix method to obtain ground motion acceleration time histories fitting a target reaction spectrum with high precision, and calculating a Fourier amplitude spectrum and a phase spectrum thereof; and considering a calculation relationship between a power spectral density function and the Fourier amplitude spectrum, taking the power spectral density function of the obtained time histories as an initial value, using a random vibration theory and Cholesky decomposition to obtain spatially cross-correlated multi-point ground motion time histories, further taking into account a phase spectrum of the ground motion time histories fitting the target reaction spectrum with high precision, and performing correction and iteration to obtain spatially cross-correlated multi-point ground motions fitting a response spectrum. The present invention overcomes the shortcomings in current multi-point ground motion time history synthesis techniques, such as the difficulty in ensuring good fitting to a target response spectrum and the non-stationarity of synthesized ground motions, helping to improve the reliability of dynamic response analysis results for large-scale engineering structures under the action of multi-point ground motions.

The invention relates to the technical field of community emergency management, and discloses a community multi-mode emergency event grading response system. The multi-modal data acquisition module acquires and divides community multi-modal sensing data, the event feature extraction module respectively analyzes and generates event feature sets, the grading evaluation module fuses features to match emergency event grades, the response decision module calls a target response scheme according to the grades, and the central processing unit verifies related parameters to generate a final response instruction. The system is further provided with a resource scheduling module for optimizing emergency resource allocation, an instruction feedback module for triggering equipment response action, and a data archiving module for storing data to form an event record chain. According to the system, multi-modal data deep processing, accurate grading evaluation and intelligent response decision making are realized, the community emergency management efficiency and the intelligent level are improved, and the community safety is guaranteed.

The invention discloses an automatic extraction method for semiconductor device model parameters, and the method comprises the steps: obtaining a multi-target response through the key parameters of a semiconductor device model, and constructing and training an agent model to predict a target response mean value and uncertainty parameters; constructing a multi-objective optimization collaborative circulation framework assisted by an agent model; generating a candidate parameter population in key parameters according to a position updating mechanism of a multi-objective optimization algorithm, predicting candidate parameters by using a trained proxy model, selecting the candidate parameters according to a prediction result to carry out real simulation evaluation, adding an obtained real sample into a non-dominated file, and updating the file based on a non-dominated relationship. Adding a training set dynamic updating agent model into the real sample, generating a potential search trajectory, selecting a representative representative Pareto approximate solution from the potential search trajectory, and updating the position of the population and the direction of the next-generation search trajectory according to the representative solution selected from the file; and terminating the circulation, outputting the solution set in the file, and checking.

The invention discloses a thinking chain controllable generation method, a related device, equipment and a storage medium, and the thinking chain controllable generation method comprises the steps: obtaining a to-be-responded statement; predicting a target thinking chain length required for responding to the to-be-responded statement; wherein the output length of the reasoning process in response to the to-be-responded statement is constrained to the length of the target thinking chain; generating a target response statement based on the target thinking chain length and the to-be-responded statement; wherein the target response statement comprises a target conclusion statement and a target thinking chain statement reflecting the reasoning process. According to the scheme, controllable generation of the thinking chain can be realized so as to ensure the output quality.

The embodiment of the invention provides a method and device for generating replies, equipment, a storage medium and a computer program product. The method includes generating a directed acyclic graph for a target problem, an intermediate node in the directed acyclic graph representing a sub-problem for the target problem. The method further includes updating the directed acyclic graph in response to one or more nodes in the directed acyclic graph needing to be updated. And according to the updated directed acyclic graph, generating a target reply aiming at the target question by using a target model. According to the method disclosed by the embodiment of the invention, whether the sub-question represented by a certain node needs to be updated can be detected, when a node needs to be updated, the directed acyclic graph can be dynamically adjusted, and the reply is generated based on the updated directed acyclic graph by using the target model, so that an accurate reply aiming at a complex question can be generated.

The invention discloses a large language model prejudice reduction method and system based on cross-model judgment, and the method comprises the steps: giving a group of multiple different LLM models, inputting preset prompts into the different LLM models, and obtaining a response generated by each LLM model; based on the composite prejudice measurement standard, evaluating the prejudice level of the response generated by all other LLM models in the same row group by using each LLM model, and obtaining a prejudice evaluation score; on the basis of the obtained prejudice evaluation scores, calculating a wave arrival score of each response by using an improved wave arrival counting mechanism, and selecting the response with the highest wave arrival score as a consensus target response which is accepted by collectively and has the least prejudice; and performing parameter fine tuning on each participated LLM model based on the obtained low-prejudice consensus target response. The invention provides a novel cross-model judgment framework for detecting and reducing prejudice in LLM by referring to concepts of human companion review and collaborative learning.

Provided are a model distillation method and apparatus. The method comprises: acquiring a fixed training dataset, wherein the fixed training dataset comprises a plurality of training text sequences; guiding, by means of a teacher model, a student model to generate target response text sequences on the basis of the training text sequences, so as to correct semantic differences between response text sequences output by the student model and response text sequences output by the teacher model; on the basis of the plurality of training text sequences and a target response text sequence corresponding to each of the plurality of training text sequences, obtaining a target training dataset; and on the basis of the target training dataset and the teacher model, performing distillation training on the student model. In the present application, semantic differences between outputs of the teacher model and outputs of the student model are marked and evaluated one by one at a sequence level, and errors are identified and corrected, such that the student model is guided to generate more reliable and diversified samples, thereby improving the training effect of the student model.

The invention discloses a cascade speculation inference method and system based on hierarchical decline KV cache compression, and the method comprises the steps: firstly inputting a context prompt text into a target model for coding, generating a KV cache, and calculating an attention score between tokens; secondly, descending sorting is carried out based on the attention scores of the final input token, KV cache blocks corresponding to the first k attention scores are selected as cascading middle layer cache, a lightweight large language model is loaded to serve as a draft model, and a hierarchical decline KV cache compression strategy is adopted to maintain draft model cache. Then, based on all the caches, a double-layer cascade speculation reasoning framework is constructed, a target reasoning path is obtained, and the caches are updated; and finally, repeating the operation until target response data corresponding to the context prompt text is output according to the target reasoning path. According to the method, the KV cache proportion is reduced, meanwhile, the draft token acceptance rate of the target model of the full KV cache is improved, and reduction of precision is reduced.

The application discloses a display panel driving method, a driving device and a readable storage medium, and relates to the technical field of display, which is characterized in that: an initial response square wave signal is converted based on a current display temperature to obtain a target response square wave signal; a target duty cycle in the target response square wave signal is used to determine a calling requirement, so as to determine a response duration corresponding to a normal output display picture of a liquid crystal display device at the current display temperature; a target response duration parameter table meeting the calling requirement is called; and the display panel at the current display temperature is driven according to the target response duration parameter table, so that the reaction speed of various pixel points to an input display signal at various temperatures is effectively accelerated, and the adverse situation that an animation picture is still blurred and ghosted due to the fact that different temperatures cannot be considered is avoided.

The invention belongs to the technical field of water traffic transportation safety and navigation mark navigation aiding, and mainly relates to a navigation mark anti-collision early warning method based on multi-sensor fusion, and the method comprises the steps: constructing a multi-sensor fusion architecture integrating an AIS terminal, a laser radar and an AI camera, and achieving the omnibearing monitoring and precise recognition of a ship in a navigation channel; a grading response mechanism is adopted, corresponding functions are dynamically started according to the real-time distance between the ship and the navigation mark, and targeted response from early warning to emergency warning is achieved; a legal evidence chain packaging mechanism is introduced, a unique ID is generated for each early warning event, formatted evidence is stored in an associated mode, and integrity and traceability of data are ensured; through a dynamic energy supply strategy, an equipment operation mode is intelligently adjusted according to a storage battery voltage threshold value, and continuous and stable work of the system under the conditions of cloudy and rainy days, night and the like is guaranteed; according to the invention, the accuracy, timeliness and reliability of navigation mark collision early warning information are effectively improved.

The invention relates to a seismic oscillation fitting method matched with a multi-damping-ratio response spectrum. According to the seismic motion fitting method, the seismic geology environment of the region is classified and the feature mother wave set is constructed, so that the seismic geology features of the region can be fully considered, and the fitting result can better conform to the actual condition of the target region and is more targeted and accurate. Through combination of characteristic seismic oscillation mother wave optimization and a dynamic parameterized particle swarm optimization algorithm, aiming at nonlinear characteristics of multi-damping ratio spectrum coupling matching, an improved particle swarm optimization PSO algorithm with a self-adaptive inertia weight and a learning factor is designed, and while the multi-target response spectrum fitting precision is improved, the multi-target response spectrum fitting precision is improved. And the time-frequency non-stationary characteristic of original seismic oscillation is reserved.

The invention discloses a flora prediction and petroleum pollution remediation method based on machine learning. The method comprises the following steps: collecting multiple groups of experimental data of a diesel oil pollution sample treated by a microbial agent, extracting environmental factors, microbial community characteristics and target response variables to construct a training data set after missing value processing, abnormal value detection and standardized pretreatment, and importing the data set into a preset machine learning model to obtain a training result; carrying out feature learning, classification training and hyper-parameter optimization by adopting a GridSearchCV method in combination with 10-fold cross validation; evaluating the correlation between a target response variable classification result and the features through a multivariable Pearson's correlation matrix, and constructing an optimal test set; and finally, selecting an optimal prediction model according to a preset index. According to the method, the model training quality is improved through data preprocessing and correlation analysis, efficient flora prediction and algorithm application evaluation are achieved by means of multiple machine learning algorithms, scientific support is provided for petroleum pollution remediation, and remediation accuracy and efficiency are improved.

The invention provides an interface dynamic simulation method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of interface testing. The method comprises the following steps: in response to a target interface of a probe injection server, the probe intercepts a target request sent by a client to the server based on at least one of a section-oriented programming technology and an interceptor technology, generates a dynamic simulation response according to dynamic configuration data through the probe, and sends the dynamic simulation response to the server, the server generates a target response according to the dynamic simulation response, and the target response is used for responding to a target request of the client; according to the embodiment of the invention, the source code of the host application program does not need to be modified, the configuration of the simulation interface and the behavior of modifying the simulation interface can be dynamically updated in real time, and the stability and safety of the application program can be improved.

The invention discloses a nuclear power unit abnormity response method and device, computer equipment and a storage medium. The method comprises: determining whether a nuclear power unit is abnormal; if the abnormality occurs, acquiring a target condition in which the major abnormality occurs; then, determining a target response organization for processing the abnormality according to the target condition of the abnormality, and determining a corresponding target person from the target response organization, so that the target person collects the fault information of the nuclear power unit according to a preset information collection list, and analyzes the fault information to obtain a fault analysis result; after the fault analysis result is received, whether the originally determined target response organization needs to be adjusted or not is judged according to the fault analysis result. Through the mode, the problem that the response speed is low due to the fact that traditional abnormal response mainly depends on manual operation and experience judgment is solved, it is ensured that when the nuclear power unit is abnormal, personnel can be rapidly and effectively organized to carry out fault processing, and the processing efficiency is improved.

The invention relates to a fine-grained target real-time image segmentation method and system based on a dynamic state modeling network, and belongs to the technical field of intelligent image processing. The method comprises the following steps: extracting multi-scale detail features by using a lightweight backbone network; through a dual-scale two-dimensional selective scanning module, the features are divided into a thin branch and a thick branch, and local scanning and global scanning are executed respectively; a dynamic cross-scale feature selection and aggregation module is adopted, redundancy is suppressed through reweighting and statistical filtering, and key target responses are highlighted; at a decoding end, local details and global semantics are fused through jump connection and an edge extractor; and finally, introducing a form-guided pseudo label hierarchical supervision strategy, and improving the structure learning ability of the model by using a coarse-to-fine morphological prior. According to the method, the segmentation precision, the boundary integrity and the tiny target recall rate of the fine-grained target under the scenes of ore separation, industrial defect detection, pavement crack recognition and the like are remarkably improved.

The invention provides an intelligent computing platform abnormal root cause positioning method fused with a large language model and related equipment, and relates to the technical field of anomaly detection. The method comprises the steps of firstly obtaining semantic expression features of interaction between a target student and a large language model, and judging whether the interaction obstacle degree exceeds a preset value or not; if yes, the feedback type of the obstacle is determined firstly, then dialogues in a preset window before the obstacle appears are searched, the starting point of the obstacle is positioned, and key semantics of the content of the obstacle in the period are extracted. And then based on key semantics, searching a target learning group and a matching dialogue of users in a set area, identifying a plurality of error types and determining a basic score, matching the error types with an obstacle feedback type to obtain a matching score, combining the error types with the obstacle feedback type to determine a basic error type, and finally modifying and generating a reply and feeding back the reply to students. By implementing the method, a closed loop from obstacle identification to root cause positioning to generation of targeted reply is realized, and the accuracy and efficiency of abnormal root cause positioning are greatly improved.

The invention discloses a dynamic interface contract mapping method based on external configuration, which comprises the following steps: obtaining routing information and mapping rules of a target interface in an external configuration source, the mapping rules comprising a request mapping rule and a response mapping rule; and dynamically registering the HTTP path according to the routing information to receive the original request data. And after original request data is received, executing a first mapping operation according to the request mapping rule, and generating first service entity data. And after the first service entity data is subjected to core service logic processing to obtain second service entity data, executing a second mapping operation according to a response mapping rule, generating target response data and outputting the target response data. Through external configuration and dynamic processing, flexible configuration and hot update of interface routing and mapping logic are realized, interface change can be adapted without code modification, the problems of high development and maintenance cost, slow response and poor flexibility of interface mapping in the prior art are effectively solved, and the system efficiency and expandability are remarkably improved.