101results about How to "Improve distribution speed" patented technology

An exhaust nozzle assembly for a gas turbine engine, the assembly comprises a main axis, an inner nozzle, an outer nozzle, a translatable centre-body and an actuator for translating the centre-body between a forward position and a rearward position. The centre-body is disposed radially inwardly of the inner nozzle thereby partly defining an inner duct for a core engine gas flow, an outer duct for a bypass gas flow is defined by the inner nozzle and the outer nozzle. The outer nozzle extends downstream of the inner nozzle and with the centre-body defines a final mixing duct having a final exhaust exit area. When the centre-body is in the forward position the final exhaust exit area is at a maximum area, and when the centre-body is in a rearward position the final exhaust exit area is at a minimum area.

A live telecast system of reciprocal-connected stream media comprises collection server, index server, a numbers of client ends and a numbers of super-seed nodes. It is featured as making said super-seed nodes have more strong download-service ability than that of client ends and forming high speed distribution network of stream media data to speed up transmission of stream media data by utilizing super-seed nodes distributed on network.

An improved nozzle vane for a gas turbine engine, comprising a vane wall having inner and outer wall surfaces, the wall surfaces being spaced from one another to define a plurality of fluid passageways for a cooling medium; discreet cavities formed by interior wall members disposed between the inner and outer wall surfaces and within the fluid passageway for the cooling medium; a plurality of impingement cooling sleeves disposed in the discreet cavities defined by the inner and outer wall surfaces and by interior wall members; and a plurality of non-round, e.g., serrated, openings in each of the impingement cooling sleeves, with the openings being sufficient in size and number to accommodate the flow of a cooling media.

The invention discloses a prefetch method for fixed-point data in Peer-to-Peer on demand system, comprising the steps of: (1) the time interval of a pointing device is decided by analyzing user's behavior and experience improvement; a medial head file and a data file are resolved by a file analyzer, metadata information is recorded and a corresponding index file is created; (2) according to media index file information, in condition of acquiring current playing data, prefetching point data is acquired from partner node or source server on 'Minimum Redundancy First' principle with residual bandwidth,; and (3) according to the drag position, integrating the media acquiring data in current channel cache window, a client automatically adjusts the play position to the user-appointed time point, or nearest prefetching point. The invention is based on observation of user's behavior and makes full use of the characteristic of Peer-to-Peer, realizes effective share of fixed-point data, lowers bandwidth consumption of a stream media server, improves user's watch experience, and is suitable for large-scale application.

In a fan filter unit for filtering and continuously supplying clean air into a clean room, the fan filter unit includes a chemical filter for removing chemical contaminants contained in the air, a fan for sucking and blowing out the air, and a particle filter for removing particles such as dusts and moisture contained in the air. A filter medium of the chemical filter has a circular tube shape and is disposed between an inner circular tube and an outer circular tube. A cover is detachably combined with top portions of the inner and the outer circular tubes, and prevents air flow through the top portions of the inner and outer circular tubes. The air is sucked only through the outer circular tube, the chemical filter and the inner circular tube by a suction force of the fan. Accordingly, velocity distribution of the air passing through the chemical filter may be improved and lifetime of the chemical filter may be increased.

The invention discloses a logistics service system, a path analysis method and a logistics distribution method. A data interface module can access a database to obtain logistics data, and calculate the logistics data to form display data; the database can obtain logistics information on the World Wide Web by accessing the World Wide Web, then converts the logistics information into logistics dataand stores the logistics data; a transportation capacity algorithm module can access the data interface module to obtain the display data, and the transportation capacity algorithm module calculates and analyzes the display data to form a path scheme; and the data interface module can further access the transportation capacity algorithm module to obtain a path scheme having at least one path plan.The logistics service system, the path analysis method and the logistics distribution method have the advantages and beneficial effects of having excellent expansibility, excellent adaptability and wide application range, ensuring data independence and security, improving the reliability of a carriage path, providing great convenience to customers, further providing a good transportation capacityinformation display platform for carriers, and greatly enhancing the fairness and transparency of the logistics industry.

The invention discloses a file releasing method and system based on point-to-point. The file releasing method comprises the following steps: obtaining pre-generated fragmented file package linking information; obtaining a pre-stored fragmented file package source data according to the obtained fragmented file package linking information and downloading the pre-stored fragmented file package source data; after downloading a fragmented file package, registering a client side downloading the fragmented file package to a fragmented file package P2P(peer-to-peer) torrent; and updating the fragmented file package interlinking information according to the registered fragmented file P2P torrent information. By use of the method and the system, the peek bandwidth of a server and the user cost can be reduced during the initial releasing of the file.

The invention discloses a printing data processing method and device and a printer. The method comprises the steps of splitting to-be-printed image data into image data according to a software processing channel; splitting each part of the image data into a plurality of parts of intermediate data according to a distribution rule of a logic processing channel; splitting each part of the intermediate data into a plurality of parts of printing data according to the arrangement rule of the nozzle rows on the corresponding nozzles; and sending each piece of printing data to the corresponding nozzlerow. According to the invention, the data processing and distribution speed and the printing efficiency are improved.

In a fan filter unit for filtering and continuously supplying clean air into a clean room, the fan filter unit includes a chemical filter for removing chemical contaminants contained in the air, a fan for sucking and blowing out the air, and a particle filter for removing particles such as dusts and moisture contained in the air. A filter medium of the chemical filter has a circular tube shape and is disposed between an inner circular tube and an outer circular tube. A cover is detachably combined with top portions of the inner and the outer circular tubes, and prevents air flow through the top portions of the inner and outer circular tubes. The air is sucked only through the outer circular tube, the chemical filter and the inner circular tube by a suction force of the fan. Accordingly, velocity distribution of the air passing through the chemical filter may be improved and lifetime of the chemical filter may be increased.

The invention provides a content distributing method and system. The method comprises: determining distance measurement values between all nodes; generating a collection tree of a plurality of shortest paths from edge nodes to a source node according to the distance measurement values between all the nodes and generating a router table; issuing the router table to all the edge nodes to generate a topological structure; and receiving a content distributing request from a request node, and distributing content, distribution of which is requested, to the request node according to the topological structure. The invention also provides the corresponding system. The method determines the shortest paths from all the edge nodes to the source node according to the distance measurement values between all the nodes, generates the router table on the basis of the shortest paths and issues the router table to all the edge nodes to generate the topological structure, so that the distance from each edge node to the source node is the shortest, and the time spent for a distributing node on distributing the content to the node requesting the content is the shortest. The method increases the speed of CDN content distribution and improves the system performance.

The invention discloses a method for selecting neighbor nodes for peer-to-peer network transmission. The method comprises the following steps: establishing a connection between a central node and target nodes, which download target files, in a cluster, and judging whether the target nodes have non-downloaded file fragmentations or not; if YES, inquiring network usage amounts and total network card bandwidths of the target nodes; calculating residual bandwidths of the target nodes according to the network usage amounts and the total network card bandwidths of the target nodes, and sequencing the target nodes according to an order of the residual bandwidths from large to small to obtain a sequence table; selecting the previous N target nodes in the sequence table as the neighbor nodes. The method can increase the data distribution speed on the premise of reducing the influence on normal operation of tasks of the cluster. The invention further discloses a system for selecting the neighbor nodes for peer-to-peer network transmission, a computer readable storage medium and a server which have the above beneficial effects.

The invention discloses a latent gaming-based method for distributing frequency spectrums in a cognition wireless system, for improving a frequency spectrum distribution speed in the cognition wireless system, and relates to the technical field of frequency spectrum distribution in a communication system. The method comprises: comparing a user randomly selected by a system and the last gaming user before each gaming begins, and if the two are the same user, commanding the system to reselect another user for this round of gaming, otherwise, performing next round of gaming. Since, repeated gaming of the same user cam not improve a signal-to-interference ratio, in the method provided by the invention, under the condition that system communication quality is ensured, the gaming frequency is reduced so that the system processing speed is enhanced. The provided method is suitable for the frequency spectrum distribution in the cognition wireless system.

The invention relates to the computer network field, more specifically a P2P acceleration content distribution network system and method. The system comprises a distribution task obtaining module, a resource management server, a resource memory, a resource tracking dispatch server and a plurality of cloud acceleration edge nodes. The distribution task obtaining module logs the information of distribution tasks in the resource management server through an operation interface by a hardware input device; the resource memory downloads and stores distribution task files, and distributes the distribution tasks to the cloud acceleration edge nodes; the resource tracking dispatch server performs P2P acceleration task distribution, monitors and records each process of task distribution, and feeds back data to the resource management server. Compared with the prior art, the system and method combine the advantages of P2P and cloud acceleration, perform MD5 verification on distribution tasks, monitor task distribution processes in real time, select an optimum download path for nodes, increase task distribution speed and improve system reliability.

The invention discloses a media information processing method, a server and a storage medium. The method comprises the steps of obtaining a source file of media information to be audited; extracting afirst digital fingerprint of the to-be-audited media information from the source file, wherein the first digital fingerprint comprises a first image fingerprint and a first audio fingerprint; according to the first digital fingerprint and a second digital fingerprint of at least one piece of target media information, determining whether the to-be-audited media information is similar to one pieceof target media information, the second digital fingerprint comprising at least one of a second image fingerprint and a second audio fingerprint; and when it is determined that the to-be-audited mediainformation is similar to one piece of target media information, determining an auditing result of the to-be-audited media information according to a preset auditing result of the target media information. By utilizing the technical scheme, the duplicate removal efficiency of the media information and the resource utilization rate of the server can be improved.

The invention provides a radial guide blade of a residual heat removal pump. The radial guide blade comprises a front cover plate, a rear cover plate and a plurality of the main blades, wherein the front cover plate and the rear cover plate are respectively fixed at two ends of the main blades in uniform circumferential distribution so as to form an annular water pressing chamber, a splitter blade is arranged every two adjacent main blades, an inlet diameter D0 of each splitter blade meets that D0=0.6-0.68 (D3+D4), and outlet diameters of the splitter blades are equal to outlet diameters of the main blades. The radial guide blade has the advantages that outlet areas of a diffusing section is reduced, diffusing efficiency of the diffusing section is lowered, speed distribution of fluid in passageways of the radial guide blade is improved, vortex backflow areas are reduced, operation efficiency of the residual heat removal pump is improved, acting force of the fluid to the main blades is improved, and operation of the residual heat removal pump is enabled to be safer and more reliable.

The invention provides a double windows based client fragment scheduling method and system. The method is applied to a P2P SVC streaming media system for scheduling data fragments in case of an equal time length or when the play time length between each layer changes in proportion. The method comprises the following steps: (101), a step for downloading a fragment needing a request again in an Emergent Window subwindow; (102) a step for downloading a fragment in a General Window subwindow; and (103), a window sliding, and returning to step (101) until fragment downloading is finished, wherein the Emergent Window subwindow and the General Window subwindow are two subwindows obtained by dividing a client sliding window according to a distance from a play point, the size of each subwindow is equal, the Emergent Window subwindow comprises all data fragments in the General Window subwindow before window sliding, and when download of some data fragments fails in the General Window subwindow before the window sliding, the Emergent Window subwindow after the window sliding requests to download failed fragments again.

The invention discloses a method for designing a centrifugal impeller splitter blade. The method includes the steps that a middle cross section, a front cover plate cross section and a rear cover plate cross section of an impeller are determined; three split flow points on the three cross sections are determined on the three cross sections respectively according to the principle that flows on the two sides of each split flow point are equal, and three bone lines on the three cross sections are obtained; and the three bone lines are mixed and scanned into a curved surface, the curved surface extends towards a front cover plate and a rear cover plate of the impeller till the curved surface intersects with the front cover plate and the rear cover plate, a work face of the impeller splitter blade is obtained, and the thickness of the work face is increased to obtain the back side of the splitter blade. The designed splitter blade can equally divide water flow in a flow channel between two adjacent long blades into two parts equal in flow, speed distribution at an outlet of the impeller can be effectively improved, hydraulic loss in the impeller is reduced, and the lift and the efficiency of the centrifugal impeller are improved.

The invention relates to the technical field of gas turbines, and provides an air inlet structure for a fuel nozzle of a gas turbine. The air inlet structure comprises a center air channel and a fuel cavity which are communicated with each other, and the center air channel is used for connecting the air into the fuel cavity. A nozzle hanging flange is arranged at one end of the center air channel, the nozzle hanging flange is provided with a plurality of air inlet channels used for connecting the air into the center air channel, and the air inlet channels are evenly distributed in the circumferential direction of the nozzle hanging flange. Each air inlet channel is provided with an adjusting mechanism, and each adjusting mechanism is used for limiting the size of the minimum flow area of the corresponding air inlet channel. Each adjusting rod can be adjusted independently, the speed distribution of the air on the head portion of a combustion chamber is effectively improved, the mixing effect of fuel and the air is improved, hence, tempering can be effectively prevented from happening at a nozzle outlet, and the reliability of the nozzle is improved. The air flow can be steplessly adjusted, stepless speed change of the air turbine can also be achieved, and the fuel can be fully burnt ultimately.

Provided is a heat exchanger which can suppress increase of pressure loss while improving nonuniform speed distribution of a fluid. In the heat exchanger (T), a fluid channel (4) is composed of flat containers (1, 2) having an inflow port (15) for the fluid at one end section and an outflow port (15) for the fluid at the other end section, and an offset type fin (5) arranged in the flat containers (1, 2). The heat exchanger is provided with a fin orthogonal intersection region (H) where the plate fin (5) orthogonally intersects with a flow direction of the fluid flowing from the inflow port (15) to the outflow port (16), and a fin parallel region (V) wherein the plate fin (5) is parallel to the flow direction of the fluid flowing from the inflow port (15) to the outflow port (16). The fin orthogonal intersection regions (H) are arranged on the sides of the inflow port (15) and the outflow port (16), and the fin parallel region (V) is arranged between the fin orthogonal intersection regions (H).

The invention discloses a peer-to-peer network transmission neighbor node selection method. The method includes the following steps that: connections are established between a central node and targetnodes in a cluster which download a target file; downloaded fragments of the central node are calculated, so that a first bit array is obtained, fragments stored in the target nodes are calculated, sothat second bit arrays are obtained; exclusive OR operation is performed on the first bit array and all the second bit arrays, so that the mutually-different data amount of the central node and eachtarget node can be obtained; and the target nodes are sequenced according to the large-to-small order of the mutually-different data amount, so that a sorted list can be obtained, and top N target nodes in the sorted list are selected as neighbor nodes. With the method adopted, data transmission rate can be improved under the premise that impact on the normal operation of a cluster task is decreased. The present invention also discloses a peer-to-peer network transmission neighbor node selection system, a computer readable storage medium and a server. The peer-to-peer network transmission neighbor node selection system has the advantages as the peer-to-peer network transmission neighbor node selection method.

A heat transfer panel, or multiple panels, utilized to absorbed heat from the turbine exhaust gas as part of the Rankin cycle which simultaneously distributes the exhaust gas through the waste heat boiler. The panel varies the gas flow characteristics across a transverse and longitudinal plane, thereby eliminating the need for a separate flow distribution device.

The method for quick distribution of web frame in the network equipment with flow control includes: according to the entry interface information of web frame counting it into correspondent virtual interface queue, regulating working condition of network interface, then adding the web frame into correspondent protocol queue to make analysis treatment, and taking out it and distributing it. and regulating total quantity of web frame and regulating working condition of correspondent network interface. Said invention can raise working stability of router, and raise its treatment speed.

The invention discloses a container mirror image processing method and device, computer equipment and a storage medium. The method comprises the steps of acquiring required file information used for generating a container mirror image, and constructing a complete mirror image according to the required file information; determining a to-be-deleted specified directory in the complete mirror image, and compressing the complete mirror image based on the to-be-deleted specified directory to obtain a compressed mirror image; obtaining directory content of a to-be-deleted specified directory, and generating to-be-mounted content based on the to-be-deleted directory content; and storing the to-be-mounted content to a block device file system, and mounting the compressed mirror image to the block device file system. According to the method, a specified directory in a constructed complete mirror image is deleted to compress the mirror image. The deleted directory content of the specified directory is used as the to-be-mounted content to be stored in the block device file system, and the compressed mirror image is mounted to the block device file system, so that the size of the to-be-distributed mirror image file is reduced, and the distribution speed of the container mirror image is increased.

The embodiment of the invention provides a network service method and device, electronic equipment and a readable storage medium, applied to the technical field of the Internet. The method comprises the steps of extracting a key feature in each user request, and according to the key feature and a pre-built corresponding relation between ISP network groups and the key features, determining the ISPnetwork group corresponding to each user request; aiming at each user request in each ISP network group, cyclically executing the steps of selecting an ISP cache node, computing the gain of the ISP network group and updating a strategy parameter in sequence according to a deep reinforcement learning algorithm, determining whether the target ISP cache node which maximizes the gain of the ISP network group and meets a user request condition exists, and if yes, responding to the user request via the target ISP cache node and transmitting request data to a user; and otherwise, responding to the user request via a CDN and transmitting the request data to the user. According to the method and device provided by the invention, the network service quality can be improved.

The invention provides a wind turbine energizer (1) based on aerodynamics for a wind driven generator. The wind turbine energizer comprises a guide plate (3) arranged in front of a wind turbine and provided with a streamline section and a supporting structure (2) thereof. Based on different hub heights and blade lengths of the wind turbine, on the premise of not interfering in rotation of wind turbine blades and not generating harmful pneumatic interference, the wind turbine energizer (1) which is proper in size is arranged near the front of the wind turbine. By forming a shrinking channel inthe upstream part of an impeller of the wind turbine, the average speed of an air flow in a region of the impeller of the wind turbine is improved, the speed is more uniformly distributed, and therefore, the efficiency of the wind turbine is improved, the starting air speed of the wind turbine is reduced, the operation stability of a wind wheel is improved, the external torque subjected to a rotary shaft of the wind wheel is reduced, and the service lives of the blades, hubs and a transmission system are prolonged.

The embodiment of the invention discloses a distributed search system, an index distribution method and a storage medium, and the system comprises a distributed index construction platform which is used for constructing an index, storing an index fragment in an index storage cluster, and writing resource description information into a state storage cluster, and the index fragment comprises a plurality of file blocks; an index storage cluster; a state storage cluster; a main control cluster which is used for sending an index downloading instruction to the state storage cluster; and a search cluster, including multiple search nodes, a plurality of search nodes form a peer-to-peer network, if the search nodes monitor an index downloading instruction, index fragments and file blocks to be downloaded are determined, and if at least two other search nodes have file blocks, the search nodes determine one of the at least two other search nodes as a downloading source of the file blocks and download the file blocks from the downloading source. According to the system of the invention, the index distribution speed is increased, and the search service stability is improved.

The invention discloses a data piece equipartition selecting system and method based on the number of client-sides and belongs to the information technology processing field. The system comprises a server-side, a Tracker server and the client-sides. The method comprises the following steps: the server-side encoding data to form a seed document, and analyzing the seed document to be connected to the Tracker server; the server-side sending parameters of selection orders of the seed document and data pieces to the client-sides; after receiving the parameters of the selection orders of the seed document and the data pieces, the client-ends analyzing the seed document and conducting data pieces selection and data loading according to the number of the data pieces of the seed document and the parameters of the selection orders of the data pieces. The data piece equipartition selecting system and method solves the problem that in a data piece random selection strategy, data pieces in a BitTorrrent system is not evenly distributed.

A cooler with a cooler main body that, in one configuration, includes a first wall portion forming a cooling surface that cools an electronic component, a second wall portion disposed opposing the first wall portion, and a side wall portion that connects a periphery of the first wall portion and a periphery of the second wall portion. In the configuration, cooling fins are attached to an inner wall surface of the first wall portion, a refrigerant introduction pipe and refrigerant introduction flow path for supplying refrigerant to the cooling fins and a refrigerant discharge pipe and refrigerant discharge flow path for ejecting refrigerant from the cooling fins are included, and protruding portions and are provided on the first wall portion side of the refrigerant introduction flow path.

The invention relates to a distribution method for the downlink resource of physical layer in a TDD-OFDMA system, comprising: distributing a DL-MAP sudden(11) according to the principle of first frequency domain and then time domain, the free region below the DL-MAP sudden is distributed for an UL-MAP sudden(12) which is smaller than the DL-MAP in length; the undistributed UL-MAP sudden(12) is distributed in the effective region in a rectangular block form according to the principle of first frequency domain and then time domain, the time domain of the rectangular block is described by the right formula, the undistributed data sudden is selected in the descending order in the rectangular block form to be distributed in the effective region in turn. The method is simple to realize, avoiding the solving complexity of NP complete problem, increasing the speed of the distributing of the downlink resource of the physical layer, the method can provide an optimized preallocated bandwidth, enhance the rates of resource utilization efficiency, and reduce the explosion frame rate.