49results about How to "Guaranteed hit rate" patented technology

The embodiment of the invention discloses a periphery searching method and device of points of interest, relating to the field of navigation. The periphery searching method and device of the points of interest can reduce the searching range and POI (point of interest) searching time in an area with higher POI density on the premise of ensuring a certain hit ratio, and enlarge the searching area and improve the hit ratio of POI searching in an area with low POI density. The periphery searching method of the points of interest comprises the steps of: obtaining the position of a reference point around the point of interest, and obtaining the distribution density of points of interest in the area unit to which the reference point belongs according to the position of the reference point; calculating the searching distance of the points of interest according to the distribution density of the points of interest; creating a searching area of the points of interest by taking the position of the reference point as the center and the searching distance as the searching radius; and searching the points of interest in the searching area according to keywords input by a user, and obtaining thepoint of interest matched with the key words. The periphery searching method and device of the points of interest, disclosed by the embodiment of the invention, are mainly used in the periphery searching process of the points of interest.

The invention provides a method and a device for updating cache resources. The method, which is applied to a cache server of a content delivery network CDN, comprises the following steps: receiving a resource request message from a user side, wherein the resource request message contains target resource information of a requested target resource; looking for the target resource information in a pre-created hit table, wherein the hit table is used to save the corresponding relationship between the resource information of locally downloaded cache resources and the aging time; if the target resource information is not found in the hit table and the target resource is locally cached, checking whether the target resource needs to be updated; and if the target resource needs to be updated, updating the target resource, and adding the target resource information and the aging time set for the target resource information to the hit table. According to the embodiments of the invention, by setting the aging time for the cache resource information and introducing the aging time to the hit table, the range of cache resource update check is narrowed, and the update efficiency is improved.

The invention discloses a CPU affinity-based virtual CPU dynamic binding method under a multi-core platform. Two modules, namely a guest operating system CPU affinity perception engine and a virtual CPU dynamic binder are used in the method. The guest operating system CPU affinity perception engine acquires CPU affinity information of a process to be executed from Guest OS, and provides the information for the virtual CPU dynamic binder by adopting quantified VAC (Value of Affinity Coefficient) CPU affinity coefficient value; and the virtual CPU dynamic binder module executes the corresponding strategy through the CPU affinity information provided by the guest operating system CPU affinity perception engine to realize dynamic binding of a virtual CPU.

The invention provides a method and a device for determining a search intention of a user. The method comprises the following steps: constructing a to-be-searched target knowledge graph and semantic concept feature vectors of entities according to an initial entity set; constructing a heterogeneous node graph and a corresponding first adjacent matrix; receiving a historical search click event, andestablishing a supervision model of the first adjacent matrix; receiving a current search keyword, and generating a candidate search intention extension entity set according to the current search keyword and the to-be-searched target knowledge graph; performing intention reasoning on the candidate search intention extension entity set by utilizing a supervision model to obtain a to-be-extended search keyword result set corresponding to the search intention; determining a target search keyword sequence corresponding to the search intention according to the to-be-extended search keyword resultset and the semantic concept feature vector. Therefore, when the search term input by the user has diversity, fuzziness and ambiguity, the knowledge graph and the semantic concept feature vector can be used for reasoning, expanding and filling the search term, so that the search is clearer.

The invention discloses an NAND flash memory-oriented buffer area management method, which comprises the following steps of: dividing a database buffer area into a cold clean linked list, a cold dirtylinked list and a mixed linked list, and numbering the cold clean linked list, the cold dirty linked list and the mixed linked list according to an overall access sequence; on the overall logic, thebuffer area is regarded as a single linked list organized according to an access sequence, a time window concept is provided, and data pages accessed in the latest period of time are covered; when buffer replacement occurs, checking whether a data page pointed by the tail of the linked list is in a time window or not according to the priority sequence of the cold clean linked list, the cold dirtylinked list and the mixed linked list; if so, searching a next linked list; and if not, directly replacing. According to the method, the flash memory write operation frequency is reduced while a relatively high hit rate is obtained, so that higher cache income is obtained, and the overall performance of the storage system is improved.

The invention discloses a Dalvik virtual machine just-in-time compilation (JIT) acceleration method based on a software Cache. The technical scheme includes that firstly, the software Cache is built in a Dalvik virtual machine, the software Cache adopts two-group connection structure and is added before a local code hash table; secondly, when a local code is found through a bytecode address, the local code is firstly found in the software Cache, when the local code is found out, the local code is directly accessed, when the local code is not found out, the local code is found in the local code hash table, when the local code is found out in the local code hash table, contents in lines of the software Cache are selected and written, when the local code is not found out, a compiling thread is waken up to compile the bytecode into the local code, and the local code hash table and the corresponding lines in the software Cache are simultaneously updated. Due to the fact that times of finding the local code hash table with a complex Hash algorithm in the Dalvik virtual machine are reduced, the Dalvik virtual machine JIT acceleration method based on the software Cache has the advantages of reducing system overhead of finding the local code hash table, thereby improving performance of a whole Android system.

The invention discloses a track monitoring system and method. The system includes: a first camera used for shooting an entrance of a monitoring range, a second camera for shooting an entrance of the monitoring range, a face detection module, and a server; the face detection module extracts first face information from a video shot by the first camera, and the tracking detection module extracts second face information and pedestrian position information from a video shot by the second camera; and the server performs face matching in a first face database according to the first face information,and writes the matched face data into a second face database, so that the matching number is reduced when the second face information extracted by the tracking detection module is matched. According to the invention, the calculation amount of face matching during pedestrian tracking can be greatly reduced. The method can be widely applied to the technical field of pedestrian tracking.

The invention provides a configurable caching system and method. The system comprises a front-end scheduler and a front-end cache server. A related location rule and an embedded script Lua language are provided in the front-end scheduler. According to the system and the method provided by the invention, by using url rewriting technology and a self-defined header field value in the front-end scheduler, factors that may interfere with caching are eliminated, thereby ensuring a hit rate of caching to a maximum extent; it is ensured that different URLs are cached according to same logic; meanwhile, it is ensured that login users are provided with separate caching logic for a same URL.

The invention discloses a popular content cache update method, a communication device and a network applicable to a mobile cellular network. Content cached in a node is partially updated at the right time according to a node location relation by LRU and LFU cache update methods. According to the cache update method provided by the invention, a data size for updating the cache is greatly reduced. The update method is adjusted aiming at characteristics of the user node of the mobile cellular network based on the existing LRU and LFU cache content update method. Particularly, in view of limited information acquisition capacity and computing capacity of the node, the method can update the cache in real time by only using information such as the node location variation, received content request and time owned by the node, in combination with a content value model of the LRU and LFU, so that balancing between the computing overhead and a cache hit ratio is achieved to a certain degree.

The embodiment of the invention provides a content central multi-hop cellular network routing method and device. The method comprises the steps that a first node broadcasts content identifiers to nodes in the i-hop range of the first node, and therefore a second node in the i-hop range of the first node determines first routing information of a forwarded interest package, and a first directory node broadcasts content identifiers and a first directory node identifier to a second directory node and nodes within the coverage range of the first directory node. According to the method, forwarding nodes which are located in the same area with the position of target content are determined through the position of the target content, a forwarding node forwards the interest package in the direction of the position of the target content, and compared with the mode that the nodes inform all nodes in the network of cache content, consumption of wireless spectrum resources in the network is reduced, consumption of electricity in the nodes is reduced, the network resource utilization rate is improved, and the survival time of the network nodes is prolonged.

The invention discloses an unstructured grid many-core coarseness parallel computing method. According to the method, on the basis of region decomposition of a first-level non-structural grid, second-level thread level region decomposition is added, each slave core solves a respective independent calculation region, the data hit rate of a slave core calculation task is ensured, and coarse-grainedparallelism of an MPI process level and coarse-grained parallelism of a slave core thread level are achieved. According to the method, the problem of adaptability of a common non-structural grid applied to a polymorphic heterogeneous processor can be solved, secondary load balance and coarse-grained many-core parallelization of a calculation core are automatically completed according to the scaleof non-structural grid data, and the calculation efficiency and the parallelization efficiency of numerical simulation of the non-structural grid on a heterogeneous system structure are improved.

The invention discloses a flash memory data buffer area replacement method based on a page access interval. In the method, all data pages cached in the data buffer area are divided into hot dirty data pages, cold dirty data pages, hot clean data pages and cold clean data pages according to access interval characteristics and reading/writing characteristics of the data pages in the buffer area, and all the data pages are managed by using a primary data buffer area and a secondary data buffer area. When the cold data page access interval is less than the last hot data page in an LRU queue of the buffer area, conversion between cold and hot data pages is performed, and meanwhile, the cold data page in the LRU queue of the primary buffer area is moved into the secondary data buffer area. When the data buffer area page needs to be replaced, the cold clean data pages in the secondary buffer area are preferentially shifted out according to an LRU rule, and the influence on performance and service life of a flash memory due to too many times of write-back operations of the dirty data pages of the data buffer area is reduced under the premise of ensuring a hit rate.

The invention belongs to the field of video coding, and discloses an HEVC intra-frame angle mode selection method, device and equipment and a readable storage medium, and the HEVC intra-frame angle mode selection method comprises the steps: obtaining a learning video frame image, and dividing the learning video frame image into a plurality of PU images; constructing a deep convolutional neural network; and learning and predicting the optimal angle mode of the current PU image from the PU image to obtain a deep convolutional neural network prediction model, embedding the deep convolutional neural network prediction model into an original HEVC coding process, and directly obtaining n mode candidate lists of an optimal angle mode through the deep convolutional neural network prediction modelto replace 35 angle modes during video frame image coding. According to the method, the deep convolutional neural network is introduced into HEVC video coding, the problem of PU angle mode selection suitable for being solved is learned from an original input video frame image, the video coding quality can be ensured, and the problem of long time in the video coding process is well solved.

The invention provides a method for saving a PostgresSQL shared memory. The method comprises the steps of tracking and counting the times of updating and searching a table; if the times of updating and searching meet the set condition, triggering data reconstruction; reconstructing the row of the selected PK to a continuous data block; on the basis of 100000 records, 1000 8K data blocks are supposed to be used, and 1000 data are supposed to be accessed often, the method is that the 1000 records are stored in the continuous data block, and 10 8K data blocks are only needed, so that the 80K shared memory is needed, and as a result, the accuracy rate of the data access in the share area of the memory can be ensured, and the shared memory can be avoided large waste.

The invention discloses a machine gun shooting door design method and system based on shooting scattering, and the method comprises the steps: carrying out the frame selection of a target through a fire control observation and aiming system after the target is found, and enabling the target to serve as a target tracking frame; after a tracking target is determined, pulling the target to an observing and aiming center, and establishing a rectangular coordinate system with the target center as an original point; adjusting the target tracking frame to enable the edge of the target image to coincide with the edge of the target tracking frame so as to correct the size of the target tracking frame; performing laser ranging on the target, calculating the size and position of the shooting door according to the distance information, and displaying the shooting door in a display control terminal picture through imaging conversion; determining whether the shooting door is located in the target surface or not, if the shooting door is not located in the target surface, adjusting the pitch angle and the azimuth angle of the machine gun to enable the shooting door to enter the target surface, andmeanwhile imaging the target in the pitch direction in the center of a view field; and monitoring the state of a firing button, judging whether a firing condition is met or not when a shooter pressesdown the firing button, and controlling the machine gun to shoot based on the judgment result.

The invention discloses a cache replacement method and device based on energy consumption characteristics of an MLC STR-RAM, and belongs to the technical field of computer storage, the MLC STR-RAM isused as a system main memory, the method comprises the following steps: when cache replacement occurs in a final-stage cache, respectively calculating N indexes of each cache block in the final-stagecache according to N preset cache replacement algorithms, respectively recording the indexes as a first index, a second index,..., and an Nth index of the corresponding cache block, respectively estimating write-back energy consumption generated by refreshing the data in each cache block back to the main memory, and taking the estimated write-back energy consumption as an (N + 1) th index of the corresponding cache block; comprehensively judging the N + 1 indexes of each cache block to select the eliminated cache block as a target cache block; wherein N is an integer greater than or equal to 1; in the N cache replacement algorithms, each cache replacement algorithm selects an eliminated cache block according to a single index, and the indexes of different cache replacement algorithms are different. According to the method, the write-back energy consumption of the MLC STT-RAM main memory can be effectively reduced.

The invention relates to a page resource acquisition method and device, electronic equipment and a storage medium, and relates to the technical field of information. The method comprises the steps that: a uniform resource locator URL of a client page is obtained, a packet identifier of an offline packet where a target resource is located is determined by analyzing the URL, and the target resource is the resource needed when the client page is loaded; according to the URL, a preset storage path of the target resource in the offline packet indicated by the packet identifier is generated; when the target resource is not obtained according to the preset storage path, a configuration file corresponding to the offline packet is obtained, the configuration file is used for recording storage path information corresponding to the preset storage path, and the storage path information is used for indicating an actual storage path of the target resource of the corresponding preset storage path in the offline packet; storage path information corresponding to the preset storage path is obtained from the configuration file; and the target resource is obtained from the offline packet according to the actual storage path indicated by the storage path information. According to the invention, the page loading speed is improved.

The application provides a cache updating method based on node displacement and LFU records and a network. A popular content cache updating method suitable for a mobile network, a communication device and the network are used to partially update the popular contents cached in a node according to the displacement of the node in the network in combination with an LFU cache updating method. By adopting the cache updating method provided by the application, the data traffic consumed by updating the cache can be reduced; aiming at the obvious difference of the popular contents in different regions of the mobile network, the node displacement is taken into account when performing cache updates; particularly, in consideration of the own limited resources of the node, by adopting the method provided by the application, a content value model can be established and the cache can be real-timely updated by only using the location information that can be easily acquired by the node and the content request, time and other information received by the node in combination with the LFU record, and the validity of the stored data in the node cache can be improved without generating an additional calculation load.