36results about How to "Overcome the disadvantage of slow speed" patented technology

The invention discloses an FPGA dynamic power consumption estimation method based on a BP neural network. The FPGA dynamic power consumption estimation method comprises the following steps that (1) the input and output quantities of four modules are obtained as sample data; (2) data screening and data preprocessing are carried out on the sample data; (3) BP neural network models of the four modules are respectively constructed according to the processed sample data; (4) part of the sample data are adopted as the training data of the BP neural network, the trained neural network is obtained after BP neural network training is carried out, and then power consumption of neural network output is obtained; (5) the sample data with the training data removed are adopted as the testing data of the BP neural network, and the obtained power consumption is compared with the testing data; (6) the power consumption output by the neural network is restored to be actual power consumption values; (7) the obtained power consumption estimation values of the four modules are summated to obtain a total power consumption value. The power consumption values can be accurately predicted through automatic study of the BP neural network.

The present invention discloses a rapid simulation method of repeat-pass spaceborne natural scene SAR complex image data. According to the method, an SAR satellite orbit is simulated, the ground scene is simulated according to the spaceborne SAR satellite imaging principle and an SAR satellite single looking complex image signal model, and multi-frame repeat-pass SAR complex image data are directly obtained. Slant-range information is obtained by using a rapid method to iteratively compute a satellite zero doppler imaging position, a backscattering coefficient is calculated by using a simplified scattering model, and a simplified repeat-pass model is provided to generate any multi-frame SAR complex image data with base length controllable. The disadvantage that speed of a traditional echo simulation method is slow is overcome, and prior information can also be provided for interference processing so as to assist in processing or evaluating a processing result.

The invention discloses a segmentation method of a high-noise gray-scale non-uniform image. The segmentation method is an image segmentation method based on combination of a distance maintenance level set method and a mark watershed method. In the segmentation method of the invention, advantages of the above two methods are used to make up mutual defects; segmentation of a target containing a high noise and a non-uniform gray scale simultaneously is realized; an edge of an interested object can be accurately segmented; and an excess segmentation problem of the image in a traditional watershed method can be overcome. Compared to a traditional level set method, by using the segmentation method of the invention, an operation speed is fast and operation time of an algorithm is not changed along with changes of an image size.

The invention discloses a machine game optimization method. Innovation of the method is using a probability matrix to analyze magnitude of enemy chess pieces force so as to decide a chess strategy of our party more purposefully. The method mainly comprises probability matrix design and an update process. The design method mainly comprises setting dimensions according to a chessboard layout and a chess piece type and quantity of the military chess, and determining an initial value of a matrix according to military chess rules and a history chess manual of an enemy; updating rules relate to updating both rows and columns in the probability matrix; and finally, after updating location information of the chess pieces, the chess strategy of our party is more purposeful, and a winning probability can be significantly increased. According the optimization method, without further expansion of a game tree, a probability distribution model is adopted to simulate a real situation, search and evaluation are carried out targeting on a current situation and prediction of a possible strategy of the enemy in a next step, and the method can be easily realized through a machine language, and the software operation speed is fast, thereby improving speed and the winning probability of machine playing chess, and overcoming shortcomings of both a slow speed and a complex search space.

The invention discloses a test method and system of a control storage chip. The method comprises: pre-storing, by a CPU (central processing unit), a first address and first data in an interface simulation accelerator through a system bus; transmitting, by the interface simulation accelerator, the first address and the first data to a storage interface control module through a data bus; writing, by the storage interface control module, the first data into a storage unit, corresponding to the first address, of a storage chip; reading, by the interface simulation accelerator, data stored under the first address through the data bus to obtain second data; acquiring, by the CPU, the second data from the interface simulation accelerator, judging whether the second data and the first data are identical or not, and giving a pass to the test if yes. The test method and system according to the technical scheme disclosed herein have the advantages that the interface simulation accelerator is innovatively designed by combining the speed of hardware test with flexibility and controllability of software simulation test, and comprehensive rapid verification of a storage interface control module is achieved via the interface simulation accelerator.

The invention discloses a multi-fusion indoor positioning method and system based on WiFi fingerprints. The positioning method comprises the following steps: obtaining a WiFi fingerprint, and constructing a WiFi fingerprint library; carrying out clustering analysis on the collected WiFi fingerprint data by adopting a clustering algorithm; carrying out secondary classification on the clustered databy utilizing a classification algorithm, and judging the sub-regions where the clustered data are located; and obtaining a regression equation of each sub-region according to the data subjected to the secondary classification, and solving coordinates of the point to be positioned by utilizing the regression equation of the sub-region. According to the positioning method and system, the limitationthat the positioning precision is easily influenced by physical environment factors is avoided, and the defect that a traditional positioning method is low in speed is overcome.

The invention discloses a surface acoustic wave regulated high-flux micro-droplet generation device and a surface acoustic wave regulated high-flux micro-droplet generation method. The device comprises an interdigital transducer, wherein an arc-shaped interdigital electrode is manufactured on the interdigital transducer, and a PDMS micro-flow channel system is bonded to the upper portion of the interdigital transducer, and is formed by bonding a top-layer PDMS flow channel, a middle-layer PDMS flow channel and a bottom-layer PDMS flow channel. The method comprises the following steps: fixing the device on an objective table of a microscope; adjusting a dispersion phase and a continuous phase to corresponding flow rates through an injection pump; at each T-shaped micro-flow channel in the device, shearing the dispersion phase into micro-droplets by the dispersion phase, pressing down the output button of a signal generator, and generating converged acoustic surface waves on the interdigital transducer, wherein an acoustic pressure field acts on the continuous phase to increase corresponding flow resistance, and is used as an acoustic valve to regulate and control the flow velocity of the continuous phase, so that the regulation and control of the droplet size are realized. According to the invention, the amplitude and the frequency of the input sine voltage are adjusted while the high-flux generation of the liquid drops is ensured, and the generation size of the micro-liquid drops can be flexibly regulated and controlled in real time.

The invention relates to a nanometer device and a nanometer machining technology, in particular to the structure of a radio-frequency single-electron transistor scanning probe and application thereof. The radio-frequency single-electron transistor scanning probe comprises an SOI (silicon on insulator) substrate of which one end has a needle-shaped lug boss, and a radio-frequency resonant circuit and a single-electron transistor integrated on the SOI substrate and matched with each other; and the single-electron transistor is integrated on the needle-shaped lug boss of the SOI. The invention further provides a manufacture method of the radio-frequency single-electron transistor scanning probe and a charge scanning imaging and oscillography and atomic power showing imaging system manufactured by the radio-frequency single-electron transistor scanning probe. The system is high in sensitivity, high in speed and high in spatial resolution, can form into the distribution morphology and the geometric morphology of the charge by means of low-temperature three-dimensional scanning, and can detect the electric potential and the dynamic change of the electric potential at any one point in the nanometer device or circuit at fixed points.

The invention discloses application of an N-oxidized tertiary amine containing polymer as a medicine or a carrier. The invention discloses application of the N-oxidized tertiary amine containing polymer as an anti-tumor medicine, an anti-tumor medicine carrier or a gene delivery carrier. The polymer has a structural frame as shown in the specification, in the formula, A1 is a fragment with an N-oxide structural unit; A2 is a fragment for connecting medicine molecules, a fragment with the N-oxide structural unit, or a fragment connected with medicine molecules; x=0-1, n0=3-300, m=3-300, and n=3-300. The N-oxidized tertiary amine containing polymer disclosed by the invention is good in both water solubility, biocompatibility and blood shielding property, can be rapidly taken into cells and can be fed into a tumor anoxia area, and thus by adopting the polymer as a carrier, medicines can be effectively delivered to the tumor anoxia area, and efficient anti-tumor functions can be brought into play. The N-oxidized tertiary amine containing polymer is an excellent medicine delivery carrier material.

The invention provides a method for file storage in continuous space of a flash memory in an equal probability manner, wherein the storage times of storage units are controlled by sequencing the storage times, and on the basis, the storage units with same storage times are subjected to continuous space sequencing to control the continuity of file storage and realize the continuous sequencing of the storage units according to the quantity from most to least. The method comprises the following steps of: supervisory sequence storage area formatting, file storage and file deletion. According to the method disclosed by the invention, the storage units with same storage times are sequenced on the basis of the storage times sequencing, the defects that the traditional file storage operation is unbalanced and a speed of a flash file storage system is low are overcome, the service life of the file storage space is prolonged, the access efficiency of the file storage space is improved, and favorable performance and practicability are achieved.

The invention provides a multilayer continuous elevator capable of continuously conveying a large number of cargoes and meanwhile capable of carrying out conveying in a one-point-to-multipoint manner. A table board is fixed to a chain set, chains are wound around in a combined manner through chain wheels, and on an inlet, through chain wheel redirection, platform face cargo carrying is formed; then, the table board vertically ascends or descends, and on an outlet, the cargoes are output through chain wheel redirection; when a floor inlet or outlet needs to be changed, a lifting car is lifted to reach the new position, and braking fixation is achieved; in the aspect of chain cycle, the table board is fixed to the four chains, and during ascending, a cargo carrying platform is formed; the front row of chains and the rear row of chains are arranged in a staggered manner, the front row of chains and the rear row of chains output the cargoes on the same plane, and through chain wheel redirection, the table board carries out conveying in the vertical direction; a balancing weight is additionally arranged on one side of the lifting car, a lifting motor drives a car body to wholly ascend or descend, driven chain wheels are arranged on the conveying chains winding around the car body, the car body ascends and descends, and chain wheels in the car body are driven along with ascending and descending of the car body; and when the elevator reaches the needed position, the motor stops and carries out braking.

The invention provides a WEB service matching method based on an overlapped oriented graph and relates to the field of computer information analysis and data processing. Description files (WSDL) of a large amount of WEB service nodes are loaded into a server side internal memory to form massive amounts of WEB description source data, a vacancy seed matching model for matching of specified service is built, the overlapped oriented graph is constructed on the basis, seeds are evaluated through two key indexes including connectivity and overlapping number to find optimal seeds, the WEB service matching is finished finally, and the WEB service matching method is quick in finding and appropriate. The WEB service matching method can be widely applied to an environment with a large amount of WEB service nodes and high WEB service finding requirement, such as a service oriented architecture (SOA) and a cloud computing environment.

The invention relates to an unmanned aerial vehicle ground detection method and system based on confidence, and the method comprises the steps: carrying out the forward reasoning of video data collected frame by frame based on a model file and a weight file of a trained target detection deep neural network, and obtaining a position region of a to-be-detected target, obtaining a candidate region corresponding to the position region of the to-be-detected target in the current video frame and a region consistent with the candidate region corresponding to the current video frame in the next videoframe, determining the position region of the to-be-detected target, taking the position region of the to-be-detected target as a target position, storing the target position in a target position setand carrying out outputting. According to the technical scheme provided by the invention, the video acquired by the unmanned aerial vehicle in real time is monitored by adopting the deep neural network target detection method, the detection efficiency of the unmanned aerial vehicle is improved, and the method and system have important application value.

The invention discloses a mildew-proof treatment coating for printed wrapping paper for artistic design. The mildew-proof treatment coating is prepared from components in parts by weight as follows: 25-30 parts of water-based acrylic modified polyurethane resin, 10-20 parts of p-hydroxybenzoate, 20-30 parts of silver-loaded zeolite, 1-3 parts of sodium bicarbonate, 5-10 parts of stearamide, 10-15 parts of a plant anti-bacterial liquid, 1-3 parts of anise oil and 3-5 parts of a surfactant, wherein the plant anti-bacterial liquid is prepared from components of raw materials in parts by weight as follows: 20-30 parts of flos caryophylli, 15-20 parts of cortex hibisci, 10-15 parts of rhizoma kaempferiae, 10-15 parts of herba eupatorii and 3-5 parts of herba schizonepetae. The invention further provides a preparation method of the mildew-proof treatment coating for the printed wrapping paper for artistic design. A chemical mildew-proof bactericide is combined with a plant mildew-proof bactericide for use, the defects of the bactericides used separately are overcome, the two bactericides are matched in use and have a synergistic function, and the mildew-proof treatment coating has a wide prospect.

The invention discloses a fast watershed detection method for a circular mark in a helicopter blade image, which comprises the following steps: 1) constructing a parallel processing frame; 2) fast extracting local extremum features based on image downsampling; 3) identifying the local extremum of the circular mark based on the preset pattern; 4) performing fast marking of the region of interest (ROI) of the circular mark based on a lookup table; 5) performing ROI local gradient calculation; 6) performing ROI local watershed transform. The invention has the advantages that the single pixel linewidth continuous edge of all circular marks in the low-exposure blade image can be accurately extracted, and the detection efficiency of the circular marks in the helicopter blade image is improved; at the same time, the method has the advantages of strong anti-jamming ability and high degree of automation, and can be used to detect and locate circular marks in the visual measurement of helicopterblade motion parameters.

The invention relates to a high-precision advance aiming angle compensation device for satellite optical communication, which relates to an advance aiming compensation device in a satellite optical communication device. It overcomes the disadvantages of relatively complex actuators and slow dynamic response in the prior art. It consists of an arithmetic processor (1), a drive controller (2), an aiming actuator (3) and an angle sensor (4). The output of (1) is connected to an input of (2), and the output of (2) The terminal is connected to the input terminal of (3), (4) is set on (3), the output terminal of (4) is connected to the other input terminal of (2), and (3) is formed by the No. 1 piezoelectric ceramic column (3-1) , No. 2 piezoelectric ceramic column (3-2), worktable body (3-3), worktable surface (3-4) and flexible hinge (3-5), (3-5) is fixed on (3-4 ) on the edge of the lower surface and (3-3) on the edge of the upper surface, (3-1), (3-2) are set inside (3-3), (3-1) and (3-2) on the The end face abuts on the lower surface of (3-4), and (3-5) is arranged on the side of (3-1) and (3-2) and has an equal distance from them.

The invention discloses a fast automatic gain control method of an intermediate-frequency digital receiver. The method comprises the following steps: firstly acquiring a signal intensity posterior knowledge table to judge signal intensity, and then acquiring an attenuation parameter posterior knowledge table, recording AD values of intermediate-frequency output peaks of the signals with differentintensities under various attenuation conditions of a radio frequency front-end device as the attenuation parameter posterior knowledge table, wherein the attenuation parameter posterior knowledge table is used for querying the attenuation gear stage needing to be set by the radio frequency front-end device when the AD value of the signal peak at a certain intensity is located in a specified range. When the signal is relatively small, once setting and querying process can be performed so as to reduce the attenuation degree of the radio frequency front-end device after the first setting and querying. But the frequency needing to be set is located within trice, and the fast automatic gain control of the receiver is realized.

The invention provides an air quality index obtaining method and device. The method comprises the following steps of: receiving atmosphere data sent by each monitor device; summarizing the atmospheredata to generate an atmosphere data file; sending the atmosphere data file to a distributed processing system; and receiving an air quality index returned by the distributed processing system, whereinthe air quality index is obtained by the distributed processing system after carrying out distributed processing on the atmosphere data file. According to the method and device, abundant atmosphere data files are processed by utilizing the distributed processing system, so that the defect that existing analytical processing manners of intensively monitoring data are low in speed is overcome, airquality indexes can be obtained more rapidly and the timeliness of air quality monitor is improved.

The invention belongs to the field of unmanned aerial vehicle communication, and relates to an unmanned aerial vehicle system based on a variable message format data link and a communication method. The system comprises a man-machine interaction module, a message generation module, a segmentation / recombination module, a sending module, a receiving module and a message analysis module. Data element extraction and data element dictionary query are carried out on a message needing to be sent, so that the function domain to which the message belongs, the message number and the message content are obtained, a binary message is generated according to the function domain to which the message belongs, the message number and the message content, the length of the message is changed along with the size of useful information, the data sending amount is reduced, and limited bandwidth resources are saved. In addition, a UDP protocol is adopted for transmission, and a segmentation / recombination mechanism and a retransmission confirmation mechanism are added, so that the transmission reliability is ensured.

The present invention provides a WEB service matching method based on overlapping directed graphs, which relates to the field of computer information analysis and data processing. By loading a large number of WEB service node description files (WSDL) into the server-side memory, a large number of WEB descriptions are formed. In order to match the source data to the specified service, a vacancy seed matching model is established, and an overlapping directed graph is constructed on this basis, and the seeds are evaluated by the two key indicators of connectivity and overlapping number, and the optimal seed is found, and finally completed Matching of WEB services is a method to quickly find suitable WEB services. The present invention can be widely applied in environments where there are a large number of WEB service nodes and the search for WEB services is highly required, such as service-oriented architecture (SOA), cloud computing and other environments.

The invention discloses an FPGA dynamic power consumption estimation method based on a BP neural network. The FPGA dynamic power consumption estimation method comprises the following steps that (1) the input and output quantities of four modules are obtained as sample data; (2) data screening and data preprocessing are carried out on the sample data; (3) BP neural network models of the four modules are respectively constructed according to the processed sample data; (4) part of the sample data are adopted as the training data of the BP neural network, the trained neural network is obtained after BP neural network training is carried out, and then power consumption of neural network output is obtained; (5) the sample data with the training data removed are adopted as the testing data of the BP neural network, and the obtained power consumption is compared with the testing data; (6) the power consumption output by the neural network is restored to be actual power consumption values; (7) the obtained power consumption estimation values of the four modules are summated to obtain a total power consumption value. The power consumption values can be accurately predicted through automatic study of the BP neural network.

The invention provides a multilayer continuous elevator capable of continuously conveying a large number of cargoes and meanwhile capable of carrying out conveying in a one-point-to-multipoint manner. A table board is fixed to a chain set, chains are wound around in a combined manner through chain wheels, and on an inlet, through chain wheel redirection, platform face cargo carrying is formed; then, the table board vertically ascends or descends, and on an outlet, the cargoes are output through chain wheel redirection; when a floor inlet or outlet needs to be changed, a lifting car is lifted to reach the new position, and braking fixation is achieved; in the aspect of chain cycle, the table board is fixed to the four chains, and during ascending, a cargo carrying platform is formed; the front row of chains and the rear row of chains are arranged in a staggered manner, the front row of chains and the rear row of chains output the cargoes on the same plane, and through chain wheel redirection, the table board carries out conveying in the vertical direction; a balancing weight is additionally arranged on one side of the lifting car, a lifting motor drives a car body to wholly ascend or descend, driven chain wheels are arranged on the conveying chains winding around the car body, the car body ascends and descends, and chain wheels in the car body are driven along with ascending and descending of the car body; and when the elevator reaches the needed position, the motor stops and carries out braking.

The invention provides a three-dimensional microfluidic device and method for generating high-flux micro-droplets. The device comprises a closed type flow channel structure formed by bonding a bottom-layer PDMS flow channel, a middle-layer PDMS flow channel and an upper-layer PDMS flow channel; the closed type flow channel structure is provided with a dispersed phase inlet connector, a continuousphase inlet connector and a collection outlet connector; the method comprises the following steps: fixing the three-dimensional microfluidic device on an objective table of a microscope and starting an injection pump; firstly, adjusting a dispersed phase and a continuous phase to corresponding flow speeds through the injection pump; and shearing the dispersed phase into the micro-droplets by the continuous phase on each T-shaped flow channel array for generating the micro-droplets in the device. By adopting the three-dimensional microfluidic device and method for generating the high-flux micro-droplets, the disadvantage that the speed for generating the micro-droplets by a two-dimensional flow channel at present is slow is overcome, and the generation of high-flux micro-droplets is realized; the whole structure is PDMS and is very firm after the PDMS is mutually bonded, so that a phenomenon that the flow channels are cracked due to excessive pressure of input fluid between the flow channels is avoided; and the PDMS is thoroughly transparent so that the device is easier to observe.

A device and method for rapid generation of core-shell droplets based on surface acoustic wave microfluidics, including a droplet control system and a core-shell droplet generation system; the droplet control system includes an interdigital transducer, and the upper part of the interdigital transducer Bonded with a voice-activated PDMS microchannel, arc-shaped electrodes are made on the interdigital transducer, and the arc-shaped electrode cooperates with the acoustic-activated PDMS microchannel. The acoustic-activated PDMS microchannel is equipped with a dispersed phase inlet joint, and the lower part of the interdigital transducer is bonded. There is a core-shell droplet generation system; the core-shell droplet generation system is formed by bonding the top layer of PDMS, the second layer of PDMS microchannels, the third layer of PDMS microchannels and the bottom layer of PDMS microchannels. Capillary glass tube; fix the surface acoustic wave microfluidic core-shell droplet rapid printing device on the stage of the microscope, generate core micro-droplets at the T-shaped microchannel, and cut them into cores at the end of the capillary tapered mouth Shell-shaped droplets; the surface acoustic wave acts on the straight channel, which can realize the regulation of the core droplet size inside the core-shell structure.

The present invention provides a method and device for realizing information demodulation in wavelength label technology. The method includes: sampling an optical signal with a wavelength label at a specified sampling rate; performing frequency analysis on the collected data to obtain the frequency spectrum of the wavelength label signal; Separate the spectrum of each modulation frequency from the spectrum of the wavelength label signal; calculate the average power spectrum of each modulation frequency in a specified time period in real time; respectively from a range that is larger than the specified multiple of the average power spectrum of each modulation frequency, Find the spectrum of the maximum value; obtain the synchronization head of the wavelength label corresponding to the modulation frequency according to the sampling time information corresponding to the spectrum of the maximum value. The invention utilizes the wavelength division multiplexing system of the wavelength label technology to realize the fine control of the modulation depth of the wavelength label signal, and overcomes the disadvantages of low efficiency of synchronous demodulation and slow information transmission rate of the wavelength label.

An adaptive online method for generating subgroup parameters. 1. For a given number of states, select the corresponding number of background section points, and interpolate to obtain the temperature of the specific problem and the section under the corresponding background section points. Combine the subgroup intermediate section and subgroup probability; 2. Select at least 20 background section points, and use Pad approximation to fit the subgroup subsection; 3. Calculate the subgroup parameters of 2-6 states separately, according to the stability criterion and the fitting error criterion to select the optimal subgroup parameters; in the method of the present invention, the Pad approximation is used to fit the subgroup parameters, the calculation speed is fast, and the purpose of generating the subgroup parameters online can be achieved; According to the selection of data and fitting error criterion, the optimal subgroup parameters can be adaptively selected; the subgroup parameters are calculated according to the temperature of the specific problem, and the calculation time will not increase due to the increase of the number of states.