212 results about "Screening algorithm" patented technology

A one-to-many identification system for access control allows a search rate of up to ˜1:30,000 in a real time. The system uses a very fast pattern based screening algorithm followed by a fast minutiae based screening algorithm. A fused score of both algorithms is used as a decision metric to screen out a vast majority of all the templates after the second stage. The remaining templates are sent to a full minutiae based algorithm to obtain a minutiae comparison score. If the result is still inconclusive after the third stage, a full pattern based algorithm is run, and its score is fused with the minutiae comparison score. The system also uses an adaptive classification technique which minimizes a distance between each template and a number of templates. The system can be realised as a standalone unit or on a server.

The invention discloses an electric system low-frequency oscillation detection method. On the basis of the existing nonlinear and non-stable signal empirical mode decomposition algorithm, the improved empirical mode screening algorithm enables the electric system low-frequency oscillation detection method to adapt to low-frequency oscillation signal processing, decomposed eigen mode components are combined with low-frequency oscillation mode physical expression meaning, and accordingly all low-frequency oscillation mode parameters of low-frequency oscillation sampling signals are recognized. The electric system low-frequency oscillation detection method reduces influence of the endpoint effect through improved mixed endpoint forecast, removes overshoot and undershoot by adopting triple B spline interpolation, simultaneously effectively reduces mode aliasing degree, analyzes and detects aliasing mode parameters through refined complex wavelets, controls algorithm screening depth through the improved energy elementary error screening stopping condition, reduces damping loss of mode components, and accordingly effectively improves low-frequency oscillation mode characteristic parameter detection accuracy.

The invention relates to a system and method for acquiring the monitoring and warning information of gas disasters of a coal mine, and belongs to the technical field of monitoring and warning of gas disasters of a coal mine. The system is realized by and the method comprises the following steps: firstly, automatically obtaining and transmitting the basic safety information of a mine by utilizing an underground transmission technology, an underground looped network and a local area network on the ground and through a data interface program of a mine monitoring and controlling system of research and development series, a data transmission module of detecting and probing equipment and an information acquisition terminal based on an anti-explosion computer and an anti-explosion mobile phone; filtering, correcting, ordering and classifying the obtained basic safety information of the mine according to an information screening algorithm for research so as to form the monitoring and warning information of gas disasters of the coal mine; finally, formulating retrieval logic of monitoring and warning information and information transmission rules of monitoring and warning information through an developed acquisition and management platform of the monitoring and warning information; automatically extracting corresponding data from an acquisition database of the monitoring and warning information; and transmitting the extracted data to a warning system of gas disasters of the coal mine so as to realize the automatic real-time acquisition of the monitoring and warning information.

The invention discloses a data processing method based on a knowledge map, which comprises the following steps: constructing a knowledge map based on the sorting of examination questions and teachingand research knowledge, school teaching, teaching materials and the difficulty degree of examination questions to obtain a knowledge base; setting search conditions and extracting test questions according to the search conditions, wherein, when the search conditions contain natural language, the search conditions are converted into structured search instructions including knowledge points and search ranges, and the search instructions are matched with the contents in the knowledge map to search out the test questions; carrying out similarity algorithm, sorting algorithm and comprehensive screening algorithm on the searched test questions to obtain a screening result; pushing the filter results to the user. The invention simplifies the retrieval process of the test questions and improves the efficiency of searching the test questions by establishing the knowledge map of the test questions and searching based on the knowledge points when searching.

The invention discloses a load balancing system for remote service calling based on a real-time load rate. The load balancing system comprises a monitoring center, a registration center, a client and a server; the monitoring center is used for status monitoring and counting the number of service connections, and regularly reporting the number of service connections to the registration center; the registration center is used for providing service release and service subscription, pushing a list of available service instances to the client, acquiring the number of service loads in real time, and calculating the load level of the service instances; and the client comprises a load balancer module, wherein the load balancer module is used for analyzing the load level of the service instances, screening the service instances according to a load balancing screening algorithm, and recommending the service instances to the client.

The invention relates to the technical field of telematics, and provides a telematics network node screening and accessibility routing construction method based on a self-encoding network. The entire routing mechanism comprises data preprocessing, self-encoding network training and node screening algorithm realization. The self-encoding network can provide a high-dimensional continuous data space and bidirectional mapping of a low-dimensional nested structure. The self-encoding network can be used to excavate and extract the related characteristics of node information in a large-scale road transport network. Dimension lowering is carried out on original information. The amount of data in a data transmission process is reduced. An effective way is provided for screening key nodes in the network. According to the invention, a great node screening method is provided for a highway scene, and the network performance is effectively improved.

A method is provided for rendering a color image with a plurality of separations with a multi-level successive-filling halftoning process using a single screen for a plurality of separations. For a separation to be processed, first, from the possible multiple levels for the separation, the set of levels that would be used and the number of dots corresponding to the different levels are decided. These decisions are based on the input level for the separation and input levels for the prior process separations. The location of the dots to be printed for the different levels for the separation is then decided by using the halftone screen, while simultaneously taking into account the placement of printed dots for the prior separations. The selection is done so as to best disperse dots and minimize overlap.

The invention provides a method, a device and equipment for determining a reality scene key frame, and media. The method comprises the following steps that: carrying out feature matching on the feature points of a current video frame and the feature points of a reference video frame to obtain matched feature point pairs, wherein the reference video frame is a video frame in a reality scene, and the collection time of the reference video frame is different from the collection time of the current video frame; according to a preset screening algorithm, screening the feature point pairs of which the similarity exceeds a first threshold value in the matched feature points; and according to the number of the screened feature point pairs, determining the current video frame as the key frame of the reality scene, and storing the key frame into a scene map corresponding to the reality scene, wherein the scene map is only used for storing the key frame of the reality scene. By use of the method, the key frame of the reality scene is determined without the restriction of a specific scene so as to construct the accurate and reliable reality scene.

The invention discloses an intelligent screening algorithm realization method of a shared parking stall system, and can solve the problems in the prior art that the parking stall resource utilizationrate is low; the method comprises the following steps: classifying car models in the market according to size specifications, wherein each type vehicle is matched with a parking stall of a specific size; reading a car model so as to classify the vehicle, and searching parking stalls and belonging parking lots meeting the requirements according to the classification result; intersecting sections soas to obtain an overlapping time between a vehicle owner parking time and a parking stall idle time, returning to the corresponding parking stall and the belonging parking lot, intersecting the result with the result obtained by the classification operation so as to obtain the final target parking stalls, and ranking the parking stalls according to user requirements. The method uses the staggeredtime and zone algorithm to solve the difficult parking problems, thus alleviating the road traffic pressure; by using said method, the parking can be more convenient and intelligent.

The invention relates to a lane line detection method based on Hough transform. The method comprises the following steps that an in-car camera collects road image data in real time; a central line of a lane line is detected according to the collected road image data; a straight line where the lane line is placed is detected by utilizing a straight line detection algorithm of Hough transform; and straight lines detected via Hough transform are screened to obtain the straight lane lines. Compared with edge detection, central line detection in the invention has the advantages that half of original object points can be omitted, the possibility that multiple straight lines are found for one lane line is reduced, and the complexity of the subsequent screening algorithm is reduced.

The invention discloses a high-resolution mass spectrum accurate molecular formula forecasting method. According to the method, candidate molecular formulas are graded and ranked by mainly adding a condition of a probability that the molecular formula exists in practice so as to improve the rank of the accurate molecular formula. Through actual tests, no matter the speed or accuracy of the method is better than that of a common method. In addition, due to the combination of the method and a common candidate molecular formula screening algorithm, the universality and accuracy of the forecast of the molecular formula can be greatly improved; through the method, accurate molecular formula forecasting can be conveniently carried out on high-resolution mass spectrum data; extremely valuable reference results are provided for operations of analyzing compound structure, and screening known substances or unknown substances.

The invention discloses a fault diagnosis method for bearing equipment, and relates to the technical field of fault diagnosis of bearing equipment. The method comprises the following steps: firstly, unbalanced data samples are processed by using a resampling method; and then feature mode extraction is carried out on newly-added data by using a deep learning algorithm, then the newly-added modes are input into an original integrated model for classification, a dynamic forgetting weight algorithm, an effective increment screening algorithm and an increment information fusion algorithm are further provided, and the weight of an increment merging mode is dynamically adjusted according to a calculated value. Finally, supervised classification is carried out on the weighted modes by using an incremental integrated SVM, so that real-time extraction of bearing equipment state features and reliable classification of fault modes are realized.

The invention discloses a method for detecting microorganisms in a sample, which belongs to the technical field of metagenome analysis and comprises the following steps: S1, acquiring metagenome sequencing data of the sample; and S2, performing species analysis on the metagenome sequencing data. The method further comprises a step of carrying out drug resistance gene analysis on the metagene sequencing data of the sample. The invention also discloses a system for detecting microorganisms and drug-resistant genes in a sample. On the basis of a comprehensive and accurate database and an intelligent analysis and screening algorithm, pathogenic microorganisms and drug-resistant genes are effectively identified by utilizing a metagenome sequencing method, false positive is effectively reduced,bacteria suspected to correspond to the detected drug-resistant genes can be prompted, and better technical support is provided for accurate diagnosis and treatment of infection; in addition, according to the method and the system, potential new pathogenic microorganisms can be accurately analyzed, and technical support is provided for early warning of new infectious diseases.

The invention relates to a load balancing method based on fog node cooperation in a fog computing network, and belongs to the technical field of mobile communication. The method comprises the following steps: firstly, carrying out related modeling on a communication process and a calculation process of a fog node cooperation scene; designing a cooperative contribution degree for screening cooperative fog nodes and a cooperative fog node screening algorithm based on multi-attribute decision; designing a cooperation contribution ratio coefficient xm based on cooperation contribution degree, andconstructing a system total overhead optimization problem model introducing the cooperation contribution ratio coefficient; finally, designing a load balancing algorithm based on the cooperation contribution degree for solving the optimal distribution result of the task loads among the fog nodes. According to the invention, on the premise of ensuring the feasibility, fairness and stability of fognode cooperation, the service load is reasonably shunted, and the system performance and QoE of users are improved.

Apparatus and methods for attacking a screening algorithm. Content to be downloaded is identified. The identified content is then partitioned into at least two sections, wherein each of the at least two sections has a duration which is less than a duration of a threshold duration value assigned by the screening algorithm, and the partitioned content is subjected to a screening algorithm. Once the content has been successfully downloaded, the integrity of the downloaded content may be restored by reassembling the sections.

The invention discloses a method for optimizing steam distribution curves of steam turbines on the basis of data mining technologies. The method includes screening data by the aid of data screening algorithms according to large quantities of historical data accumulated in DCS (distributed control systems) and correcting main steam flow under different working conditions according to main steam flow correction formulas to obtain main steam flow under rated working conditions; carrying out clustering analysis on the corrected main steam flow, unit load, main valve location instructions and valve location values of various valves by the aid of data mining tools to obtain main steam flow, main valve location instructions and valve location values of the various valves and correcting steam distribution curves of the valves according to relations among original main valve location instructions and four adjusting valves to obtain novel steam distribution curves of the valves. The unit load is used as a mass center of the main steam flow, the main valve location instructions and the valve location values which are obtained by means of clustering analysis. The method has the advantages that the main valve location instructions and the main steam flow of the steam turbines can linearly change owing to the novel steam distribution curves of the valves, and accordingly the problems of fluctuation of existing unit load, poor quality of primary frequency regulation, load jumping, long response time of units and the like can be solved.

The invention discloses a brain section coupling analysis method based on synchronous screening. The method comprises first synchronously acquiring EEG signals in 32 channels and EMG signals in 12 channels when different gripping power is output; in order to research a coupling relation of an EEG motor area and a sensory area in a gripping power output process, analyzing the EEG signals and brachioradialis signals of multiple channels, such as leads C3, C4, CP5, and CP6, in the motor area and the sensory area of the brain; then extracting the synchronous information of the EEG and the EMG by using a synchronous screening algorithm to obtain SSEM (Synchronous Screening of EEG Signals Based on EMG Signals); and finally calculating the symbol transfer entropy of the SSEM to determine its coupling relation. The method can remove data unrelated to motion and reduce a data size.

The invention discloses a remote fault diagnosis method of a generator excitation system. The method comprises the steps that local abnormities of the generator excitation system are associated and screened according to the data screening algorithm, and effective data are compressed; abnormity and fault information of the generator excitation system is sent to a remote server; the remote server establishes an expert knowledge base to compare the abnormity and fault features of the generator excitation system and locate faults; early warning is carried out on accidents. The method has the advantages that input / output signals of a sensor and an analog quantity of the local generator excitation system are screened, and therefore extraction of effective signals is simplified; data are transmitted to the remote server in a one-way mode, and therefore the remote side is prevented from interfering with or controlling the local excitation system; the server compares the data features through the expert knowledge base so as to rapidly analyze the abnormity and fault information, find out fault causes and predicate the accidents.

The invention provides a system for detecting potential pathogenic variants of exome based on a family. The system comprises a comment integration module for reading a result file of exome sequencing data processing flow, a matrix segmentation module for extracting variants in all samples t, solving a union set and then combining all the samples to form a matrix, a potential pathogenic variant mining module for extracting variant information from the obtained matrix for being enumerated, evaluating pathogenicity of individual variants and combined double-site variants to obtain a list of potential pathogenic variants, and a potential pathogenic variant recording module for recording occurrence of a site in each sample and a target gene and displaying with icons. Output results of conventional exome sequencing low processing are used as input conditions to complete the integration and basic filtration of data; a candidate set of potential pathogenic variants is obtained according to a special variant screening algorithm, so that the system provided by the invention solves the problem of mining of potential pathogenic variants of sequencing data with high heterogeneity, high mutation rate and high noise.

The invention provides an apple soluble solid near-infrared detection part compensation method and system. The method includes the steps an absorbancy spectrum of the preset portion of an apple is obtained; the soluble solid content of the preset portion of the apple is obtained; according to the absorbancy spectrum and the soluble solid content, a characteristic wavelength screening algorithm is applied to select characteristic wavelengths related to the apple soluble solid content in a full-wave band spectral range, and an apple soluble solid near-infrared detection model is set up. Spectral information of multiple portions of the apple is fused, the characteristic wavelength screening algorithm is combined, the apple soluble solid near-infrared detection model is set up, the influence on the apple soluble solid near-infrared detection model by changes of spectrum detection is overcome, the stability of the model is improved, and a theoretical basis is laid for portable detection and online detection of apple soluble solid in practical production.

The present invention proposes a real estate valuation method based on a random forest, comprising: step 1, establishing a basic database step: collecting real estate data to form a basic database, and using a clustering algorithm or / and a deduplication algorithm or / and a screening algorithm to perform data processing on the basic database , carry out characteristic variable quantitative processing to real estate data; Step 2 establishes the step of random forest model, step 3 steps the step that random forest model is tested, step 4 real estate valuation step, the random forest algorithm of the present invention does not need pre-set function form, can Accurately fit samples of complex nonlinear relationships.

The invention discloses a configuration file collecting and monitoring method based on OGG database replication. File transmission from the production end to the disaster recovery end is achieved by deploying a disaster recovery cascading server end, wherein isupload_rz fields are added into a category classification table; if the disaster recovery end is matched, the collected information is directly input into the disaster recovery end, and if the production end is matched, the information is transmitted to the disaster recovery end through the RDT service according to the kpiid attribute configuration situation in con.properties; gibberish in the collected file is cleared away and effective configuration information is reserved based on the screening algorithm; compliance comparison is carried out according to the comparison algorithm and the established configuration rule; difference comparison between configuration information at the production end and the conuration information at the disaster recovery end is carried out, and the detecting result is shown. By means of the configuration file collecting and monitoring method, the configuration file information of the production end and the disaster recovery end is collected, the key configuration file information is automatically detected and analyzed through tools, monitoring on the conuration content of the OGG is achieved through the rule defining and analyzing, and the blind spots of original OGG replication monitoring are made up.

The invention discloses a converter valve voltage-sharing state consistency statistics method, which includes the following steps of: configuring a register array which is used for storing deviation data for each thyristor valve level of the converter valve in a converter valve control unit, and configuring a storage region for storing the deviation data with time mark information; configuring a high speed processor in the converter valve control unit as a key core device; designing a deviation data screening algorithm in the converter valve control unit used for screening the data with large deviation from state signals fed back from the each thyristor control unit of the converter valve, and classifying and storing the data; designing and providing an interface for reading the deviation data in the converter valve control unit, the interface being used for on-line monitoring the deviation data or reading the data and off-line analyzing the data; and finding the thyristor valve level with large deviation accumulation quantity from a data statistics result, analyzing a specific location in a thyristor valve voltage-sharing loop where the difference exists, thereby pre-judging the case that this type of thyristor valve level may have fault, and taking the pre-judgment as a reference basis of device replacement and optimization design.

The invention provides an asymmetric encryption method using quantum random numbers. The specific implementation steps of the method are as follows: step 1: a quantum random number generator generatesunpredictable random numbers; step 2: these quantum random numbers are subjected to a large integer prime number screening algorithm to obtain a seed random number satisfying the generation of a public key and private key pair; and step 3: the seed random number is subjected to a public key and private key pair generation algorithm to generate the public key and private key pair. By adoption of the method provided by the invention, the problem that the existing public key encryption system generates the random number by using the determined algorithm, resulting in extremely high insecurity, is solved. The unpredictable random number generated by the quantum random number generator is used as the seed to ensures the security and reliability of the generated public key and private key pair.Therefore, the asymmetric encryption technique using the quantum random numbers will further improve the security under the existing encryption technology.

The invention discloses a power grid equipment data flow cleaning method based on an association rule, which comprises the following steps of: calculating association strength of historical data of each data sequence in a data flow by utilizing an Apriorri algorithm, and outputting an association relationship among different data sequences; using an abnormal data screening algorithm based on a sliding time window to detect the data sequences with weak correlation strength one by one; carrying out abnormal data identification processing on the data sequence with higher correlation degree at thesame time; and applying the neural networks with multiple wavelet basis functions to data cleaning to complete combined prediction. The data flow cleaning method is accurate in power grid equipment risk assessment and stable and reliable in data.

An embodiment of the invention provides a remote sensing image time sequence change detection method. The method comprises the steps of performing cloud removal on a to-be-detected time sequence imageto obtain a cloudless time sequence curve of each pixel of a cloudless time sequence image at each waveband; obtaining a seasonal breakpoint position in each cloudless time sequence curve; and basedon a preset screening algorithm, detecting whether the change of each seasonal breakpoint position is true change or false change. According to the remote sensing image time sequence change detectionmethod and system provided by embodiments of the invention, by utilizing spatial information of seasonal earth surface change, the position of each seasonal breakpoint in a time sequence is accuratelydetermined, and whether the change of each seasonal breakpoint position is the true change or the false change is judged, so that effective detection of the time sequence is realized.

The invention discloses a college student idle bicycle leasing software screening system and a realization method thereof. The main functions of the method are collection, integration, classification and screening of leasable bicycle information. The method comprises the following steps: after a user registering software, an information obtaining interface is accessed, and the software pushes the leasable bicycle information updated in real time in the form of a list; and if the user wants to screen bicycle information in a specific scope, by touching a function key on a screen, he can screen models, leasing prices, bicycle brands and leasable time of coincidence and obtain bicycle information according with requirements. According to the invention, a screening algorithm is innovatively applied to a college student bicycle leasing system, the trouble of manual information screening is greatly reduced, and fast and intelligent information processing is realized.

The invention discloses a wheel recognition based induction method for side-azimuth towing by road rescue equipment. The method combines the structure of road rescue equipment and the characteristicsof side azimuth towing operation, and firstly determines the area to be detected of the wheels of the towed vehicle. Then Gaussian smoothing filtering and edge detection are performed on the image ofthe region, and the gradient directions of each image point are calculated, and the non-zero edge image point set is obtained after binarization. Then the image points are clustered according to the edge continuity feature. Then the gradient direction histogram is established and the related parameters are calculated for the edge point set obtained by clustering. Then a two-step screening algorithm of wheel edge point set is proposed. Therefore, towing guidance is performed according to the selected wheel center position. Finally, the Kalman filter is used to predict and set the area of the next frame of the wheel to be detected. The guidance method provided by the invention has good real-time performance, environmental adaptability and anti-interference ability, and effectively improves the rescue efficiency of the road rescue equipment.

In a coal mine underground mobile object positioning system, a reliable wireless communication algorithm with a good concurrent identification capability is designed, such that the reading leakage rate of the positioning system can be effectively reduced, and through control of a wireless bump probability, the system reliability is improved, thus the goals of improving the performance of the mine mobile object positioning system and providing a safety guarantee for coal mine production are realized. On the basis that several wireless communication technologies of the mine positioning system and influences on the reading leakage rate are analyzed, the invention discloses a "multi-concurrent leakage screening" algorithm and introduces a "time slice scheduling" algorithm on the multi-concurrent leakage screening algorithm, such that the algorithm effectively reduces the reading leakage rate on the basis that bidirectional communication between a card reader and a label is realized, thus two purposes are achieved through one move.

The invention discloses a lane target detection method, apparatus and device, and a computer readable storage medium. The method comprises the steps of obtaining a to-be-detected image collected by animage collection device on a vehicle; preprocessing the to-be-detected image to obtain a processed image corresponding to a region of interest in the to-be-detected image; tracking, detecting and processing lane lines in the image by using a threshold screening algorithm and an instantiation scoring screening algorithm. according to the method, the to-be-detected image is preprocessed, the processed image corresponding to the region of interest in the to-be-detected image is obtained, and interference factors in the to-be-detected image are removed, so that the operand is reduced; through theuse of a threshold screening algorithm and an instantiated scoring screening algorithm, the detection efficiency is improved on the basis of ensuring higher precision and robustness of lane line detection, so that real-time and accurate detection on a vehicle gauge level chip can be achieved; and the use of a millimeter-wave radar and a laser radar is avoided, and the vehicle cost is reduced.