55 results about "Dynamic search" patented technology

The invention provides an intelligent question and answer method and system based on a knowledge graph, and relates to the technical field of knowledge graphs, the method comprises the following steps: S1, integrating multi-source heterogeneous data to construct an initial knowledge graph; s2, analyzing the questions of the user, executing single-hop query and judging whether a result meets requirements or not; if not, entering S3, performing multi-hop dynamic search based on an adaptive path extension algorithm, and generating a reasoning path and evidence; s4, calling a large language model to combine with an attention mechanism to perform semantic matching on a result or a path, and screening an optimal answer; and S5, correcting the knowledge graph and performing closed-loop feedback to the query process to realize continuous optimization. Through cooperation of the knowledge graph and the large language model and combination of dynamic path search and semantic matching, the problems that a traditional method is insufficient in knowledge coverage, low in reasoning efficiency and poor in answer interpretability are solved, the method has the advantages of being efficient, accurate and self-optimized, and the performance and user experience in a complex scene are remarkably improved.

The invention discloses a lunar radar horizon tracking method based on a dynamic search center, and solves the problems that an existing method is high in dependence on artificial experience and low in processing efficiency, and the false horizon misjudgment rate is high due to lunar soil non-uniformity interference. Comprising the following steps: performing Hilbert transform on a radar signal to extract an envelope; based on historical horizon data, predicting a dynamic search center of a current channel through Gaussian weighted average and determining a search window; screening local maximum value points in the window as candidate points, and calculating signal intensity, smoothness and edge direction scores of the candidate points; selecting an optimal candidate point as a horizon position through multi-feature linear fusion; and according to the lunar soil interface form and the signal-to-noise ratio dynamic adjustment parameters, tracking and outputting a final horizon path one by one. According to the method, historical path constraints and local signal features are fused, the identification precision and robustness of the lunar soil layered interface in the complex electromagnetic environment are remarkably improved, and the method is suitable for radar data processing of lunar exploration tasks such as Chang'e 4 and the like.

The invention discloses an automatic driving automobile obstacle avoidance trajectory planning method based on space-time corridor dynamic search in a curve scene, and relates to the field of intelligent driving and traffic safety, and the method comprises the steps: S1, information collection and reference line fitting; s2, coordinate system conversion of the vehicle state parameters; s3, generating an obstacle space in the three-dimensional space-time map S-D-T; s4, generating a subsection passable space-time corridor; s5, generating feasible track points in the space-time corridor area; and S6, generating an optimal track point. According to the method, track point sudden change and sharp turning conditions are avoided, and safety and efficiency are brought to driving.

The application discloses a kind of water electrolysis hydrogen production rectifier transformer harmonic active inhibition optimization method, comprising the following steps: constructing rectifier circuit, rectifier circuit includes switch tube, the voltage vector and zero vector of switch tube output are combined, and combined output voltage vector is obtained;The harmonic amplitude is obtained by processing combined output voltage vector, Levy tangent flight strategy and confidence domain variation strategy are introduced in black-winged kite optimization algorithm to construct improved dynamic search algorithm, the total harmonic distortion is obtained by iterative calculation of harmonic amplitude by improved dynamic search algorithm, i.e. optimal switch tube combination, the switch tube of rectifier circuit is adjusted by optimal switch tube combination, and the total harmonic distortion of rectifier circuit is reduced.The optimal switch tube combination of rectifier circuit is iteratively calculated by improved dynamic search algorithm in the application, effectively reduces resource consumption, and processing speed is accelerated, and the efficiency and accuracy of electrolytic water hydrogen production rectifier are improved.

The invention discloses an online car-hailing order allocation method and system based on an artificial intelligence algorithm, and relates to the technical field of intelligent scheduling, and the method comprises the steps: obtaining to-be-allocated order data and driver data; determining a candidate driver set by adopting a dynamic search radius and a multi-state driver screening strategy, and calculating an estimated arrival time in combination with a real-time traffic condition; establishing a multi-factor order receiving willingness evaluation model, and calculating the order receiving probability of each order-driver combination; constructing a comprehensive matching degree evaluation model, calculating the matching degree of all order-driver combinations, and generating a matching degree matrix; processing the matching degree matrix by adopting a global optimization algorithm, and outputting an order-driver allocation result; and pushing an order to a driver in the order-driver allocation result, monitoring a response state, and performing reallocation processing on a non-response condition. Through the multi-factor order receiving willingness evaluation and the global optimization algorithm, the success rate and matching precision of order allocation are improved, the order waiting time is shortened, and the overall allocation efficiency is improved.

Exemplary embodiments include systems and methods for generating dynamic searches and responses for a customer of an ecommerce store, the systems and methods comprising: a source database storing context information regarding products listed on the ecommerce store; a user interface element supported by the ecommerce store, configured to receive a prompt-query pair comprising a user query and a pre-configured prompt, and further configured to populate a contextual response to the prompt-query pair; a server coupled to the user device; and a large language model coupled to the source database and the server. The large language model is configured to: generate an initial set of prompt-query pairs using context information for products listed on the ecommerce store; receive the prompt-query pair; and generate a contextual response using the pre-configured prompt and the information regarding the product stored in the at least one source database.

The present disclosure relates to transmitting and receiving improved monitoring reports. A base station can configure a user equipment (UE) to monitor a channel / signal on multiple beams. The base station can transmit the multiple beams to the UE, which are at different time periods in a monitoring window. The UE can then monitor the multiple beams within the monitoring window and determine a monitoring report. The UE can then transmit the monitoring report to the base station, wherein the monitoring report indicates a subset of multiple beams as being more or less suitable for communication. The base station can also indicate that it will stop transmitting one or more beams and then stop transmitting the beams. If the UE does not send the report, the base station can continue to transmit the multiple beams.

The invention discloses a hard tag text adversarial sample generation method and system based on momentum and variance. The method comprises the following steps: collecting a text data set of which categories are labeled; preprocessing the data set; randomly initializing an adversarial sample, providing a global fixed word replacement strategy to improve the attack success rate, and generating a plurality of initial adversarial samples; a plurality of adversarial samples are used for searching, momentum and variance are applied to guide the searching process, a'noise adding-noise reducing 'strategy is provided to enrich the randomness of the searching, a quality evaluation index is provided to balance the quality of the generated adversarial samples, and a dynamic searching threshold is provided to reduce redundant searching; and putting forward a greedy optimization strategy, optimizing the optimal adversarial sample generated in the previous step, and obtaining an adversarial sample with higher quality. On the five classification data sets and the three reasoning data sets, the similarity between the adversarial sample generated in the hard tag black box scene and the original text is improved, and the attack success rate is guaranteed.

The invention relates to a protection section dynamic search display method and device and a medium, and the method comprises the steps: loading train information through an ATS system, recognizing a WTC vehicle, verifying the positioning effectiveness and position change of the train, and verifying the turnout bifurcation state and direction consistency in combination with bifurcation direction value parameters. Legal path equipment is screened, the section occupation state is updated, and finally accurate calculation and display of the protection section are achieved. Compared with the prior art, the method has the advantages that the real-time performance and positioning precision of train occupation display are greatly improved, scheduling errors caused by information lag or positioning fuzziness are effectively reduced, and key technical support is provided for high efficiency and safety of subway train scheduling.

The invention relates to the technical field of photovoltaic power generation control, in particular to a rapid maximum power point tracking method based on an optimization algorithm and a variable step size strategy. According to the method, an improved intelligent optimization algorithm based on biological heuristic is adopted, and the performance is optimized through an adaptive weight factor and a dynamic search strategy; acquiring a current maximum power point by utilizing a self-adaptive step length adjustment strategy based on a power change rate and combining a momentum factor and a step length reset mechanism; identifying a global optimal region through second derivative information of a power-voltage curve and a sliding window technology; and a state evaluator based on fuzzy logic automatically selects an optimal tracking mode according to the illumination change rate and the power fluctuation amplitude. According to the method, the maximum power point of the photovoltaic system can be quickly and accurately tracked under complex and changeable environmental conditions, and the energy conversion efficiency is improved.

The invention relates to the technical field of federated learning and privacy protection, and discloses a federated learning privacy protection method for dynamic search hybrid encryption and contribution value perception, which mainly comprises the following steps of: constructing a privacy protection scheme of a local evaluation and hybrid encryption mechanism combined with contribution degree perception to realize differentiated privacy protection of client privacy data; the calculation time overhead and the communication overhead based on a single homomorphic encryption privacy mechanism are effectively reduced; a central storage server access control and dynamic key rotation mechanism driven by a security token is constructed, and key non-replay and historical data non-revealing are realized by combining a searchable relevance trap door index stored by a ciphertext; according to the method, on the basis of a security aggregation algorithm of contribution perception, the contribution value of each local model in a ciphertext state is evaluated by a joint secrecy calculation protocol, aggregation screening of a global model under the condition that the contribution value plaintext is not revealed is ensured, power dispersion of a central storage server is completed, and the overall efficiency and privacy of the system are remarkably improved.

The application provides an image generation method based on federated learning and an optimal time allocation strategy, comprising the following steps: determining the sparsity according to the client communication budget and the data volume, and randomly initializing the diffusion model parameters and the personalized sparse mask; determining the neighbor clients of each client according to different optimal time allocation strategies, and determining the mask intersection parameter aggregation mode of the current client and the neighbor clients according to the current round and the aggregation probability, to obtain the aggregated initialization model; performing gradient update on the initialized model parameters after aggregation according to the data batch and the parameter mask state, to obtain the updated model parameters; and calling a gradient information dynamic search algorithm to update the sparse mask in each client based on the updated model parameters. The application can protect the training data privacy of users of each party, can jointly use the data of each party to improve the generalization ability of the model, and can efficiently allocate limited communication resources under the condition of bandwidth limitation.

This invention belongs to the fields of artificial intelligence and natural language processing. To address the problems of redundant retrieval and the inability to perform multi-step reasoning and dynamic programming in traditional retrieval enhancement generation methods, it provides a method and system for determining dynamic retrieval decision schemes based on Monte Carlo tree search. The method includes receiving an initial query question as the starting point of the reasoning process to trigger the invocation of a large language model; invoking the large language model to perform subquery generation and action selection operations; iteratively exploring the reasoning actions of the subquery question based on Monte Carlo tree search; and dynamically filtering the reasoning path by considering the confidence score and honesty score of the large language model to obtain a complete set of reasoning trajectories; constructing preference pairs for each subquery question based on the complete set of reasoning trajectories to optimize the reasoning retrieval strategy, enabling the large language model to autonomously determine the optimal retrieval time. This reduces invalid retrieval and multi-step reasoning and dynamic programming.

The application discloses a tool replacement strategy dynamic search method, comprising the following steps: constructing a tool time sequence feature extraction model; constructing a tool replacement strategy dynamic search model comprising a tool running state space and a replacement strategy space; extracting tool time sequence features through the tool time sequence feature extraction model; combining the processed multi-source production data and the extracted tool time sequence features with the tool replacement strategy dynamic search model to output tool replacement strategy results and evaluation results; continuously optimizing the tool time sequence feature extraction model and the tool replacement strategy dynamic search model based on the output tool replacement strategy results and evaluation results; and performing tool replacement strategy dynamic search by combining the optimized tool time sequence feature extraction model and the tool replacement strategy dynamic search model. The application reduces the number of unqualified products produced due to untimely tool replacement and the tool replacement time of line shutdown, and improves tool utilization, equipment production efficiency and overall efficiency of the production line.

A query for a data object search may be received. The query may include a data object attribute to be suppressed or excluded in results of the data object search. The query may be tokenized to identify the data object attribute, and transformed into a vector, where a portion of the vector corresponds to the data object attribute. A modified vector may be generated by adjusting this portion of the vector, and a plurality of data objects responsive to the query and a plurality of scores corresponding to the plurality of data objects may be determined based on the modified vector. Computer-executable instructions configured to cause a user computing device to construct and display a user interface that displays at least a portion of the data objects according to the scores as the results of the data object search may be generated and transmitted to the user computing device.

The invention discloses a low earth orbit satellite CSK signal joint capturing method based on Beidou system assistance, and relates to the technical field of satellite navigation signal processing, and the method comprises the steps: processing a Beidou satellite signal, and extracting prior information from the Beidou satellite signal; based on the prior information and pre-stored orbit parameters of a low-orbit satellite, Doppler frequency and time uncertainty of the CSK modulation signal are acquired; constructing a two-dimensional dynamic search window for searching a CSK modulation signal by taking Doppler frequency and time uncertainty as a center; and in the two-dimensional dynamic search window, performing joint capture processing on the received CSK modulation signal to determine the final code phase and the final Doppler frequency of the CSK modulation signal. The method can effectively adapt to high Doppler frequency shift and rapid change thereof caused by low-orbit satellites, and the capture success rate in a high dynamic environment and the overall reliability of a navigation system are improved.

The invention discloses an automatic boxing method based on reinforcement learning and dynamic search, and relates to the technical field of intelligent manufacturing and logistics automation, and the method comprises the following steps: establishing a time sequence observation layer from decision to execution, collecting a strategy score gradient, an action withdrawal frequency and execution energy consumption, and generating baseline data representing boxing sequence oscillation; and performing causal coherent decomposition based on the baseline data, identifying an approximate optimal path switching cluster, calculating switching frequency, staying duration and withdrawing cost, and constructing an oscillation cause map. According to the method, through construction of scoring and execution whole process monitoring, causal decomposition, disturbance playback, energy barrier modeling, a buffer execution mechanism and an inverse phase shock suppression strategy, identification, suppression and extinguishing of boxing path shock are achieved, a strategy stable closed loop is formed, the stability and execution efficiency of the boxing process are effectively improved, and energy consumption and the failure rate are reduced.

The invention discloses a satellite MCSK signal capturing method and a related device, and belongs to the technical field of satellite navigation. Firstly, an improved parallel code phase search algorithm provided by the invention is used for carrying out coarse search on an MCSK signal; wherein in the local code generation process, phase rotation is carried out on a data component local code to generate local codes of various data component code phases, so that a Doppler frequency shift coarse estimation value and a main code phase are calculated; second-level code phase search and Doppler frequency shift fine estimation are realized by using a PMF-FFT search method of double-layer matched filtering, and a dynamic search factor is introduced into signal search in a further embodiment, so that the problem of related peak attenuation in a high-dynamic Doppler scene is effectively solved. Further, the embodiment also provides a signal capture state judgment mechanism based on multi-search round capture results, and by comparing and screening the multi-moment capture results, the accuracy and stability of the algorithm on signal capture in a low-orbit and high-dynamic scene are enhanced, and the misjudgment probability is effectively reduced.

The application relates to the technical field of federated learning and privacy protection, and discloses a dynamic search hybrid encryption and contribution value perception federated learning privacy protection method, which mainly comprises the following steps: constructing a privacy protection scheme combining a contribution degree perception local evaluation and a hybrid encryption mechanism, realizing differentiated privacy protection of client privacy data, effectively reducing the calculation time cost and communication cost based on a single homomorphic encryption privacy mechanism; constructing a security token driven center storage server access control and dynamic key rotation mechanism, and combining a searchable relevance trapdoor index of ciphertext storage, realizing key non-replayability and historical data non-revealability; and the method is based on a contribution perception security aggregation algorithm, a joint secret computation protocol is used to evaluate contribution values of each local model in a ciphertext state, contribution value plaintext is not revealed, aggregation screening of a global model is ensured, power dispersion of a center storage server is completed, and the overall efficiency and privacy of the system are significantly improved.

Systems and methods for searching, identifying, scoring, and providing access to companion media assets for a primary media asset are disclosed. In response to a request for companion content, metadata within a predefined time period of a play position when the request was made, is downloaded. A dynamic search template that contains search parameters based on the downloaded metadata is generated. In response to the search conducted using the search template, a plurality of companion media assets are identified and then verified. A trust score for the companion media asset is accessed. The trust score may be analyzed and modified based on its contextual relationship to the play position of the primary media asset. If the trust score is within a rating range, then a link to access the companion media asset, or a specific segment or play position within the companion media asset, is provided.

Described is a system and method for enabling dynamic selection of a search input. For example, rather than having a static search input box, the search input may be dynamically positioned such that it encompasses a portion of displayed information. An image segment that includes a representation of the encompassed portion of the displayed information is generated and processed to determine an object represented in the portion of the displayed information. Additional images with visually similar representations of objects are then determined and presented to the user.

Described is a system and method for enabling dynamic selection of a search input. For example, rather than having a static search input box, the search input may be dynamically positioned such that it encompasses a portion of displayed information. For example, a user may touch a touch-based display using two fingers to invoke the dynamic search input and then determine a size and a position of the dynamic search input by moving their fingers on the display. An image segment that includes a representation of the encompassed portion of the displayed information is generated and processed to determine an object represented in the portion of the displayed information. Additional images with visually similar representations of objects are then determined and presented to the user.

The application relates to the technical field of path planning, in particular to a path planning method based on an improved A* algorithm and a grid and a positioning system, which comprises the following steps: constructing an abstract network model of a parking lot; finding a next node in the form of a dynamic search tree; selecting an optimal node after calculating the weighted value of the node by using a weighted heuristic function, and updating the weighted heuristic function in real time, so that the path planning of the addressing of the parking lot is realized. The application solves the problems that, when an intelligent vehicle frequently avoids obstacles and turns, the dependence on a map terrain is high, the type of obstacles is fixed, the emergency capability in the face of multiple conditions is poor, and the planning time and the occupied memory present exponential growth with the expansion of the grid map.

The invention discloses a self-adaptive cascade controller parameter self-tuning method, a medium and computer equipment, and the method comprises the steps: calculating initial parameter ranges of a cascade primary controller and a cascade secondary controller by means of an internal model controller and first-order inertia pure lag transfer functions of a primary loop and a secondary loop for a controlled system containing a primary loop object and a secondary loop object; and expanding to obtain a target parameter range. The population is initialized based on the range, a new population position is generated by utilizing a spiral position updating strategy of an original whale algorithm, a dynamic search range adjusting strategy is newly added to adjust a border crossing position, and a chaos sine movement strategy is added to correct the position during global search. And calculating an individual fitness value in iteration, and updating the optimal solution of the individual and the group. And after the maximum number of iterations is reached, assigning the optimal individual position information to the cascade primary controller and the cascade secondary controller. The whale algorithm is improved to optimize parameters, the search range is widened, the search efficiency is improved, the cascade controller parameter self-tuning is more accurate, and the system performance is better.

The application relates to a hybrid architecture expert-driven remote sensing semantic segmentation self-learning network search method, which comprises the following steps: acquiring a data set of remote sensing semantic segmentation; designing a hybrid architecture expert network as a feature encoder based on prior parameters; constructing a heterogeneous coding dynamic search space, adaptively searching and learning to fuse multi-granularity visual features of the hybrid architecture expert; constructing a coarse-fine granularity representation search space, adaptively searching and learning to fuse the multi-granularity visual features; adopting a step-by-step up-sampling strategy from bottom to top, adaptively searching and learning to restore the multi-granularity visual features from coarse to fine to the original resolution of an image; outputting a semantic segmentation mapping result, and adopting an accuracy evaluation to evaluate the performance of the network. The application fuses multi-architecture and hybrid expert visual features and multi-granularity visual features, constructs a remote sensing image semantic segmentation network with richer feature expression, and enhances the adaptability to multi-scale characteristics of remote sensing ground objects.

The application discloses an automatic packing method based on reinforcement learning and dynamic search, and relates to the technical field of intelligent manufacturing and logistics automation, and comprises the following steps: a time sequence observation layer from decision to execution is established, strategy score gradient, action withdrawal times and execution energy consumption are collected, and baseline data representing packing sequence oscillation is generated; based on the baseline data, causal coherence decomposition is carried out, an approximately optimal path switching cluster is identified, switching frequency, residence time and withdrawal cost are calculated, and an oscillation cause atlas is constructed. Through the construction of score and execution whole process monitoring, causal decomposition, disturbance playback, energy barrier modeling, buffer execution mechanism and inverse phase shock suppression strategy, the application realizes the identification, suppression and extinguishing of the packing path oscillation, forms a stable closed loop of strategy, effectively improves the packing process stability and execution efficiency, and reduces the energy consumption and failure rate.

The invention provides a text and question conformity evaluation method, system and device and a medium, and the method comprises the steps: obtaining a target composition, extracting title subject terms from a title of the target composition according to a keyword extraction algorithm, and calculating subject degree weights of words in a text of the target composition; according to a preset composition subject term extension knowledge base, subject-related word extension is carried out on the title subject term, and the extension knowledge base carries out keyword indexing on the corpus based on the target composition and generates a subject-related word cluster by adopting word clustering; according to the full-text word number of the target composition, text subject words of the composition are extracted according to a preset dynamic rule; generating a weighted topic degree vector according to the topic degree weights of the words, and calculating the cosine similarity between the topic words of the title and the topic words of the main body of the composition to obtain a topic conformity score; selecting an optimal threshold value from the candidate threshold value range according to a dynamic search algorithm; and outputting a text and question conformity grade of the target composition according to the text and question conformity score and an optimal threshold.

The application discloses a data processing method and device for an energy storage battery and a nonvolatile storage medium. The method comprises the following steps: acquiring real-time operation data of a target energy storage battery at a current time; extracting an operation state of the target energy storage battery based on the real-time operation data; receiving a query instruction sent by a user; determining a target data source based on the query instruction; searching in the target data source based on the operation state to obtain target candidate data; constructing a structured enhanced prompt word based on the target candidate data; inputting the structured enhanced prompt word into a preset language model to obtain a data processing result based on the query instruction and the real-time operation data. The application solves the technical problems that the current energy storage system operation and maintenance method fails to integrate real-time operation states and multi-source historical knowledge, lacks an intent-driven dynamic search mechanism, and causes edge analysis response lag, high diagnosis misjudgment rate and natural language interaction failure.

The present invention discloses a Parkinson's speech recognition method and device based on a PSO dynamic search classification network. The method comprises the following steps: acquiring speech data of a Parkinson's patient, editing the speech data, extracting speech features from the edited speech data, and preparing a speech feature data set; initializing parameters of a dynamic search classification network, and generating a classification network particle swarm by random initialization; training a classification network in each generation of the classification network particle swarm using the speech feature data set to obtain a fitness function value of each particle, updating the classification network particle swarm according to the fitness function value of each particle, cyclically training and updating the classification network until an optimal classification network is obtained, and performing Parkinson's speech recognition on the speech feature data to be recognized based on the optimal classification network obtained through training.

The application discloses a kind of based on artificial intelligence algorithm's online car-hailing order distribution method and system, it is related to intelligent scheduling technical field, including obtaining order data and driver data to be distributed;Adopt dynamic search radius and multi-state driver screening strategy to determine candidate driver set, the estimated arrival time is calculated in combination with real-time traffic condition;Establish multi-factor order acceptance willingness evaluation model, calculate the order acceptance probability of each order-driver combination;Build comprehensive matching degree evaluation model, calculate the matching degree of all order-driver combinations, generate matching degree matrix;Global optimization algorithm is handled matching degree matrix, and the order-driver distribution result is output;Order is pushed to the driver in order-driver distribution result and monitored response state, and re-distribution processing is executed to unresponsive condition.The application improves the success rate and matching accuracy of order distribution by multi-factor order acceptance willingness evaluation and global optimization algorithm, reduces order waiting time, and improves overall distribution efficiency.