204results about How to "Guaranteed processing speed" patented technology

The invention relates to an image postprocessing method for removing a metal artifact from a computed tomography (CT) image, which comprises the following steps of: converting an original CT image into a polar coordinate image from an orthogonal coordinate image; determining a metal projection region in the polar coordinate image; establishing a model in the polar coordinate image; performing model correction by adopting the model; correcting front and back projection errors brought into the model correction; and converting the polar coordinate image into the orthogonal coordinate image. According to a model establishing method provided by the invention, the good model can be established, and finally, a good correction result can be obtained; the errors brought in by front and back projection are corrected, so that the result after the correction is better, and the processing speed is ensured.

Disclosed is a system for improving local electric network power supply ability analysis and providing entire process online monitoring. The system comprises a sub-system receiving original power data in targeted areas, wherein the sub-system classifies the original power data according to node data, line data and transformer data; a sub-system computing power supply capability according to power data, wherein the sub-system uses the improved trial-and-error method based on repeated trends to compute the power supply capability of the local electric network. The computing process comprises that a critical point is obtained under a predetermined operation mode and a load growth mode, a limited condition exactly at the critical point is effective, a threshold crossing occurs when load increases over the critical point, the critical point corresponds to the maximum power supply capability of the local electric network, and the available power supply capability is the difference between the maximum power supply capability and the present interlocal transmission load. The system also comprises an aid decision-making sub-system evaluating the power supply capability of the targeted areas according to computed results, wherein the sub-system provides detailed parameters of weak devices with threshold crossing priority in operation, thereby performing an alarm mechanism.

The invention discloses a 5G network element task processing method and relates to the field of communication. The method comprises the steps of sending network element state monitoring instructions to registration network elements when an NRF (Network Repository Function) receives network element registration request messages; receiving state information returned by each registration network element; carrying out statics on a real-time state of each registration network element according to the state information and generating network element state records; determining task execution networkelements according to received task processing requests and network element state records; and sending task processing instructions to the task execution network elements, thereby enabling the task execution network elements to carry out corresponding task processing. According to the method, the network element selection accuracy can be improved, and network task processing speed is ensured. Theinvention also discloses the NRF, a 5G network element task processing system and a readable memory medium which have the beneficial effects.

An embodiment of the invention discloses an anti-deadlock method, a device and an on-chip system of a system bus and relates to the on-chip system technical field to solve the problem that the existing deadlock prevention technology can affect the performance of the system bus to a greater extent. The embodiment of the invention blocks the operations which can possibly cause deadlock, and by sending a priority improving command, the embodiment helps the equipment to speed up the processing speed of the previously sent operation command which causes deadlock root. The embodiment of the invention is mainly used in the bus-required systems or equipment, such as AXI system bus.

The invention relates to the data backup technology, and discloses a fast incremental backup method aiming at the problems that the incremental backup method in the prior art is too long in freezing system written time and is easy to cause system jamming. The incremental backup method places the generation bitmap and the comparison bitmap outside the written time of freezing and unfreezing system, greatly reduces the written time of continuous freezing system, and relieves the influence on the normal operation of the users. The method generates the bitmap needing incremental backup directly by comparing the hash value of file description, namely the file description different from the file description in images should be backup. The method adopts the comparison mode for incremental backup, does not need continuous monitoring rolls or sectionalized change, and can avoid the problem caused by discontinuous monitoring under abnormal condition. The fast incremental backup method is mainlyused for disc data incremental backup, especially suitable for the data incremental backup in NTFS system.

The invention relates to a method for processing a subsurface gate by discharging, comprising the following steps of: manufacturing a rough processing electrode and a fine processing electrode according to the shape of a gating port; using a turning machine to turn the rough processing electrode and the fine processing electrode; arranging an embedding piece on an electric discharge machine working platform and enabling the gating direction of the embedding piece to be vertical to the electric discharge machine working platform; finding the coordinates of X axis and Z axis of the gating port, using the rough processing electrode for rough processing to the position 0.4mm-0.6mm above the bottom of the gating port; and replacing the rough processing electrode with the fine processing electrode and carrying out fine processing till the required value is reached. By the method, the subsurface gate can be fast processed, and simultaneously the processing accuracy also can be guaranteed.

A method and apparatus for training a model are disclosed in embodiment of that present application. The method comprises the following steps: determining M computing nodes of the first data subset from the block chain system, wherein the client device can select K computing nodes from the M computing nodes for model training and obtain the training results of P computing nodes from the K computing nodes; Further, the client device may obtain the model parameters according to the training results of each data subset in the Y data subsets. In the embodiment of the present application, by adopting the training model of the computing node in the block chain system, it is unnecessary to construct a dedicated machine learning algorithm group, thereby saving the cost and reducing the waste of the social resources; Furthermore, the training efficiency of the model can be improved by selecting the computing nodes with strong storage and processing ability in the block chain system, and then the better training effect can be obtained.

The invention discloses a processing method and device of a transparent material, and provides an optical fiber laser of which the wave length is 1020 nm to 1090 nm. The processing method is characterized in that laser beams output by a seed laser are subjected to energy amplification through an optical fiber amplifier, laser pulse trains are output, each pulse train comprises at least two laser pulses, pulse width is smaller than 200 ps, peak power of the pulses is larger than 100 kW, the time among the laser pulses in each pulse train is smaller than 90 ns, the interval among the pulse trains is larger than 240 ns, and the total pulse number in each second is larger than 50,000, wherein ultrafast laser beams output by the laser are subjected to energy amplification only through the optical fiber amplifier; and the laser is focused on a to-be-processed position of the transparent material, a focus position is moved, processing is carried out according to a set track, a focus point isgradually risen, and the transparent material is processed from bottom to up. According to the processing method, material removal efficiency can be greatly improved, a heating or etching follow-up procedure is not required, and bevel angles are avoided.

The invention relates to the technical field of waste gas treatment. The invention further discloses an industrial waste gas treatment device utilizing a Tesla valve principle. The device comprises ashell, the outer part of the shell is fixedly connected with an air inlet pipe, a brush shaft is movably connected into the air inlet pipe, a partition plate is fixedly connected to the outer part ofthe brush shaft, a brush is fixedly connected to the outer part of the partition plate, a screen is movably connected to the outer part of the brush, a dust passing pipe is fixedly connected to the outer part of the air inlet pipe, a dust collecting box is fixedly connected to the outer part of the dust passing pipe, and an air vent is formed in the outer part of the air inlet pipe. The screen, the brush and the partition plate are matched for use so that the effect of preliminary waste gas treatment is realized, large particles in the waste gas are removed through pretreatment on the waste gas, so that the situation that waste gas treatment is affected and the waste gas treatment effect is improved due to the fact that a subsequent filter screen is blocked is prevented, and the waste gastreatment effect is improved, and the brush continuously brushes and cleans the screen, so that the effects of preventing the screen from being blocked and improving the filtering effect are achieved,and therefore the waste gas treatment effect is guaranteed.

The invention discloses a method and device for the encryption and decryption of an IP message in satellite communication. The method comprises the steps: receiving message source data which comprises a data part and a TCP/UDP packet header part; carrying out the encryption of data carried by the data part, and obtaining a ciphertext data part; transmitting the IP message comprising the TCP/UDP packet header part and the ciphertext data part, wherein the TCP/UDP packet header part is not encrypted. According to the invention, the method and device meets the encryption and QoS requirements.

The invention discloses an ultrafast ultraviolet laser processing method and device of a transparent material, and provides an optical fiber laser of which the wave length is 1020 nm to 1090 nm. The processing method is characterized in that energy amplification is carried out through an optical fiber amplifier; optical third harmonic generation is carried out, and ultraviolet laser pulse trains are acquired; each pulse train comprises at least two laser pulses, pulse width is smaller than 200 ps, peak power of the pulses is larger than 50 kW, the time between the adjacent laser pulses in thepulse trains is smaller than 90 ns, the interval among the pulse trains is larger than 240 ns, and the total pulse number in each second is larger than 50000; ultrafast laser beams output by the laserare subjected to energy amplification only through the optical fiber amplifier; and the ultraviolet laser beams are focused on the lower surface of the to-be-processed transparent material, processing is carried out according to a set track, a focus point is gradually risen, and the transparent material is processed from bottom to up. According to the processing method, material removal efficiency can be greatly improved, a heating or etching follow-up procedure is not required, and bevel angles are avoided.

The invention discloses a processing method and device for an ultrafast green laser transparent material and provides an optical fiber laser, the wave length of which is ranged between 1020 nm and 1090 nm. By adopting the optical fiber laser for amplifying energy, the processing method is characterized by comprising the following steps: then carrying out optical frequency doubling to obtain a green laser pulse string, wherein each pulse string comprises at least two laser pulses, the pulse width is smaller than 200ps, the peak power of the pulse is greater than 100 kW, the time of adjacent laser pulses in the pulse string is shorter than 90 ns, the interval time among the pulse strings is greater than 240 ns, and the total pulse number per second is greater than 50,000; carrying out energyamplification on the ultrafast laser string output by the laser only with the optical amplifier; and focusing the green laser string to the lower surface of a to-be-processed transparent material, and processing the material according to a set track and raising the focus point gradually so as to process the transparent material from bottom to top. According to the processing method, the removal efficiency of the material can be improved greatly without a follow-up step of heating or acid invasion and generation of an oblique angle.

The invention provides an automatic impedance matcher for a magnetostrictive transducer. The automatic impedance matcher comprises an impedance matching module, a signal collection module, a feedbackmodule, and a single-chip microcomputer control module; the impedance matching module comprises two sets of 8-channel 5V relay modules with optocoupler isolation functions, a 5V DC power supply, 16 CBB matching capacitors and 16 resistors with resistance of 1kohm; the two sets of relay modules have the same structure and respectively comprise 8 relays connected in parallel; DC + and DC- Ports of each set of relay modules are connected to a positive electrode c and a negative electrode d of the 5V DC power supply respectively; and an assorted H jumper cap is respectively inserted into each of high and low level trigger selection ends S1 to S16 of each relay. The automatic impedance matcher provided by the invention enables the transducer to continuously and stably work in a resonant state under different environments, working conditions and loads, thereby significantly improving the working efficiency of the transducer.

The invention discloses invalid template generation method and device, relating to the technical field of information processing. The ineffective template generation method comprises the following steps of: acquiring a seed invalid template set, wherein the seed invalid template set includes one or multiple seed invalid web page templates; generating a candidate invalid template set according to the seed invalid web page templates and web pages inside a local web page database; and screening to obtain a finial invalid template set from the candidate invalid template set. The invention also discloses invalid web page identification method and device. The invention can be used for fast and accurately automatically identifying invalid web pages.

The invention discloses a data processing flow design method based on NIFI, which relates to the data processing field. The method includes giving full play to the cluster performance to get the processors that collect the list of file names; Distributing the list of files to processors on each node of the nifi cluster through kafka; Processor capable of multi-node and multi-thread parallel downloading of remote computer files; A processor that parses different acquisition files according to different file parsing rules, solves the single-point bottleneck problem of data processing, and can solve the problem of re-customizing the code when the file format changes by flexibly configuring collection instances, concurrency, scheduling policies, and file parsing rules.

The invention relates to a cable forming mechanism, in particular to a flame-retardant power cable processing system and method. The system comprises a processing bracket, a power mechanism, connecting frames, a reel mechanism, a pressing mechanism, a winding frame and a winding mechanism. Linear flame-retardant rubber is arranged on the winding frame, and the flame-retardant rubber is wound on acable core. The winding mechanism is fixedly connected to the processing bracket. The winding frame is in contact with the winding mechanism. One end of the cable core is fixed on the reel mechanism.The cable core is pulled by the power mechanism to pass through the winding frame, the winding frame is driven by the winding mechanism to wind wire-rope-like glass fibers and the flame-retardant rubber on the cable core in turn, and hot-press forming is carried out when the cable core and the materials wound on the cable core pass through an extrusion forming wheel I and an extrusion forming wheel II. An efficient flame-retardant power cable can be processed and produced efficiently and quickly, and a cable core can be coated with a flame-retardant material efficiently and quickly.

The invention relates to a stage curtain integrity identification mechanism. The mechanism comprises a curtain unfolding device connected to a stage curtain for controlling the stage curtain to be switched between an unfolded state and a contracted state and sending an unfolding prompt signal when the stage curtain is in the unfolded state; a hemisphere camera used for image data capture for the stage curtain to obtain and output a current screen image; a noise detecting device connected to the hemisphere camera for receiving the current screen image and performing noise detection on the current screen image to obtain a noise amplitude corresponding to the current screen image to output the amplitude as the current noise amplitude. The mechanism achieves effective monitoring of the overalldegree of lack of the curtain.

The invention discloses a dark room broadband RCD measurement method based on compression sensing. The method comprises steps of (1) designing a sparse measurement matrix, and configuring compression sampling testing parameters; (2) placing a scaling body and a measurement object on a rotary table in sequence, and completing echo data sampling of the measurement object, wherein the rotary table rotates at the uniform speed in the sampling process; (3) completing scaling processing to the measurement object according to the obtained sampling data; (4) subjecting object data after scaling to sparseness by the sparse measurement matrix; and (5) using a two-dimensional data reconstruction algorithm to perform data reconstruction of other angles and other frequencies. The method utilizes the coupling of the two-dimensional signal, and improves the imaging resolution under a precondition of the same data size.

The invention discloses a method and a system for detecting the uniformity of pixels of a display screen. The method processes display screen photos by calculating the center distance between light points of the display screen and performing skipping according to the center distance. The system comprises a communication interface, a user input interface, a storage module, a scanning module and ananalysis module. The method and the system can detect the uniformity and dead pixels of the pixels of the display screen, and accelerate the detection speed and the detection accuracy rate.