23results about How to "Avoid fragmentation" patented technology

The application discloses a customer intention recognition method and device, an electronic device, and a storage medium. The method comprises the following steps: obtaining a data processing result of a chat record, and inputting the data processing result into a pre-configured agent group; obtaining an analysis result and a user portrait through a first agent and a second agent in the agent group, wherein the first agent is used for planning and summarizing customer emotions, and the second agent is used for extracting a user portrait; and obtaining a customer intention recognition result according to the analysis result and the user portrait. Through the application, not only basic data or simple labels can be output, but also targeted optimization suggestions can be provided for follow-up actions of customer managers based on large model agent technology, and planning schemes for subsequent customer development and customer retention are generated in combination with customer insight results, so that the follow-up direction of employees is clear, customer insight is converted into an actual business growth strategy, and customer conversion and retention efficiency is improved.

The application is suitable for the technical field of data processing, and provides a negative feedback monitoring method applied to media data, which comprises the following steps: step S1: media data collection and fusion: collecting media data and user interaction behavior data from different channels, and constructing a unified user session sequence model to time-correlate multi-modal data within a single user session; step S2: multi-modal negative feedback feature fusion: processing the media data, extracting negative feedback features of text, vision, audio and user behavior modes, and analyzing the correlation between different modal negative feedback features by using a cross-modal correlator. The application effectively avoids the problem of cross-session data fragmentation by limiting the operation range of the user session sequence model to a single continuous access period.

The application discloses an AI computing resource management method and system, relates to the technical field of computing resource management, and comprises the following steps: step S1, priority is set and the priority is adjusted; step S2, a real-time dynamic adjustment strategy is realized through an LLM scheduler; step S3, a basic quota and a shared quota are allocated, and the shared quota is allocated based on segmentation; step S4, devices with continuous numbers are forced to be allocated, high-priority tasks preempt low-priority resources, and states are automatically saved; and step S5, members and structures are automatically updated, and pre-processing is used instead of real-time topology calculation. According to the application, high-priority basic quotas and shared quotas are more, low-priority basic quotas and shared quotas are less, and interval distribution is realized, so that the resource intervals of different priorities are both layered and staggered, interval distribution is realized, fragmentation of quota resources is avoided, shared basic quota intervals are realized, resource utilization is improved, and balance between guarantee and efficiency of resource allocation is realized.

The invention relates to a weather state display method and device, equipment, a medium, a product and a chip system, and the method comprises the steps: obtaining real-time weather information in response to a triggering operation; based on the weather information, determining rendering parameters of a display object under the weather information; page rendering is conducted on the basis of the rendering parameters and the weather information, a weather state page is generated, the weather state page is displayed, and the display effect of content in a text information area of the weather state page is determined on the basis of the rendering parameters. Therefore, the display effect of the text in the text information area of the weather state page is determined based on the rendering parameters, and the display effect of the weather information can be combined with the content, such as characters, pictures, buttons and other elements, in the text information area, so that the display of the weather information in the weather state page and the display of the prompt information are interacted; and the authenticity of weather scene display is improved.

The application discloses a kind of based on end side's big model key value cache dynamic optimization method and system, method includes: S1, the context resource of multiple inference sequences is centrally controlled, and corresponding context space is pre-allocated;S2, whether the currently available context space satisfies the allocation demand of inference sequence is judged, if it satisfies, then the intelligent context allocation is carried out to inference sequence, if it does not satisfy, then whether the remaining memory space of end side equipment is enough is judged, according to the judgment result, step S3 or S4 is executed;S3, if space is enough, then the available context space is increased by dynamic expansion, and the context space allocation and context update of inference sequence are re-performed;S4, if space is insufficient, first carry out context recovery, then discard and retain context by sequence perception sliding window mechanism, and execute context update after allocation is completed.The application can realize the efficient inference of big model under the condition of end side limited resource.

The application provides a three-dimensional metallogenic prediction method based on spatial correlation correction and related equipment, relates to the technical field of metallogenic prediction, adopts a local neighborhood optimal estimation strategy to reduce the calculation complexity, uses residual errors instead of target values to compensate space, accurately captures spatial correlation information, and improves the accuracy of three-dimensional metallogenic prediction; the prediction result takes into account the non-linear fitting of attribute characteristics and the continuity of spatial distribution, avoids the problems of high-grade area fragmentation and oversmoothing, and improves the stability and spatial reasonableness of three-dimensional metallogenic prediction.

This invention provides a document segmentation method, apparatus, electronic device, storage medium, and program product, relating to the field of document processing technology. The method includes: acquiring a converted image of the current page in a document to be segmented, and extracting multiple local visual features from the current page image; performing attention fusion calculation based on the multiple local visual features and the visual feature set of the preceding page to generate multiple fused visual features containing global contextual information; inputting the multiple fused visual features and the semantic segmentation result of the preceding page into a semantic segmentation model to obtain the semantic segmentation result of the current page to be processed output by the semantic segmentation model; and segmenting the document to be segmented based on the semantic segmentation results of each page in the document to be segmented to obtain multiple segmented blocks of the document. This invention can utilize the global visual contextual information of the entire document when processing any single page, thereby improving the accuracy of document segmentation.

The invention relates to the technical field of electric digital data processing, in particular to a clustering-based real-time hotspot event discovery method and device and a medium. The method comprises the steps that if current time meets a preset triggering condition, whether an initial cluster similar to a first event text exists in an initial cluster set or not is judged for each first event text; if yes, adding the first event text to a similar initial cluster; otherwise, adding a new constructed cluster to the initial cluster set; performing density clustering on the cluster center of each cluster in the updated initial cluster set to obtain a plurality of classes; combining the clusters corresponding to the cluster center included in each class to obtain a target cluster of the class corresponding to the current time; and determining the popularity of each target cluster according to the number of the event texts included in each target cluster, and determining events corresponding to a first preset number of target clusters with the highest popularity as hotspot events corresponding to the current time. According to the invention, the efficiency of the hot event discovery process can be improved.

The invention discloses a data processing system and method based on a vehicle-mounted CAN / LIN bus hybrid application layer, and the system comprises a data collection layer which is used for receiving and transmitting a CAN message, and receiving and transmitting an LIN message; the cache management layer comprises a cache strategy and cache partitions, the cache strategy comprises cache according to a matrix table and cache index counting, and the cache partitions comprise a CAN channel partition and an LIN channel partition; the network transmission layer comprises a communication protocol and frame header filling, the communication protocol comprises a UDP (User Datagram Protocol) and CAN / LIN (Controller Area Network / Local Interconnect Network) packaging and reassembling, the frame header filling comprises a filling format and filling content, and the method further comprises a CAN / LIN message circular cache management method, a dynamic template CAN data acquisition method and a timestamp synchronous calibration method. The method has the advantages that the CAN message processing efficiency is improved, the memory occupation is reduced, the timestamp synchronization precision is enhanced, dynamic configuration and template switching are supported, and CAN / LIN bus unified management is realized.

The application discloses a kind of Cu-Ti alloy foil precipitated phase morphology and distribution regulation method, belong to copper alloy aging precipitation strengthening and phase change regulation technical field.The present application includes: Cu-Ti alloy foil is carried out high-temperature solid solution treatment and water quenching;Afterwards, Cu-Ti alloy is treated on surface, to remove the dirt, oxide and impurity on alloy surface;After surface treatment, Cu-Ti alloy is sequentially carried out cross asynchronous rolling and aging treatment, and the rolling roller ratio control is 1.15~1.16, and the aging treatment temperature is 440~460 DEG C, and the aging treatment time is 20~40min.The present application is based on cross asynchronous rolling pretreatment regulation matrix distortion organization, can optimize rolling matrix dislocation distortion field, to realize β'-Cu4Ti fine dispersion uniform precipitation, inhibit β-Cu4Ti coarse organization state of precipitation, so that Cu-Ti alloy foil strength and conductivity are synergistically optimized.

The utility model relates to curtain wall installation technical field, concretely relates to a glass curtain wall construction equipment, including mounting seat, one side of mounting seat is provided with curtain wall body, is provided with curtain wall fixing mechanism between mounting seat and curtain wall body, and the curtain wall fixing mechanism includes four groups of setting in the side of curtain wall body and is pasted and is installed arc plate, and the side fixedly connected with two groups of first sucking disc of arc plate near curtain wall body, the inside of pasted and is installed arc plate is provided with limit slot, and the inside of limit slot has the fixed disc of sliding, and the side fixedly connected with four groups of fixed tubes of fixed disc near curtain wall body, and the end fixedly connected with vacuum suction tube of fixed tube away from fixed disc, and the end of vacuum suction tube away from fixed disc is pasted with curtain wall body. The utility model cooperates mounting seat with curtain wall fixing mechanism, utilizes the cooperation of pasted and is installed arc plate and fixed disc, makes it utilize fixed disc to increase stress area on the basis of vacuum suction tube and first sucking disc to improve safety and stability.

This invention relates to the field of equipment monitoring and fault diagnosis technology, and discloses a method and system for equipment fault diagnosis based on knowledge graphs and large language models. The method includes: acquiring multiple types of data sources for the target equipment, preprocessing and extracting features to obtain a structured dataset; extracting knowledge triples from the dataset based on a natural language processing model and constructing an equipment health status ontology; constructing evaluation functions for each fault phenomenon to quantify the adaptability of monitoring methods, forming fault phenomenon-monitoring method mapping triples; merging the knowledge and mapping triples and storing them in a graph database to form a knowledge graph and optimizing it; using a pre-trained semantic embedding model to encode triples and construct a retrieval index library, receiving user query codes and retrieving and constructing a semantic subgraph; combining the semantic subgraph and the query into prompt words and inputting them into a large language model to generate recommendation results, updating the knowledge graph or index library based on user feedback. This invention improves the accuracy and adaptability of fault diagnosis and reduces operation and maintenance costs.

The application provides a front-end and back-end message processing method, device and medium. Based on a first operation of a user on a user terminal, at least one interaction request information is determined. The interaction request information is sent to a server to parse the interaction request information by the server and obtain an interaction information list corresponding to the interaction request information. The interaction information list includes at least one back-end message box, and the back-end message box includes at least an interaction information label, an interaction information attribute, an interaction prompt information and an interaction information content. Each interaction information in the interaction information list is sequentially added to an interaction information array. The interaction information array includes at least one front-end message box. The front-end message box is a pre-constructed information interaction container. Based on the interaction information array, an interaction feedback information corresponding to the interaction request information is generated. The interaction feedback information includes at least one interaction information content.

The application provides a kind of based on deep learning's commodity material classification method, system, medium and equipment, comprising: step S1: obtaining material data, carries out data preprocessing;Step S2: design and build deep learning algorithm model, and utilize the data after processing training;Step S3: the effectiveness of deep learning algorithm model is proved by optimizing deep learning algorithm model architecture and data features and carrying out comparison test;Step S4: utilize the deep learning algorithm model after optimization, complete the classification of given material.The application avoids the situation that the vector length of corresponding text of each field is too low and semantic information is split by constructing feature vector independently for the text of each field and performing feature learning.

This invention discloses a method and apparatus for crack morphology inversion combining full-time step and physical model, comprising: inputting crack height and crack offset in the crack height direction into a physical model; first, processing the crack height through a crack propagation sub-model to output crack geometric parameters for the full-time step; then, processing the crack geometric parameters and crack offset in the crack height direction through a fiber strain response sub-model to output axial strain rate at simulated measurement points for the full-time step; adjusting the axial strain rate at simulated measurement points based on positive and negative strain rate scaling factors; calculating the loss value between the adjusted simulated axial strain rate and the measured axial strain rate; if the loss value is greater than a preset loss threshold, adjusting the corresponding optimization control parameters until the loss value is less than or equal to the preset loss threshold, or a preset iteration stop condition is reached; and outputting the target crack geometric parameters for the full-time step based on the finally adjusted optimization control parameters. This invention can improve the accuracy and efficiency of crack morphology inversion.

This disclosure presents embodiments of a method, apparatus, and electronic device for water resource quality testing applied to water diversion projects. One specific implementation of the method includes: conducting water quality testing on the water area where the water diversion project is located to obtain surface water quality datasets and groundwater quality datasets; constructing a surface data structure network based on a pre-built water area simulation model and the surface water quality datasets; constructing a groundwater data structure network; conducting water pollution testing on the water area to generate surface water quality test result sets and groundwater quality test result sets; performing correlation analysis on the surface water resources and groundwater resources of the water area to generate water pollution correlation features; identifying surface water pollution areas and groundwater pollution areas of the water area; and determining the surface water quality test result sets, groundwater quality test result sets, surface water pollution areas, groundwater pollution areas, and water pollution correlation features as the water resource quality testing results. This implementation can improve the accuracy of the overall water area quality testing results.

The invention discloses a question and answer data processing method and device, and relates to the technical field of intelligent question and answer. A specific embodiment of the method comprises the steps of determining a priority of each session message in a received message sequence sent by a user, determining a message state of the session message according to the priority, and pushing a corresponding response message to a session for the session message of which the message state is a first state. According to the embodiment, the response messages are pushed according to the priorities of the session messages in a multi-round question and answer scene, the response speed of important messages is increased, and then the communication efficiency with users is improved.

The present application relates to nucleic acid compositions targeting the homologous region of the human chromosome 7q11.23 region, a method for constructing a Williams syndrome animal model and applications. The nucleic acid compositions include a first composition and a second composition, the first composition includes a first sgRNA and a second sgRNA, the nucleotide sequences of which are shown in SEQ ID NO: 1 and SEQ ID NO: 2 respectively, and the second composition includes a third sgRNA and a fourth sgRNA, the nucleotide sequences of which are shown in SEQ ID NO: 3 and SEQ ID NO: 4 respectively. By targeting the specific nucleic acid composition of the homologous region of the human chromosome 7q11.23 region, the complete genomic fragment of the Gtf2i gene and the Trim50 gene interval can be deleted at one time, and the synchronous deletion of multiple genes in the region is realized.

The invention provides an aperiodic sampling fault-tolerant control method and system for a vehicle active suspension system, and relates to the technical field of vehicle chassis control, and the method comprises the steps: building a kinetic equation of the vehicle active suspension system, and converting the kinetic equation into a state space model; based on the state space model, introducing constraint conditions of actuator saturation, tire dynamic load and suspension maximum stroke, and constructing a system output model containing the constraint conditions; and combining the system output model with an input time-lag model and a Markov jump linear system model, and reconstructing the state space model to obtain the vehicle active suspension system input time-lag model containing the random fault. According to the method, a unified data processing framework is constructed, multi-source data are integrated through the system, full-process closed-loop design is achieved, and finally efficient aperiodic sampling fault-tolerant control of the vehicle active suspension system is achieved.

The present application relates to the field of intelligent content generation, in particular to a sports event commentary real-time generation method, comprising constructing an event commentary trigger monitoring sequence, generating a commentary candidate semantic unit set, screening a target commentary semantic based on a commentary continuity constraint, determining a commentary trigger time and generating a commentary instruction, generating a commentary text based on a commentary language organization strategy, outputting and dynamically updating the commentary content, through constructing the commentary trigger monitoring sequence, introducing an instant semantic activation amount judgment mechanism and a commentary continuity maintenance mechanism, accurate identification of key changes in the event and continuity control of the commentary content are realized, so as to generate event commentary content in line with real-time commentary habits.

The utility model discloses rolling seasoning equipment for popcorn processing, which relates to the technical field of popcorn processing equipment and comprises a base, two vertical seats arranged at intervals are fixed on the base, an annular cover is arranged at the upper end of the base, the annular cover penetrates through the two vertical seats, and the annular cover is rotatably mounted on the vertical seats. A driving mechanism is arranged on the base, the driving mechanism is connected with the annular cover, limiting mechanisms are arranged at the two ends of the annular cover, and the limiting mechanisms are arranged on the strip-shaped parts and used for limiting the container. According to the rolling seasoning equipment for popcorn processing, popcorn and granular seasoning in the container are located between the adjacent ribs, the container is driven by the servo motor to rotate, so that the popcorn and the granular seasoning in the container are continuously rolled, and the popcorn and the granular seasoning can be uniformly mixed; compared with the mode that popcorn is stirred through a stirring tool, the phenomenon that the appearance of the popcorn is broken can be avoided, and the integrity of the appearance of the popcorn can be protected.

The application discloses a new low-carbon city carbon emission collaborative prediction method and system based on digital twinning, relates to the technical field of low-carbon city carbon emission prediction, and comprises the following steps: constructing a city multi-source carbon emission data access channel, collecting dynamic streaming data and static benchmark data to form an original data lake, generating normalized carbon emission element data set through standardization cleaning and space-time alignment, and driving a city-level digital twinning basic model to generate a twinning body real-time state sequence. Through the improved attention fusion algorithm introducing a learnable collaborative adjustment factor, an adaptive bias term and a cross-head attention interaction layer, multi-dimensional collaborative features are extracted, a carbon emission collaborative prediction model is used to output multi-time point prediction results, the results are fed back to the digital twinning model to carry out prediction scenario virtual deduction, and a prediction verification closed loop is formed. The method optimizes carbon emission feature extraction and prediction verification process, and improves the accuracy and rationality of city carbon emission collaborative prediction.

The application discloses an automatic driving lateral cruise simulation evaluation method, device and equipment and a storage medium. The method comprises the following steps: constructing a real vehicle-simulation test environment matched with a running design domain of an automatic driving system; collecting motion track data of real vehicles and simulation vehicles by using a unified target driving track; and performing comprehensive comparison based on multi-dimensional evaluation parameters. The method can simultaneously evaluate the lateral modeling precision of vehicle dynamics, the integration precision of a simulation tool chain and a measured automatic driving controller / algorithm, covers all factors influencing the automatic driving lateral control result, and avoids the fragmentation problem of evaluation results caused by separate evaluation in the prior art.