142 results about "Incremental computing" patented technology

A treatment planning method and system for optimizing a treatment plan used to irradiate a treatment volume including a target volume, such as a tumor, is disclosed. According to the method, two dose calculation algorithms are used to develop the optimized treatment plan. A first dose calculation algorithm is used to obtain substantially complete dose calculations and a second, incremental, dose calculation algorithm is used to make more limited calculations. The incremental calculations may be performed, for example, with less precision, less accuracy or less scope (e.g., focused on a specific subvolume within the treatment volume) in order to reduce the time required to achieve an optimized plan. Each of the dose calculation algorithms may be iterated a plurality of times, and different cutoff criteria can be used to limit the number of iterations in a given pass. A treatment planning system of the invention uses software for implementing the complete and incremental dose calculation algorithms. The method and system are especially useful for IMRT and arc therapy where treatment plan optimization is particularly challenging.

The invention relates to the field of a social network and is applied to a network system with social elements. The invention discloses a method and a system for calculating user influence in a social network, and in the method, the behaviors of users are analyzed so that a model of the behavior of influence spread among users is established and the mold is used for identifying the most influential people in the social network. The system records user access logs in real time. The method comprises the following steps that: (1) an influence behavior extracting module filters the interactive behaviors in the logs; (2) an influence spread calculating module determines the influence of the current behavior spread according to the type and the actor of the behavior and the emotion tendentiousness analysis of the relative content; (3) an influence updating module mergers the data generated through the increment; (4) a user ranking module incrementally calculates the most influential people; and (5) a service module appoints a target user range and sequences the target users according to the influence. The method and the system can dynamically reflect the change condition of the user influence according to the interactive behaviors of the users in the social network.

A method and a system implementing web-based incremental computing. A webpage updates when a data value relevant to displayed information is altered and updates only those affected elements. A watch system monitors relevant data values and in conjunction with a dependency tree, may execute functions to alter only altered elements of a webpage. A framework to automatically track data dependencies and propagate data changes is disclosed.

The invention discloses a fast statistical query method for mass medical data. The method includes: firstly, collecting, filtering and processing medical data from different data sources, and storing processed medical data into an HDFS (Hadoop distributed file system) or HBase; using the HDFS or HBase to create and initialize statistics trees for one or certain statistics-based medical data; subjecting the newly added medical data to single or multiple incremental computations, and updating the statistics tree by results of the incremental computations; allowing a server to respond to a statistical query request from a user according to the one or a plurality of created statistics trees, to analyze the statistical query request into sub-queries based on one or the plurality of statistics trees, and to combine into a plurality of sub-query results fed back to the user. The position of statistical data can be located quickly by the method, access to mass data is avoided, system load is lightened, and query efficiency and access concurrency are increased.

A method is provided to update information in a computer system comprising providing a reactive value hierarchy structure in a non-transitory computer readable storage device; in response to a change in a reactive value corresponding to a node in the hierarchy, marking nodes within the hierarchy that share as dirty a path with a node corresponding to a changed reactive value; traversing down through the hierarchy to a level where a reactive value is located that is either dirty or clean; if the reactive value is dirty, traversing back up to a dependent node in the hierarchy and compute a new reactive value and repeating the traversing down step; if the reactive value is clean, traversing back up to a dependent node and repeating the traversing down step; and producing at least one side-effect in the computer system in response to a compute of a new reactive value.

Systems and methods for instance counting and for the identification of a temporal pattern in an event sequence. The system addresses the use of “earliest-first” and “no-reuse” policies as criteria for the correctness of counting. The system also achieves higher performance than conventional methods by utilizing incremental computation.

Systems and methods for data processing using incremental algorithms. Embodiments of the invention decompose complex or monolithic data processing problems into one or more incremental computations called flows. These flows may be distributed across a networked cluster of commodity computers, facilitating the easy scaling of the system and robust recovery functionality. Once a request is submitted to the system, its solution may be maintained from that point in time forward, such that whenever changes are made to a problem's data the solution is efficiently recomputed.

Disclosed are methods and for incrementally updating the accuracy provided by documents in training set of used for automatic categorization. A k-nearest neighbor database includes the documents in the training set, categories, category assignments of the documents and category scores for the documents. A list made up of the nearest neighbors of the documents and corresponding similarity scores contains is maintained by the method. On adding or deleting documents or category assignments, the documents influenced by the changed documents or category assignments are identified. The category scores of the identified documents are updated to be consistent for the updated training set and a new precision and recall curves are computed for the categories including updated category scores. The precision and recall curves may be used to determine an optimal number of documents to maximize the return of relevant documents while minimizing the total number of documents.

In the data storage system the storage area network performs XOR operations on incoming data for parity generation without buffering data through a centralized RAID engine or processor. The hardware for calculating the XOR data is distributed to incrementally calculate data parity in parallel across each data channel and may be implemented as a set of FPGAs with low bandwidths to efficiently scale as the amount of storage memory increases. A host adaptively appoints data storage controllers in the storage area network to perform XOR parity operations on data passing therethrough. The system provides data migration and parity generation in a simple and effective matter and attains a reduction in cost and power consumption.

A treatment planning method and system for optimizing a treatment plan used to irradiate a treatment volume including a target volume, such as a tumor, is disclosed. According to the method, two dose calculation algorithms are used to develop the optimized treatment plan. A first dose calculation algorithm is used to obtain substantially complete dose calculations and a second, incremental, dose calculation algorithm is used to make more limited calculations. The incremental calculations may be performed, for example, with less precision, less accuracy or less scope (e.g., focused on a specific subvolume within the treatment volume) in order to reduce the time required to achieve an optimized plan. Each of the dose calculation algorithms may be iterated a plurality of times, and different cutoff criteria can be used to limit the number of iterations in a given pass. A treatment planning system of the invention uses software for implementing the complete and incremental dose calculation algorithms. The method and system are especially useful for IMRT and arc therapy where treatment plan optimization is particularly challenging.

The invention discloses a virtual machine position selection system facing load balance in a cloud computation environment, which comprises a front end module and a rear end module, wherein the front end module comprises a comprehensive load computation module and a host selection module, and the rear end module comprises a plurality of distributed node agent modules in an integrated manner. Each node agent module comprises a node load collection module, a node load predication module and a node load incremental computation module. Cloud computation serving as a novel computation and service method is put forward and is increasingly widely paid attention to, popularized and applied, and the virtual machine position selection system serves as a vital module in resource allocation of cloud computation, so that the virtual machine position selection system has the advantages that the virtual machine network throughput rate in the cloud computation environment is increased, the resource preemption probability of virtual machines is decreased, the service quality of a virtual host is improved, and the load balance of a distributed system in the cloud computation environment is enhanced. Hence, the virtual machine position selection system is wide in application range and capable of generating remarkable economic benefit.

The present invention relates to communication field, wherein an embodiment of the invention discloses an ECMP path soft recovery method in a MPLS-TP packet transfer network. The embodiment of the invention comprises the following steps: realizing functions of congestion discovering, congestion positioning, etc. by the expanded OAM information; transmitting TDM grouping dynamic clock by MPLS-TP for adjusting operation; path cracking of the bypass heavy routing; when congestion occurs, firstly adjusting TDM service data grouping length through a source end for improving the delay characteristic; adopting a local service for shunting to the service with overhigh speed for steering clear of the congestion routing; if local link resource is insufficient, informing the source end for adjusting the bandwidth allocation, allocating higher bandwidth resource on the other N-1 paths; if no enough resources can be allocated or the allocated resources are not adjusted under the principle of simplifying the operation, informing the routing module by the source end, and completing the calculation of the other routing through the mode of incremental computation according to the bandwidth request amount. The method of the invention reduces the time for recovering network congestion. The network flexibility can be enhanced through the mode of self-adapting grouping variable-length transmission, and the using efficiency of the protected resource is increased. The quick performance recovery of the service is realized through the mode of bypass heavy routing cracking. The quick combining operation at the branching point is completed with a special numbering and a signaling process.

The invention discloses a user interest model incremental update method of a personalized recommendation system. According to the basic principle of the method, the method includes storing and generating an intermediate result of calculation of a current user interest model, and performing incremental calculating on the basis of the intermediate result when the user interest model is updated. On the premise that interest information is protected from losing during updating process, the requirements that the user interest model can be updated rapidly and continuously on condition of large data amount can be met, performances of the recommendation system can be improved, and higher-quality service is provided for users.

The invention discloses a construction method and system of a human society person portfolio database on the basis of big data. The construction system comprises a human society business database, a business data synchronization module, an archive information initialization module and an incremental message storage module, wherein the human society business database provides an original data source for the initialization of the human society person portfolio database, and provides message flow LCRs (Logical Change Records) for data increment; the business data synchronization module is used for finishing data initialization from business data to human society process data, so as to realize increment synchronization from the human society business database to a human society business process database; the archive information initialization module is used for defining, scheduling and executing an archive information initialization task, finishing the extraction, the conversion, the cleaning and the integration of human society business process data and storing results to the human society person portfolio database; the incremental message storage module receives an incremental message of an incremental synchronization module, and provides an incremental message flow for triggering flow calculation for an archive information incremental calculation module. According to the construction system, the human society business data is taken as the data source, and a big-data processing technology and a distributed storage technology are used for constructing an accurate and real-time person portfolio.

The invention relates to an arithmetic decoder based on context self-adaptive binary, wherein said decoder comprises an arithmetic decoder controller, a data interface, a mark subscript, a context subscript increment counter, a context manager, a decode engine and a mark detector; the controller is used to command external modules, the communication between nearby information and grammar element value, to simplify the interfaces of whole decoder; the data interface is used to communicate code flow with outer space, to reduce the communication times between decoder and external code flow; the mark subscript and the context subscript increment counter can quickly calculate out the mark subscript and context subscript increment; and the decode engine can realize the decode at several modes, to save the hardware consumption and accelerate the decode.

The invention discloses an NPC three-level inverter optimal sequence model prediction control device and method. The device comprises an NPC three-phase three-level inverter, a digital processing control module and a driving circuit, wherein the digital processing control module comprises a sampling unit, a sector pre-selection unit, a vector action time calculation unit, a prediction calculationunit and a modulation unit. The method comprises the following steps of: calculating a reference voltage vector and carrying out sector selection on a pre-generated three-vector sequence, calculatingincrements of the current and the neutral-point voltage according to the sampling information, calculating vector action time and correcting the vector action time, predicting, calculating and selecting an optimal vector sequence and action time, and generating a switching signal by the modulation unit to drive the NPC three-phase three-level inverter to work. The method is low in hardware cost, accurate in control and wide in application range, fixed switching frequency grid-connected control over the NPC three-phase three-level inverter can be achieved, the total harmonic distortion rate ofthe network access current is reduced, and meanwhile fluctuation of the midpoint voltage of the direct-current side is effectively restrained.

The utility model provides a calculation method of block size of cylindrical panoramic video coding and a transmission method, which is characterized in that: a code rate of the panoramic video coding is got without direct blocking pre-coding to the cylindrical panoramic video; blocking pre-coding to the video is operated according to an optional block size, and the code rate in the block size is got; the increment of the two code rate is computed; the impact factor of the video is computed based on the increment; and the optimal coding block size is computed based on the impact factor. The utility model also provides a transmission method of the cylindrical panoramic video block coding based on the block size computed by the calculation method. The utility model has the advantages of computing the optimal coding block size by twice pre-coding, and having lowest transmission data size in the same visual quality of users by using the optimal block size to code the cylindrical panoramic video.

The invention relates to a real-time large data processing system and method based on distributed internal memory calculation. The real-time large data processing method comprises the following steps: a resource pool is formed by respective internal memory resources of multiple distributed internal memory nodes; when a user originates a task, a client transfers the task and incremental data related to the task within a fixed time interval to a resource scheduler of the resource pool to be applied for execution; the resource scheduler can be used for carrying out deployment in the resource pool according to the internal memory resources needed by the task, corresponding internal memory nodes in the resource pool are appointed for carrying out calculation on the incremental data related to the task, obtained local results of respective calculation are stored in the resource pool; and local results, obtained by virtue of calculation of the appointed internal memory nodes, are calculated in a combination mode to obtain a global calculation result, and the global calculation result is sent to the user. The real-time large data processing system and method is capable of achieving centralized use of distributed internal memories, restoring data in the process of the internal memory calculation and achieving real-time data incremental calculation, is high in data access speed and can be used for providing a real-time analysis result for the user.

The invention relates to an entity finite element calculation method for a time-varying effect of a concrete structure, which comprises the following steps of: 1) discretizing the stress process of concrete according to a stepwise method, and deducing creep strain increment in an nth time period; 2) fitting the creep coefficient expression into an exponential series form by using Matlab software;3) substituting the creep coefficient fitting expression into a creep strain increment calculation formula, and establishing a concrete uniaxial creep strain increment recurrence calculation formula;4) considering the creep Poisson effect, and establishing a concrete time-varying constitutive equation in the constitutive structure of the entity unit material; and 5) developing the time-varying constitutive structure of the concrete in the USERMAT subprogram of the ANSYS software by using the Fortran language. Compared with the prior art, time-varying effect value prediction of the concrete structure in a complex stress process under a spatial stress state can be realized, and the joint with a creep and shrinkage model of a common design specification is good.

The invention discloses an incremental correction method for a tidal atlas of an electric network. The incremental correction method comprises the following steps of: firstly performing structure hierarchical reorganization on an electric network model, defining a voltage class range of the electric network model, then performing incremental calculation on the model, further calculating a graphic increment according to calculation result of the incremental calculation of the model, and finally correcting the graphic increment of the electric network model according to the calculation result of the graphic increment. According to the incremental correction method disclosed by the invention, incremental updating of the tidal atlas of the electric network within a full voltage class range or a single voltage class range of the electric network in the jurisdiction scope of a certain control center of an electric power system can be realized, namely on the basis of realizing full graphic updating, the parts which are not changed during every correction can be prevented from being deleted and re-generated; and simultaneously, according to the incremental correction method disclosed by the invention, the graphic increment is in one-to-one correspondence with a model increment based on an IEC (International Electrotechnical Commission) 1970 / CIM (common information model) model standard, so that the changes in tidal flow of the electric network is very easy to identify, and the working difficulty of dispatching staff and operation staff of the electric network can be greatly reduced.

The invention provides a tunneling method and system for automatic surveying, mapping and positioning. The tunneling method comprise the steps that a well lane model given by a design system is dispersed into a bottom plate middle axis multi-section line and a well lane section contour polygon; a machine body coordinate system of a heading machine is set with the right side direction of the heading machine as the X axis, the head direction of the heading machine as the Y axis, and perpendicular to the XOY plane and the overall upward direction as the Z axis; the relative position coordinate increment from the laser target ball position to the turning center of the heading machine is calibrated according to the coordinate axis increment of the machine body coordinate system; the position coordinates of a laser target ball of the heading machine are positioned by using a positioning robot and a north pointing device; according to the position coordinate of the laser target ball and the relative position coordinate increment of the laser target ball position in the machine body coordinate system, the motion base point coordinates of a cantilever of the heading machine is calculated; the heading machine is controlled to cut the radial section according to the multi-section line of the bottom middle axis, the well lane section contour polygon, and the motion base point coordinates of the cantilever of the heading machine. According to the tunneling method and system, the heading machine can be accurately positioned, and the operation is unmanned.

The present invention discloses a dynamic social network community structure evolution method based on incremental clustering to solve the problems of community structure detection and communication evolution tracking in a large scale network. The method comprises a step of extracting the core node of a whole network to form a core sub graph, a step of running a hierarchical clustering algorithm on the core sub graph at a time t=0 to obtain the initial structure of a core community, and using an extended algorithm on the above basis to obtain the community structure of the whole network, and a step of using an incremental clustering algorithm to obtain the core community structure of the whole network at present time according to the dynamic evolution condition of an adjacent time network at a time t which is larger than 0 and extending the core community structure to obtain a whole community structure. Through introducing the core sub graph, the incremental calculation in the whole network is avoided, the processing speed is accelerated, and thus the method is suitable for the community discovery in the large scale network. In addition, through introducing the concept of a community structure shift, the large error of the community structure after long time evolution is avoided, and the accuracy of community evolution tracking is improved.

The invention discloses a graph structure matching-based social network analysis method, and relates to the fields of a graph structure matching technique, an incremental calculation technique and the like. The graph structure matching-based social network analysis method mainly comprises the steps as follows: (a) a graphic interface helps a user manage graph data, and constructs a mode pattern and a visual query result; and (b) a query engine invokes a distributed algorithm to carry out structure matching calculation; meanwhile, the matching results are evaluated; and top-K results are selected and are visualized on the graphic interface; and (c) incremental calculation is carried out on the existing matching results by an incremental algorithm, so as to ensure the accuracy and the integrity of the results. Aiming at ''large data'' in a social network, efficient and convenient data analysis and maintenance are achieved through distributed, visual and incremental calculation techniques, so that a key technical support is provided for hotspot applications such as expert recommendation, social circle identification and social position analysis.

A method of selecting a motion vector for use in a motion estimation system in which the motion vector defines movement of a block of pixels between a search window and a reference frame. The method includes the steps of scanning a search window in a fixed order for a suitable match with an area of the reference frame such that each search position within a current row or column of the window is one pixel away from a previous search position, defining a preferred point in the search window, calculating a primary norm function with reference to the preferred point for a given position in each row or column, calculating a norm function for each search position based on a linear increment of the primary norm function for the respective row or column, calculating an Absolute Error (AE) for each search position, based on a difference between a given property of the search position and a position in the reference window, and selecting a motion vector, the displacement of the motion vector defined in terms of the search position having the smallest AE and the smallest norm function. An apparatus for performing the method is also disclosed.

The invention provides a method and a system for realizing checksum updating of a 1588UDP (User Datagram Protocol) packet in multiple MAC (Media Access Control). The method comprises the following steps that: a message outlet processing module reads message data, edits a message header and modifies the message data, substitutes a modified 64-bit time stamp or a time modification domain value with 0, and obtains the value of the original 16-bit UDP checksum by calculation; the current 64-bit time stamp is obtained from a PTP (Precision Time Protocol) engine through a message time identification point, and the value of new 16-bit UDP checksum is obtained based on the value of the original 16-bit UDP checksum by incremental computation; and the MAC substitutes the value of the new 16-bit checksum and the 64-bit time stamp for the previous data while transmitting the message to the position on which the data are modified. In the method, the UDP checksum computing mode before the PTP message enters into an MAC end is modified, and the 64-bit original time stamp or the time modification domain value is not needed to be stored in multiple MAC, so that use of a register or a memory is reduced greatly, and design cost of chips is also reduced.

The invention discloses a Hadoop platform time series data incremental computation method and system. The method includes the steps that when a time series data incremental computation task is started, the historical computational state of time series data is obtained from a cache server; incremental computation is carried out by means of a segmented time series data incremental computation method containing a SubCp sub-operation and a ReduceCp sub-operation according to the historical computational state, wherein the SubCp sub-operation is used for self-definition of segmented time series data and storage of intermediate results, the ReduceCp sub-operation is carried out in an operation merging stage and used for merging operation on computed results of the segmented time series data according to the self-defined operation, and the computational state of the SubCp sub-operation and the computational state of the ReduceCp sub-operation are maintained through the cache server. By the adoption of the method and system, plenty of unnecessary repetitive computation can be saved through incremental computation, and therefore data processing efficiency is improved.

The invention provides an object pose estimation method, device and electronic equipment, and belongs to the field of image and video processing. The object pose estimation method is to utilize multiple cameras to perform pose estimation on the same estimated pose object, the method includes determining the camera pose difference between the two cameras in a system which consists of the multiple cameras and the estimated pose object; utilizing views taken by the multiple cameras to perform the pose estimation to obtain multiple initial pose estimation results; selecting at least two initial pose estimation results capable of fusing from the multiple initial pose estimation results according to the camera pose difference; utilizing at least two initial pose estimation results to build optimized objective functions and constraint conditions thereof, and performing solving to obtain the correction increments of each initial pose estimation result; and utilizing the correction increments to obtain a target pose estimation result through calculation. The precision and confidence degree of the object pose estimation can be enhanced according to the technical scheme.

The invention discloses a log real-time monitoring and early warning method and its device. The log real-time monitoring and early warning method comprises the following steps: collecting server logs; aggregating and summarizing the collected server logs according to predetermined time and business; Incremental calculations are performed according to the aggregated server logs; and the pre-set threshold is used to determine whether to issue an early warning based on the results of incremental calculations. The present invention realizes the monitoring and early warning with high timeliness, high accuracy and low calculation cost through the continuous real-time incremental calculation mode.

The invention discloses a modeling system based on a logical relationship, which comprises an operation task integrating module, a task interpreter, a graph layout correcting module, a graph wiring correcting module and a database model increment correcting module, wherein the database model increment correcting module comprises a graph increment calculating unit and a model increment calculatingunit. Furthermore, the invention discloses a modeling method using the modeling system. The system, through description of the logical relationship and based on support of the automatic wiring technology, realizes the pattern-model integrated generation of a new grid model of the power system which is based on the logical relationship description; the grid model is completely defined and modifiedon one net, which is different from the original way that the power system model is built on countless net patterns, thereby achieving and helping the grid operation manager to accurately and rapidlyestablish and modify the grid model of the full system; and the modeling system is clear and direct, reliable and accurate, and is easy.

The invention relates to a netuser behavior data real-time processing method based on distributed computation. The netuser behavior data real-time processing method includes a starting process and a real-time computing process which are sequentially operated. The starting process includes extracting data from a netuser behavior database, subjecting the data to distributed batching computation to acquire a statistical result initial value, and storing the statistical result initial value into a statistical result cache region and a result database. The real-time computing process includes that a data source end continues to receive event data streams generated from user visit interbehavior data and stores the same into the netuser behavior database, all current event data streams on the data source end are distributed in multiple nodes in distributed manner, incremental computation and agreement stipulation are performed on the event data streams according to the statistical result initial value, and the final result is saved in the statistical result cache region. Compared with the prior art, the netuser behavior data real-time processing method has the advantages such that distributed multi-node parallel computing is supported, incremental computational algorithm of the event stream is designed based on the actual needs, coupling degree with other functions on the network platform is low, and real-time computing is effectively guaranteed.