39 results about "Association strategy" patented technology

The invention discloses a target movement situation information data association strategy-based target tracking method. The method includes the following steps that: step 1, the track of a target to be selected is created; step 2, tracking processing is carried out on the track of the target to be selected, a reasonable track of a formal target is established according to a target starting condition; and step 3, a Kalman filtering model is used to carry out filtering estimation processing on the tracks of a tracked target and the formal target, so that a track state estimation value can be obtained, the time to closest point of approach (TCPA) of the tracked target and the formal target is calculated according to a ship collision avoidance radar measurement model and a TCPA value is compared with a target separation time threshold, a target approach time threshold and a target overlap time threshold, so that the movement state information of the targets can be obtained; and step 4, and the association strategy and associated algorithm of the trace points and tracks of the tracked target are determined according to the movement state information of the targets, and the association relation of the tracked target and the trace points can be obtained.

The invention discloses an echo fused target track association method based on track state estimation.The method comprises the steps of conducting polar coordinate sector partitioning, conducting track starting/updating, conducting track prepushing and associated sector calculation, and judging the state of the prepushed track in an associated sector according to a corrected target fusion threshold and setting the state of the track as non-fusion, pre-fusion or fusion.Different track association strategies are adopted for the three track states; track-trace point association is conducted on the non-fusion track with the conventional nearest neighbor method; track association is conducted on the pre-fusion track with the target feature information assisted nearest neighbor method; for the fusion track, within the fusion wave gate range, feature information is used as the main basis for track association after echo defusion, the fusion track and a track cluster are updated in real time, and the fusion track and the track cluster are deleted if the number of tracks in the track cluster is lower than 2.By the adoption of the method, the probability of mis-tracking or track losing after target fusion can be reduced, and tracking stability can be improved.

The invention discloses an implicit discourse relation recognition method based on multi-granularity generated image enhancement representation, and provides a multi-granularity generated image and aneural network for enhancing argument vector representation by simulating an association strategy for the first time due to the problems of ambiguity, fuzziness and the like of texts. Corresponding images according to different granularities (sentence level and phrase level) of texts are introduced, which helps understand semantics of chapters. In order to better capture context information of a text image, text and image features are integrated according to the sequence information of the text .Important image-text information and interaction information are captured in an image-text vector sequence representation whole formed by splicing two arguments by utilizing a self-attention mechanism; therefore, argument vector representation is further enriched, feature vector representation usedfor recognizing the discourse relations is obtained, and finally the feature vector representation used for recognizing the discourse relations is input into the discourse relation recognition layerfor discourse relation recognition.

The invention discloses a mass data distributed storage method. A data storage area is divided into a data area, a copy area and a coding area; the data area is used for storing source data; the copy area is used for storing copy data; the copy area and the data area form a multi-copy full backup disaster recovery system; the coding area and the data area form a complete information association disaster recovery system. Compared with the prior art, the invention is advantaged by that on the basis of mass data storage cost, efficiency and load equalization, a new method combined with a data information association strategy and a dynamic copy strategy is proposed; the cluster data security is ensured, at the same time, the data recovery speed of the system is remarkably improved, the load equalization capability of the system is optimized, and the integrated storage cost of the system is reduced.

The invention provides a product data packet management system, which performs cross-domain collaboration on data packets of a product and process driven and closed-loop management on the data packetacceptance checking process in a non-networked environment. The system comprises a data packet collection subsystem and a data packet acceptance checking and managing subsystem, wherein the data packet collection subsystem comprises a data packet template management module, a data packet collection module, a data packet import and export module and a site and user authority management module and is used for collecting data packets and establishing an association strategy among a site, a user and data so as to realize strong association among the site, the user and the data, thereby realizing effective transmission and collaboration of data among a single-machine unit, a subsystem unit and an overall unit in the non-networked environment; and the data packet acceptance checking and managingsubsystem comprises a data packet import and export module, a data packet parallel acceptance checking module, a to-be-handled closed-loop management module and a data packet application module and is used for performing process-driven and closed-loop management on the acceptance checking process of the data packets collected by the data packet collection subsystem.

The invention provides a resource allocation method and device based on reinforcement learning in an ultra-dense network, which can realize load balance of the network and improve the energy efficiency of the network. The method comprises the following steps: analyzing the current state of a network based on Q learning to obtain an association strategy of a user and a base station and a transmitting power control strategy of the base station when the energy efficiency of the network is maximized; associating the user with the base station according to the obtained association strategy of the user and the base station when the network energy efficiency is maximized; and on the basis that the user is associated with the base station, controlling the transmitting power of the base station inthe network according to the obtained transmitting power control strategy of the base station when the network energy efficiency is maximized. The invention relates to the technical field of communication.

Aiming at access management problems and epidemic prevention problems displayed in the epidemic situation period and the whole society, split access place management is changed into a system driven by dynamic data and risk data, and each access manager is assisted to provide necessary real-time data for superior epidemic prevention, command or big data units; and each access person actively or passively learns the individual static data, dynamic data and risk data by the access management personnel during access code scanning access application, and decides access or forbidding or local control and the like of the access person according to the attributes, admission, export and association strategies (such as notification and report to related units) of the area.

The embodiment of the invention discloses a communication method based on a social identity, which can be applied to a third-party server. The communication method comprises the following steps: obtaining a first social identity and a first communication identifier corresponding to the receiver terminal; selecting an information association strategy corresponding to the receiver terminal; establishing an association between the first social identity and the first communication identifier according to the information association strategy corresponding to the receiver terminal to obtain the first associated information; and controlling the third-party server to establish a communication link between the sender terminal and the receiver terminal by using the first associated information, wherein the communication link is used to characterize a communication link established between the sender terminal and the receiver terminal through the third-party server. The embodiment of the invention also provides a server.

The invention discloses a virtual-medium-based material storage management method and system. The method comprises the following steps that: introducing high-quality audio-video file paths and material metadata information of materials through invoking a material import interface by an external system; copying high-quality audio-video files of the materials into a high-quality storage subsystem; converting the high-quality audio-video files of the materials into audio-video stream media files, storing the audio-video stream media files in a stream media storage subsystem, and associating high-quality files into a virtual medium according to an association strategy; determining whether the virtual medium is fully filled with the high-quality files, if yes, writing contents in the virtual medium into a physical medium by a read-write device; storing the physical medium in a strip rack; and finding out a required physical medium by a user through browsing stream media so as to take back the stream media physical medium for use. According to the method, material storage, access and browsing efficiencies are improved to a large extent, the management efficiency and work efficiency of media contents are improved, and the system cost is greatly reduced. In addition, the invention discloses a system for implementing the method.

The embodiment of the invention relates to a dialogue processing method and device, a server and a storage medium, and the method comprises the steps: determining an association strategy correspondingto a first text in a first text display process; receiving an input second text, wherein the second text is associated with the first text; associating the second text with the first text by adoptingthe association strategy, and displaying the associated first text and second text. According to the method, the dialogue message can be associated and the associated message can be integrated and displayed, the dialogue message can be logically and clearly displayed, the reading of a user is facilitated, and the user experience is improved.

The invention discloses a reservoir dispatching system automatic calculation method based on information association. The method comprises the steps of firstly analyzing an association relation among calculation quantities of a reservoir dispatching system, defining and configuring the calculation priority of each calculation quantity of the reservoir dispatching system, performing configuration according to characteristics of the calculation quantities and the association relation among the calculation quantities for forming an automatic calculation information association strategy, and finally performing automatic calculation according to the information association strategy, the priority sequence and a corresponding algorithm which is bond with the calculation quantities, and storing a calculation result into a corresponding list of a database. According to the reservoir dispatching system automatic calculation method, on the condition that partial acquired information are discontinuous, because the automatic association relation exists among calculation data items, associated calculation can be performed in manual calculation for some quantities, thereby reducing workload in complementary calculation in automatic calculation and improving data consistency.

The invention discloses a query method and device, electronic equipment and a medium. In the method, after the to-be-processed content is acquired, the to-be-processed content is processed based on the preset filtering rule and the association strategy to obtain the request content, and the query result is acquired based on the request content. By applying the technical scheme of the invention, the query content input by the user can be optimized based on the filtering rule and the association strategy, and the corresponding query result is searched from the database based on the optimized query content. According to the invention, a large amount of useless information that a user is not interested in in in a query result in the prior art can be avoided, and operation resources of equipment are saved.

The invention provides a migration method which includes the steps: satisfying a migration condition, determining a to-be-migrated object, obtaining at least one associated object of the to-be-migrated object at least based on a determined association strategy which is at least related to an access relationship between the to-be-migrated object and the associated object, and migrating the to-be-migrated object and the at least one associated object from a current electronic device to a target electronic device at least based on the determined migration strategy, wherein the migration conditionis at least a condition for triggering the current electronic device to execute a migration operation. In addition, the invention further provides a migration device.

The invention discloses a device and a method used for dynamically controlling an application, and a mobile terminal. The device comprises an embedding unit, a notification unit, a decision unit and aprocessing unit, wherein the embedding unit is a controlled application embedding state monitor which needs to be dynamically controlled in user equipment; the notification unit analyzes a control instruction from control service to obtain a notification message which aims at the specific controlled application, and sends the notification message to the state monitor of the specific controlled application so as to accelerate the state monitor to generate a response message; the decision unit determines a processing strategy which carries out dynamic control by aiming at the specific controlled application on the basis of the response message received from the state monitor of the specific controlled application, and determines the association strategy of an association application relatedto the dynamic control on the basis of the processing strategy; and the processing unit carries out dynamic control on the specific controlled application according to the processing strategy under asituation that the association application is controlled on the basis of the association strategy.

The embodiment of the invention discloses a control method applied to a terminal. The terminal is internally provided with an ultrasonic sensor group; the method comprises the steps of obtaining a division instruction for the ultrasonic sensor group and dividing screen regions corresponding to at least one receiver into N regions according to the division instruction, wherein the ultrasonic sensor group comprises at least one emitter and at least one receiver, and N is a positive integer greater than or equal to 2; determining the first running emitter from the emitters; according to a preset association strategy, associating the first emitter with the N regions separately to obtain the emitting power of the N regions; utilizing the emitting power of the regions to recognize gesture instructions corresponding to the different screen regions, and achieving the control over the terminal. A mode of isolated-type terminal control is put forward and presents the diversity of terminal control modes. The embodiment of the invention also discloses the terminal and a storage medium.

The invention provides a method for associating a user to a network under small cell base station ultra-dense deployment for improving the energy efficiency of a hybrid energy heterogeneous network under ultra-dense deployment. The method comprises the following steps: firstly, a base station sets and transmits a base station weight value and a priority according to own power consumption situation; and secondly, the user perceives the use condition of green energy in an association process and preferentially accesses the base station with sufficient green energy. In the embodiment, a hybrid energy ultra-dense heterogeneous network scene in which power energy and green energy supply power at the same time is considered, modelling is performed on large-scale channel fading by using a dual-slope path loss model in 5G, and the user adaptively adjusts the association strategy according to the collection condition and use condition of the green energy of the base station. In the embodiment,the resource allocation and the transmitting power of the base station in the association process are optimized in combination with the matching theory, thereby effectively reducing the power consumption and improving the energy efficiency.

The invention relates to an M2M communication network joint clustering and power distribution method based on a non-orthogonal multiple access technology, and belongs to the technical field of wireless communication. The method comprises the following steps: S1, modeling channel gain between M2M equipment and a base station; s2, modeling channel gain between M2M devices; s3, determining a cluster head selection strategy; s4, pairing the cluster members, and determining all combination modes; s5, aiming at various cluster member combination modes, optimizing and determining a cluster head association strategy; s6, determining a cluster member power distribution strategy based on an intra-cluster energy efficiency optimization criterion; and S7, comparing all cluster member combinations, and determining a global optimal clustering and power distribution strategy. According to the invention, the M2M equipment association and resource allocation strategy can be optimized and designed, so that the equipment energy efficiency maximization is realized.

The invention provides a trusted system suitable for an industrial Internet of Things sensing computing layer, comprising a cache unit configured to be connected with an association unit and used for temporarily storing pre-detection information; the strategy library is configured to be connected with the association unit and used for storing association strategies; the association unit is configured to be connected with the acquisition end, the strategy library and the cache unit, and is used for receiving the post-detection information of the acquisition end, associating the pre-cached pre-detection information and the newly detected post-detection information into transmission information according to an association strategy of the strategy library, and transmitting the transmission information to the acquisition end; and caching the later detection information as the previous detection information of the next group of detection information. According to the invention, the prior detection information and the later detection information are associated into the transmission information to be sent, so that the prior received detection information can be verified, the cracking difficulty in the transmission process can be improved, the operation efficiency of the receiving end can be improved, and the method has the advantages of safety in use and high credibility.

The embodiment of the invention discloses a method and device for normalizing elements of network data, an apparatus and a storage medium. The method Includes the following steps: extracting the object relation from the original data set by using the objectification extraction strategy; Analyzing the object relationship according to a bridge association strategy and/or a weight setting strategy toobtain an identical relationship, wherein the identical relationship comprises a weight value of a relationship between two objects; Updating a weight value of a relationship between two objects in the same relationship according to a statistical weight calculation strategy; According to the same updated relationship, constructing the network and obtaining the normalized results of the elements.The method for normalizing elements of network data provided by the embodiment of the present invention updates the weight values of two object relationships through statistical weight calculation strategies after the same relationship is determined, and normalizes the elements by using the updated same relationship, thereby improving the accuracy of normalizing the elements.

The invention discloses a terminal tracing method based on traffic learning. The method comprises the following steps: 1, extracting terminal information used for tracing and dimension information used for forming machine learning in network traffic; 2, preprocessing the information extracted in the step 1; 3, establishing a feature classifier; 4, according to the extracted terminal information and the feature classifier obtained in the step 3, enabling all data to carry the terminal information through a related association strategy to complete terminal tracing. Compared with an existing active traceability method, the method is simpler and more suitable for big data development, and terminal information can be traced.

The invention provides an SLAM data association method based on a self-adaptive local and grouping association strategy. The method comprises the following specific steps: firstly, adaptively determining a local data association region according to the position of the mobile robot at the current moment, the measurement range of a sensor and the distribution condition of observation values at the current moment; secondly, adaptively determining the grouping number of the observation values according to the local association region and the number and distribution of the current observation values, and grouping the current observation values by using a Gaussian mixture clustering algorithm on the basis of the grouping number; thirdly, completing data association between the observed value ineach group and the feature points in the local area by using a joint compatible branch and bound algorithm, and screening out an optimal association solution of each group; and finally, integrating each group of optimal solutions to obtain a final data association result. Experimental results prove that the method can reduce the association calculation amount while ensuring high association accuracy.