45 results about "Online search" patented technology

The invention discloses a three-level ANPC converter model prediction control method and device based on Newton iteration method vector optimization, and belongs to the technical field of converter control, and the method comprises the steps: building a mathematical model of a three-level ANPC converter under a three-phase static coordinate system based on the Kirchhoff theorem; performing forward Euler discretization on the mathematical model of the three-level ANPC converter to obtain a current prediction equation model; establishing a cost function according to the control target; performing online search on the optimal voltage vector according to a cost function; voltage vector synthesis is carried out, voltage vectors selected when the optimal voltage vector falls in 36 different small sectors, a voltage vector action sequence and action time distribution are obtained through calculation, and then the action time of a pair of redundant positive and negative small vectors in the voltage vectors is adjusted to carry out neutral point potential balance control. According to the invention, rapid optimal voltage vector search and three-level ANPC converter output current and neutral point potential control are realized.

The invention discloses a position searching method and device based on implicit intention recognition, and relates to the technical field of information retrieval and intelligent recommendation. The method comprises the following steps: sampling from a historical log and constructing prompt words, and performing weighted voting labeling on keywords and resume pairs by utilizing a plurality of large language models to generate an intention recognition training sample; and the cue word is iteratively optimized through manual evaluation, so that the labeling accuracy is ensured. And training an intention recognition model by using the samples. In the online search stage, structural features of search terms and resumes of a user are extracted and input into the pre-training model to recognize the implicit job hunting intention of the pre-training model. Finally, personalized retrieval conditions are dynamically generated according to the intention recognition result, and related positions are recalled and sorted from the position database. According to the method, the accuracy and the individuation degree of position searching are effectively improved, and the user experience of job seekers is remarkably improved.

Today neural networks are used to enable autonomous vehicles and improve the quality of speech recognition, real-time language translation, and online search optimizations. However, operation of the neural networks for these applications consumes energy. Quantization of parameters used by the neural networks reduces the amount of memory needed to store the parameters while also reducing the power consumed during operation of the neural network. Matrix operations performed by the neural networks require many multiplication calculations, so reducing the number of bits that are multiplied reduces the energy that is consumed. Quantizing smaller sets of the parameters using a shared scale factor improves accuracy compared with quantizing larger sets of the parameters. Accuracy of the calculations may be maintained by quantizing and scaling the parameters using fine-grained per-vector scale factors. A vector includes one or more elements within a single dimension of a multi-dimensional matrix.

The invention discloses a control method of a flat wire motor. The method comprises the following steps: constructing an optimal current instruction table taking a rotating speed and a torque as indexes; if the motor is in the dynamic working condition, directly looking up a table to obtain a current optimal current instruction; if in a steady-state working condition, performing online search by taking a table look-up result as an initial point: generating two-dimensional Bernoulli random disturbance, and sequentially applying positive and negative disturbances to the initial point; respectively measuring the input power under the two disturbances; according to the ratio of the positive and negative input power difference to the corresponding disturbance component, synchronously calculating gradient estimation vectors of the system input power to d-axis and q-axis current; according to a steepest descent method, the current instruction is updated in the reverse direction of the gradient, an optimal instruction enabling the input power to be minimized is obtained, and an inverter is driven to control a motor to operate. According to the method, through a hybrid control strategy combining dynamic table look-up and steady-state online search, the total loss minimization of the flat wire motor under all working conditions in consideration of alternating current copper loss is realized, and the rapidity of dynamic response is ensured.

The invention belongs to the field of artificial intelligence, particularly relates to a multivariable self-optimization control coal mill agent operation method, and aims to solve the problems of variable coupling interference, parameter solidification lag and poor working condition adaptability in traditional control. According to the strategy, an attention mechanism-based LSTM dynamic model is constructed, and future states of six types of variables such as air volume and temperature are predicted; defining a weighted optimization function fusing powder making efficiency, energy consumption, wear and stability, and introducing a soft constraint to guarantee a safety boundary; an optimal control combination is searched on line every five seconds by adopting a depth deterministic strategy gradient algorithm, and the four-dimensional action space of an air door, a hot blast valve, a coal feeder and a grinding roller is covered; and triggering an incremental learning fine tuning model through residual feedback to prevent disastrous forgetting. According to the scheme, multivariable cooperation, online self-optimization and intelligent safety control under all working conditions are realized, and the dynamic response capability and the comprehensive operation efficiency of the coal mill are remarkably improved.

This invention discloses an online search method for particle spelling variants and properties usable by intelligent agents. The steps are as follows: inputting the baseline spelling of a particle and the constructed particle-specific prompt words into a large language model to generate a core variant corresponding to the particle; integrating the core variant of each particle with the authoritative benchmark information corresponding to the particle to form a full variant set for each particle; each record in the full variant set contains a unique identifier for a particle, a multi-scene benchmark name, a core variant, and extended attributes; dynamically mapping a remotely callable interface set on the server to a local proxy object on the client; after receiving a query request, the local proxy object parses the particle spelling parameters in the query request and performs multi-dimensional spelling matching in the full variant set to realize online search for particle spelling variants and properties. This invention effectively improves the ability of AI intelligent agents to handle domain-specific tasks.

A method may include receiving data related to a plurality of items and processing the data using a machine learning model. The machine learning model may have been trained to output a score for each of the plurality of items based on one or more target variables and to process the data using a grouped linear regression for groups of items based on sub-divisions of the groups. The method may include storing the output in a data store. Each entry in the data store may include at least an item identifier for an item, a group name, and the score. The method may include receiving search criteria for a search and identifying a set of search results in a group of items. The method may include determining an order of the set of search results and outputting the set of search results.

The invention provides a historical character knowledge proofreading method and system, a storage medium and electronic equipment. The historical character knowledge proofreading method comprises the steps of constructing a first historical character knowledge database and a second historical character knowledge database; obtaining a to-be-proofread text, and performing assertion detection on the to-be-proofread text to obtain an assertion detection result; obtaining an identification field in the to-be-proofreading text based on the assertion detection result; according to the identification field, the corresponding historical figure is retrieved in the first historical figure knowledge database online, and related reference information of the historical figure is retrieved in the second historical figure knowledge database online; and obtaining a proofreading result of the historical character knowledge based on the historical character, the related reference information and the to-be-proofreading text. According to the method, the historical figure knowledge database is constructed offline, and in the online proofreading process, the historical figures are determined through the identification fields, and then the corresponding related reference information is retrieved, so that the problem that the retrieved information is redundant and irrelevant is effectively solved, and the proofreading efficiency and accuracy are improved.

The invention provides an agricultural machinery equipment identification method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining an agricultural machinery equipment target image, carrying out the first type and second type information labeling of the target image according to the standard of agricultural machinery classification (NY / T1640-2021), and building a target image data set; according to an SIFT + PCA data preprocessing method, performing feature extraction and screening on the target image data set to obtain a target image feature data set; training a two-layer classification model corresponding to the target image feature data set through an SVM algorithm; inputting a target image online; and according to the SVM two-layer classification model, analyzing the preprocessed online target image feature data, and identifying the agricultural machinery equipment type online. According to the method, agricultural machinery equipment of different types and different models can be quickly and accurately identified, and reliable technical support is provided for online search, consultation, popularization and application of agricultural machinery technology and equipment thereof.

The present invention discloses an online automatic tuning and compilation system and method for large-scale model training, which belongs to the field of computing framework programming and compilers. The system includes an automatic tuning client, an automatic tuning server, a configuration manager, a Triton operator register, a communication manager, an automatic tuning scheduler, and a data collector. The present invention improves the computing performance of the fusion operator by searching online for the optimal hyperparameter configuration of the fusion operator during large-scale model training, thereby improving the efficiency of utilizing GPU hardware resources during large-scale model training. The online automatic tuning and compilation system of the present invention provides a search for a larger hyperparameter space, and quickly finds the optimal hyperparameter configuration of the fusion operator for end-to-end performance during large-scale model training.

The invention relates to an automatic method for technical assistance in the correction of vulnerabilities of a computer system, where the aforementioned method comprises at least the following steps: a) receiving information relative to computer system vulnerabilities by text, voice, file input or through data exchange; b) launching tools based on artificial intelligence algorithms and machine learning to understand the information and requests entered relative to vulnerabilities; c) performing parsing and text mining of the information received and process it through intelligent data matching with information relative to the computer vulnerabilities present in a database or through interaction with external online vulnerability databases; d) in case of unsatisfactory results, starting an extended online search by web scraping using deep learning algorithms and unsupervised machine learning; e) updating an internal vulnerability database with the additional information found; and f) generating a vulnerability report accompanied by relative remediation.

This invention discloses an automated tuning method and system for large language model service configurations, belonging to the fields of artificial intelligence and cloud computing technology. The invention includes an offline analysis module for performance profiling of the large language model (LLM) service system under a finite configuration set and for collecting training data; an online search module for exploring and evaluating the performance of different configurations in the configuration space based on a Monte Carlo tree search algorithm; and a simulator module for receiving the configurations generated by the online search module and predicting the performance metrics of the configurations under a target workload based on a logical step simulation model. This invention effectively solves the problems of low search efficiency and poor performance of existing automatic tuning methods in existing large language model service frameworks (such as vLLM, SGLang, etc.) due to the huge number of configuration parameters (reaching a petabyte-scale search space) and the complex dependencies and mutual exclusion relationships between parameters.

Systems and methods for product recommendation are described. One aspect includes receiving a user selection of a first item, and a user request for an online search based on the first item. The online search may include at least one modification to at least one property of the first item. The online search may be conducted, and one or more results of the online search may be presented to the to the user. The results may include at least a second item that differs from the first item with respect to the at least one property.

The invention discloses a maximum aftershock prediction method based on online enhancement, and belongs to the technical field of earthquake prediction. The core of the method is that a personalized prediction model is dynamically generated for a current event through online retrieval of similar historical sequences and data enhancement aiming at a newly occurring main earthquake. Firstly, a historical seismic database is constructed from a global seismic directory, and an initial prediction model is trained. When a new main earthquake occurs, an online process is triggered immediately: based on the characteristics of the new main earthquake, K most similar sequences are quickly retrieved from historical data; then, the similar sequence data are enhanced (such as noise adding and interpolation), so that a diversified enhanced training set is constructed; and finally, performing rapid fine adjustment on the initial model by using the enhancement set to obtain an optimization model for the current main earthquake, and predicting the maximum aftershock magnitude. According to the method, the limitation of one-time cutting of the existing static model is overcome, the problem of inaccurate prediction in an earthquake small sample scene is effectively solved, and the accuracy, robustness and timeliness of prediction are remarkably improved.

The present invention relates to the field of machine learning technology, and in particular to a large language model capability assessment method and system based on dynamic data assessment. The method comprises: obtaining a question input by a user and extracting core knowledge points and main content from the question; based on the core knowledge points and main content, using a pre-trained large language model to perform an online search and generate a knowledge description related to the question; generating an assessment question based on the core knowledge points, main content, and knowledge description; performing difficulty control and optimization on the assessment question to obtain a final assessment question; performing multi-dimensional capability assessment and quality testing on the final assessment question, obtaining an assessment result, and completing the capability assessment of the large language model. The present invention can ensure the quality and consistency of assessment data and improve the reliability and fairness of LLMs capability assessment.

The invention discloses an optical fiber vibration sensing differential step length adaptive optimization method based on a DE algorithm. The method comprises the following steps: establishing an objective function between a difference step length and a signal-to-noise ratio of a vibration signal, and introducing a differential evolution (DE) algorithm to perform online search and update on the difference step length. And in a constrained step size space, carrying out iterative evaluation and screening on the candidate difference step size by utilizing variation, crossover and selection mechanisms of a differential evolution algorithm to obtain a difference step size parameter which enables the signal-to-noise ratio to be optimal. Compared with a traditional fixed step length and experience parameter adjustment method, the method has the advantages that dependence on artificial experience can be effectively reduced, local optimum is avoided, the signal-to-noise ratio of the vibration signals is increased, and meanwhile high calculation efficiency and robustness are achieved.

Systems and methods for product recommendation are described. One aspect includes receiving a user selection of a first item, and a user request for an online search based on the first item. The online search may include at least one modification to at least one property of the first item. The online search may be conducted, and one or more results of the online search may be presented to the to the user. The results may include at least a second item that differs from the first item with respect to the at least one property.

A unified item recognition driver is presented. The driver manages a plurality of disparate item scanning capable peripheral devices of a transaction terminal and computer vision item recognition applications on behalf of a transaction manager of the terminal. The driver prioritizes item codes provided by the devices and applications to select a highest ranked item code for any given item of a transaction. The item codes are provided to the transaction manager of the terminal to complete the transaction. In an embodiment, the driver searches online to resolve an item code when the peripherals and applications fail to resolve the item code for any given item.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for distributing content are disclosed. In one aspect, a method includes storing, in a data structure, data specifying a future live event. An opportunity to provide the specified content to a user at a user device is identified. It is determined that (i) a time of the opportunity is between a start time and an end time for the live event, (ii) that a user device is located in a same geographic region as the live event based on geographical data provided by the user device, and (iii) the user is interested in the live event based on attributes of the user matching attributes of other users that were identified as interested in the live event (e.g., based on evaluation of online search data). The content is provided for display at the user device.

The invention discloses a motor optimal efficiency control method based on minimum loss online search, which aims at the over-loss problem of a servo drive system, establishes a novel full frequency domain harmonic loss model through double Fourier analysis, and can effectively reduce torque ripples by combining a minimum loss fusion algorithm and a golden section online search algorithm. The control precision is improved, dynamic optimization of energy efficiency and minimization of system loss can be achieved under different working conditions, and the system control efficiency and the overall energy efficiency performance are remarkably improved.

The invention discloses a collaborative search method and system based on heterogeneous unmanned aerial vehicles, and the method comprises the steps: constructing a dynamically updated target probability map based on the last known position, loss time and maximum moving speed of a target; establishing an area uncertainty map reflecting the unknown degree of an undetected area, and generating a search difficulty map in combination with prior information such as terrain, obstacles and illumination; distributing a search area according to the initial position of the unmanned aerial vehicle, the sensor capability and a search difficulty map; executing a search task, and updating the target probability map and the regional uncertainty map; and dynamically planning the next waypoint based on the updated target probability map and the regional uncertainty map. The system comprises a map construction module, a task allocation module and an online search module. By using the method, various prior intelligence information and real-time detection information can be fully utilized, and the search efficiency is effectively improved. The method can be widely applied to the field of unmanned aerial vehicle planning.

Systems and methods for measuring impact of online search queries on user actions. The method includes capturing clickstream data entered via a website, the clickstream data including text-based queries associated with web searches, and clustering the queries to generate query clusters. The method also includes assigning each query cluster to an intent such that each assigned intent estimates a desired action behind the queries in the corresponding query cluster. The method further includes mapping each intent assigned to a query cluster to at least one action motivated by the intent. The method also includes computing metrics using the mapping to quantitatively measure the impact of the queries on the mapped actions by tracking performance of the actions within a predefined time period after the queries.

The present invention discloses a Chinese text error correction method based on online search assistance. First, the sentence to be corrected is segmented, and an online query is performed through a search engine to crawl and count the word frequencies to construct a word frequency table. Then, the original sentence is tokenized, and suspicious words are detected according to the word frequencies and perplexities of the tokenized results in the word frequency table. The suspicious words are searched based on the context of the original sentence and the context information retrieved from the search engine. The topsis algorithm is used to score according to the word frequency, pinyin edit distance, and structural similarity to form candidate words, and some near-sound and near-shape words are also added as candidate words. The candidate words are used to replace the suspicious words in the original sentence, and the perplexity is calculated using the original GPT-2 model. The sentence with the minimum perplexity is selected as the final corrected result. This method can automatically correct Chinese sentences that may contain errors immediately without the need for additional training of an error correction model or a large-scale data set.

The invention relates to the technical field of cloud computing, and particularly provides a cloud storage file searching method and device. The method comprises the following steps: collecting a plurality of cloud storage files uploaded to a cloud disk by a target user; wherein each cloud storage file comprises at least one type of modal data; performing feature extraction on each modal data to obtain a modal feature vector of each modal data in each cloud storage file; performing index construction based on a fusion feature vector of each cloud storage file determined by the modal feature vector of each modal data to obtain an initial offline index of the plurality of cloud storage files; dynamically updating the initial off-line index by utilizing the dynamic interest characteristics of the target user to obtain personalized on-line indexes of the plurality of cloud storage files; and providing a search service for the target user by using the personalized online index. According to the method, the response speed of online search can be increased, the search result can be more targeted and practical, and the user satisfaction and stickiness are enhanced.

The application provides a training method of a search relevance model, a relevance evaluation method and related devices. The method comprises: inputting a sample into a plurality of teacher models to obtain a basic relevance score output by each teacher model; inputting the sample, a search term type and a teacher model scoring confidence into a gating model to dynamically obtain the weight of different teacher models; weighting and fusing the basic relevance scores output by the plurality of teacher models based on the weight output by the gating model to obtain a target relevance score; inputting the sample into a student model to train the student model with the target relevance score as a target; and adding an entity extraction auxiliary task to the teacher model and extracting key attributes by using an external knowledge graph, and improving the attention weight of the key attributes through a post-interaction link when training the student model. The student model serves as an online search relevance model. The application can improve the relevance discrimination ability of search terms and commodities in a search system to a certain extent.

Embodiments of the present application provide an end-side model inference dynamic optimization method, device, equipment and medium. The method surrounds the end-side single-flow CPU inference path, establishes a performance, energy consumption and thermal state measurement system covering the prompt processing stage and the self-recurrence generation stage, and adopts a fixed word window unified index statistics and configuration decision. The offline measurement is performed on the CPU frequency, thread number and affinity configuration space, the feasible boundary and preferred priori are refined, and a discrete neighborhood search dynamic optimization method under the offline priori constraint is proposed. Thus, taking a unified comprehensive cost function as an optimization target, the online search is limited in the feasible neighborhood obtained through offline screening under the first word delay, thermal safety and frequency accessibility constraint, and combined with the minimum residence, anti-shake, limited memory and controlled jump-out mechanism, the low-overhead joint adjustment of the CPU frequency and thread number is realized, while the adaptability to state changes is improved.

This invention discloses a snow ablation-optimized weak-sensitivity Kalman filter method and system for UAV state estimation, belonging to the field of UAV autonomous navigation and state estimation technology. The method includes: obtaining the posterior state estimate, posterior covariance matrix, and posterior sensitivity matrix of the previous time step; obtaining the prior statistics of the current time step based on ensemble Kalman filtering with ensemble point sampling and prediction propagation; searching online for the optimal scaling factor using the snow ablation optimization algorithm and scaling the preset sensitivity weight matrix; calculating the Kalman gain based on the weighted prior sensitivity matrix; updating the posterior state estimate using measurement data and outputting the flight state estimation result. This invention, while retaining the advantages of fast analytical calculation of the weak-sensitivity ensemble Kalman filter gain, achieves online adaptive adjustment of the sensitivity weights, improving the accuracy and robustness of state estimation in dynamic flight environments, and has low computational overhead, making it suitable for UAV airborne navigation systems.

Online content is received in a user computer over the public Internet. A target text of the online content is selected to check for misinformation. A search query is generated using the target text. An online search is performed using the search query to generate a first set of search results. Another online search is performed using an opposite search query, which conveys a meaning that disagrees with the target text, to generate a second set of search results. Misinformation in the online content is detected based on how much the target text aligns with collected online contents from the first and second search results.

The present application relates to a kind of container assets precision marketing method and system, the method is based on the original method acquisition basis data, increase the online order of customer in trade platform, online search data, and utilize the customer information data builds a bottom database;And according to customer data information attribute constructs a new unified customer label system, to realize the precision group of customer, and customer group and individual insight;In addition, through data interface, label and crowd package are provided to other data products to realize risk control management, marketing management and trade platform management and other customer label scene applications.The present application can help container leasing enterprise to comprehensively understand customer data through data analysis, while realizing the precision analysis of specific customer, to effectively propose marketing scheme and improve customer activity, at the same time, can assist container leasing enterprise to identify and control risk, formulate personalized marketing strategy suitable for different customer groups, realize precision marketing.

A scholar system and method transforms search of academic knowledge. The system uses a combination of face recognition and language generation technology to effortlessly capture scholars in photos to give rich summaries of profiles of their work. The system also offers an interactive chat element for users to get more insights about a scholar, reducing online search efforts. The system ensures privacy and ethical compliance to rules and regulations.