207 results about "Combination algorithm" patented technology

Data management systems and methods that can catalog, store, and protect data from hard disk failure on a mobile or stationary computer, using a computer program and a portable data storage device are provided. The program uses a combination message-digest and check-function based algorithm that can back up, manage, and protect data independent of specific computer operating system or file system. The program can check files for block-level content changes from previously stored data blocks and catalogs and stores only new data blocks. By storing only new data blocks and cataloging information within a comprehensive, backup archive composed of these blocks, the size of each backup is reduced.

A global policy store, in which policies applicable to multiple applications in an enterprise environment can be stored, can be stored in association with that environment. An application-level policy combining algorithm can be associated with a specific application to resolve conflicts between the results of evaluating policies that pertain to that application's resources. A persistent model is defined for an Extensible Access Control Markup Language (XACML) target definition.

The invention discloses authentication system, including a user's terminal body for applying user's label certification by user's label information; a subsystem key generation basis to generate a user key generation basis for user key accounting including a public key generation basis and a private key generation basis; a certification management subsystem for responding the application of the user's terminal body to generate the public key and private key of the user with the user's label, the user key generation basis and the user's label combination accounting, and to sign the user's label certification including the public key and private key of the user and send them the certification to the applied user's terminal body. In addition, the invention discloses a corresponding method of releasing user's label certification. The invention takes advantages of low cost without a third party certifying, achieving key producing with scale and low requirement to the system resource and scale.

A method of monitoring brain activity is provided, wherein the method includes receiving a signal associated with neuronal activity of a mammalian brain. The method also includes processing the signal using a linear, non-linear, or combination algorithm to extract a signal feature. A neuromarker may be determined based on an association between the signal feature and a library of features, wherein the library includes a plurality of signal features correlated with a plurality of disease states.

Techniques for highly parallel evaluation of XACML policies are described herein. In one embodiment, attributes are extracted from a request for accessing a resource including at least one of a user attribute and an environment attribute. Multiple individual searches are concurrently performed, one for each of the extracted attributes, in a policy store having stored therein rules and policies written in XACML, where the rules and policies are optimally stored using a bit vector algorithm. The individual search results associated with the attributes are then combined to generate a single final result using a predetermined policy combination algorithm. It is then determined whether the client is eligible to access the requested resource of the datacenter based on the single final result, including performing a layer-7 access control process, where the network element operates as an application service gateway to the datacenter. Other methods and apparatuses are also described.

The invention provides a Maxout multi-convolution neural network fusion face recognition method. The technical problem of unequal classification of positive and negative samples in a single network is solved. The method comprises the steps that a regression tree combination algorithm and affine transformation processing training data are used to acquire a database after alignment; face images in the aligned database are extracted to acquire a database simply with face area images; two networks are built, and a Maxout method is used to acquire a fused network; the database simply with face area images is used to train the fused convolution nerve network to acquire a trained network model; and a test image is pre-processed to test the trained network model. According to the invention, a Maxout module is used to fuse two or more subsystems, and then is successively connected with a full connection module and an output module to acquire a complete face recognition system; the classification sensitivity of positive and negative samples of the convolution neural network can be balanced; and the face recognition accuracy is improved.

The invention discloses a line customization method and a line customization system based on travel planning. The method comprises the following steps: grabbing POI data in a WEB server, and finding out the key words in the POI data; building a tag library according to a set dimension, and establishing the corresponding relationship between the keywords and user preferences; getting a travel library according to the travel of a user, and decomposing the travel library into travel fragments, scoring POI attribute tags contained in the travel fragments to get the score of the user preference, and getting the routes of the travel fragments according to the score of the user preference; and combining the routes in the travel fragments through a POI combination algorithm to get a customized line. According to the invention, the fragments are combined into a finished recommended line based on POI attributes and tags, and thus, a travel fitting the personalized travel preference of a user can be recommended to the user.

The invention provides a GNSS / MINS (global navigation satellite system / micro-electro-mechanical systems inertial navigation system) super-deep combination navigation method, aiming to realize super-deep combination of a GNSS and an MINS. The GNSS uses a direct location estimation method to estimate location errors, speed error and clock errors according to location and speed auxiliary information provided by an INS (inertial navigation system) to close loops; a system takes an MIMU (MINS inertial measurement unit) and a base band correlator as sensors to realize all the navigation functions in a top combinational algorithm, in other words, the algorithm takes the GNSS base band correlator as the sensor of sensitive space-time locating fields and takes the MIMU as the sensor of sensitive inertance fields to realize integral combination of the GNSS and the MINS; and by the aid of the INS, the GNSS combines multichannel information to perform vector phase discrimination and vector locating. The invention further provides super-deep combination system and device applied to the method. The method, the system and the device have the advantages of high navigation accuracy, good dynamic property, high GNSS tracking sensitivity, high GNSS anti-interference performance, wide GNSS dynamic pulling range and the like; and in theory, the dynamic range is limited by trends of the MINS, and -160dBM (decibels above one milliwatt in 600 ohms) signal tracking can be realized by the GNSS by the aid of the MINS.

A method for the processing of images in interventional radioscopy to produce a sequence of de-noised images in which elongated objects of guiding instrumentation equipment are preserved and enhanced. The method of the invention uses a tool for indicating the confidence to be placed in a given pixel as regards its belonging to a piece of guidance equipment. This tool is used to create a combination algorithm capable of re-injecting the pixels having a low contrast-to-noise ratio into the de-noised image. This combination algorithm enables the enhancement solely of the pixels of the guidance equipment with a high confidence value so as not to produce artifacts with the enhancement of false positives.

The present invention discloses a voting based classification method for cross-language subjective and objectives sentiments. The method comprises the following steps of: S1, constructing a target language sentiment dictionary according to a source language sentiment dictionary; S2, separately adopting three algorithms of a rule algorithm, a combination algorithm of machine translation and statistical machine learning, and a polarity eigenvalue calculation algorithm to extract words from a sentence of a to-be-tagged text, determining sentiment polarity of the words according to the constructed target language sentiment dictionary, and further determining subjective and objective nature of the sentence; and S3, acquiring subjective and objective nature determination results of the sentence, obtained according to the three algorithms, and determining the subjective and objective nature of the sentence by voting. The method can fully consider contexts and usage habits of a target language on the premise of keeping certain accuracy, effectively solves the problem that a tagged corpus of the target language is scarce, and further improves classification accuracy on the premise of ensuring the recall rate.

An automatic generation method for a software test case comprises the following steps: establishing multiple predetermined case generation specification modules and combinational algorithms; inputting at least one test parameter value according to test demands to create an extensible markup language (XML) file; reading in the XML file and selecting test parameters and corresponding combinational algorithms; establishing multiple valid parameter test cases and multiple invalid parameter test cases with the multiple predetermined case generation specification modules according to the test parameters which are read in; synthesizing the multiple parameter test cases into at least one actual test case according to the selected combinational algorithms; and outputting the actual test case as a receipt file. Therefore, testing personnel can finish output of a test case receipt automatically after creating the XML file according to the demands, workloads of the testing personnel are greatly reduced, and the universality of a test platform and the reliability of test results are improved.

The invention discloses a vehicle-mounted strapdown / satellite tight-combination seamless navigation method. The method comprises the following steps that a system is initialized; messages are output by a GPS, inertial measurement unit data are output by a SINS, and synchronization is carried out; the GPS messages are analyzed to obtain rough pseudorange, Doppler shift and satellite position parameters, and compensation for the parameters is carried out to obtain fine pseudorange, pseudorange rate and satellite position parameters; navigation calculation is carried out on the inertial measurement unit data to obtain vehicle state, speed and position parameters; if the number of visible satellites is larger than or equal to 4 is judged, if the number is larger than or equal to 4, 4 satellites are selected to carry out tight combination filter calculation, and is the number is smaller than 4, an intelligent combination algorithm is used; vehicle navigation parameters are output, and if navigation needs to be continued, the step 2 is redone. Tight combination navigation and the intelligent combination algorithm are used for realizing seamless switching of different visible satellites, and vehicle seamless navigation is reliably guaranteed.

A global policy store, in which policies applicable to multiple applications in an enterprise environment can be stored, can be stored in association with that environment. An application-level policy combining algorithm can be associated with a specific application to resolve conflicts between the results of evaluating policies that pertain to that application's resources. A persistent model is defined for an Extensible Access Control Markup Language (XACML) target definition.

The invention relates to the technical field of computer networks, in particular to a self-adaptive load balance scheduling mechanism for an internet-of-things device access processing platform. The invention provides a novel scheduling algorithm framework in which a routing policy and a routing algorithm are independently separated, the routing policy can be replaced, the evaluation indicator is expandable, the routing algorithm is open-ended, and three routing policies including a single algorithm policy, a combined algorithm policy and an adaptive algorithm policy, are also provided to deal with the abruptness and the randomness of a device connection request in a real application environment. With the adoption of the self-adaptive load balance scheduling mechanism, an internet-of-things device access system effectively adjusts the distribution condition of the device connection request according to the performance difference and the load conditions of various nodes in a cluster, so that the overall performance of a cluster system is improved. Compared with the prior art, the self-adaptive load balance scheduling mechanism for the internet-of-things device access processing platform has the advantages of novel thought, flexibility in use, and good convenience for expansion, fully meets the scheduling requirements of the connection request of an internet-of-things gateway, and effectively breaks the limitations that the single router is not applicable to all scenes.

According to the invention, a probability integral method parameter prediction model for optimizing the BP neural network based on a combined algorithm (GP) of a genetic algorithm and a particle swarmoptimization algorithm, and an input layer of the BP neural network is optimized by adopting an average influence value algorithm (MIV), so that the complexity of the network is reduced, and the purpose of improving the prediction precision is achieved. An MIV-GP-BP model is established by taking actual measurement data of 50 working surfaces as a training set and a test set of the BP neural network, and the precision and reliability of a model prediction result are analyzed; the results show that: in five parameters, the root mean square error ranges from 0.0058 to 1.1575; the maximum relative error of q, tan beta, b and theta is not more than 5.42%, the average relative median error is less than 2.81%, the s / H relative error is not more than 9.66%, the average relative median error is less than 4.31% (the parameter itself is small), and the optimized neural network model has higher prediction precision and stability.

A marketing system comprising a receiving means for receiving a request for researcher characteristics from a database comprising an array of researcher characteristics, a filtering means for filtering the array of researcher characteristics in order of relevance to the requested characteristics, and delivery means for delivering the relevant filtered researcher characteristics derived from the filtered array. Marketers search a database containing information pertaining to scientific researchers, including recent grant applications, recent publications, history of attending research-oriented events, purchase history of the laboratory, research focus of the laboratory, techniques used under the research focus, geography, contact information for members of the laboratory, services provided by the laboratory, world wide web address associated with the laboratory, individual and group research survey information, equipment, consumables and software accessible to the individual or group, URL, and any combination thereof. An algorithm ranks matches by relevance to the marketer-specified search criteria, and specific information pertaining to database matches is filtered and delivered to the marketer. Targeted marketing campaigns can be directed toward potential customers based on retrieved information.

The invention discloses a method for applying a combined algorithm to a combined learning structure digital predistortion system. Through combination of a classic crest factor restriction (CFR) technology and a digital predistoriton (DPD) technology, a combined learning structure design digital predistoriton system combining an indirect learning structure and a direct learning structure is provided according to advantages and disadvantages of the direction learning structure and the indirect learning structure; then according to advantages of an LMS algorithm, an RLS algorithm and a variable step size LMS, namely, NLMS algorithm, a combined algorithm combining RLS and NLMS is provided for estimating and updating parameters of a digital predistorter. The combined algorithm has the characteristic of rapid convergence of the RLS algorithm and also simplifies computation amount. CFR and DPD are sequentially performed on an input peak average ratio signal, then the processed signal is sent to a power amplifier, so that ideal linear output is obtained, and in-band distortion and out-band spectrum expansion caused by power amplifier non-linear are effectively relieved. Through adoption of the method, performance degradation of a broadband power amplifier caused by memory effect is relieved, furthermore, the linearity and efficiency of the broadband power amplifier are improved.

This invention relates to an intellectual map-matched automobile navigation method based on traffic information, which comprises continual receiving satellite positioning information by a vehicle-carried GPS receiver, receiving vehicle angle variation and mileage data by vehicle-carried sensor; according to GPS quality index and mileage difference, setting range taking reliability analysis with received data; matching map with E-map to the output positioning information, taking feedback check to received information, and outputting vehicle position. This invention obtains optimal estimation for the vehicle position in road network by combined optimized algorithm, corrects the error from DR system to keep system continual work for a long time without GPS signal, and applies new target-oriented map-matching method to improve match reliability and auto error correction function.

The invention discloses a code coverage rate computing analysis method, a code coverage rate computing analysis device, equipment and a storage medium, and belongs to the technical field of computers.The method comprises the following steps of acquiring a code of a current version and commit id corresponding to the code from a code library via a git command; packaging and compiling the code, andstoring the code into a folder named by the commit id corresponding to the code; receiving a coverage rate file named by the commit id of the code, and storing the file into the folder named by the commit id; generating a test coverage rate port of the current code via a code coverage rate tool; generating a full coverage rate report via a combination algorithm; and generating an incremental coverage rate report via a git diff command. According to the method, device equipment and medium provided by the invention, combination of coverage rates between the different codes after the code is frequently changed is supported, and thus the certain version of accurate coverage rate file can be acquired.

The invention provides a ship water gauge image automatic number reading method based on a deep learning algorithm. According to the method, the deep learning algorithm is used for performing identification positioning on the number in a ship water gauge image, thereby obtaining a digital number and a frame coordinate in the image, and a digital frame coordinate is corrected. By means of a numberarrangement rule and a human eye number reading principle in the image, a digit combination algorithm based on the human eye number reading principle is designed. By means of image color information of a ship body and water body, a method for identifying a water level in different color body color backgrounds is designed, wherein different identification methods can be selected according to different backgrounds, thereby ensuring high accuracy and high speed in water level identification. According to the position coordinate of the oblique number in each line and the position coordinate of thewater level, a length reference and oblique number correction algorithm is designed; wherein the actual pixel distance between water level and the number is calculated, and percentage of the pixel distance between the water level and the number in the actual length reference is calculated. Furthermore the number of the ship water gauge image is obtained.

The encoding method for LDPC code based on LU decomposition for optimized search matrix is realized by software pretreatment and hardware encoding. Wherein, eliminating correlative rows of check matrix, realigning the matrix; selecting nonsingular matrix with non-correlative column composite algorism for decomposition with simple LU algorism; optimizing search by diagonal element asymptotic one of nonsingular matrix; finally, selecting nonsingular matrix with lowest density, and reducing delay of whole coding system by time interval control parallel process hardware structure. This invention holds advantages in prior art, reduces calculation amount near Greedy algorism, and fit to any LDPC code special to abnormal code and normal node with correlative rows.

The invention discloses an online compression method and device for mass ship AIS track data. The online compression method is a threshold combination algorithm for limiting double factors (synchronous Euclidean distance and steering angle). The method comprises the following steps: taking a water area as a research object, firstly finding historical characteristics of a ship track in a certain water area according to historical AIS data, then carrying out online compression based on a mode of combining change of a motion state with historical characteristic data, and finally constructing shipAIS track data based on kernel density estimation. According to the method, the AIS data can be efficiently compressed, the track feature points of the AIS data can be completely and accurately reserved, efficient management of the AIS data is facilitated, and the information communication efficiency and supervision efficiency between ships are improved.

The invention provides a simple and reliable inertia / GPS (global positioning system) / milemeter combined algorithm for solving the precision positioning problem of vehicles in complex environments, such as bridges, trees, tunnel barriers, urban canyons and the like. The method comprises the following steps of: respectively independently calculating position information of a carrier through an inertia / GPS combined navigation algorithm and a milemeter dead reckoning algorithm, further fusing the position information obtained by calculation through the two different algorithms through an output correction filter with simple design, and finally giving out high-precision vehicle position information. Compared with a traditional single-filter deep combined algorithm, the method has the advantages of stronger environmental adaptability, no sensitivity to error models of GPS, inertial navigation and a milemeter, high reliability, high output precision, easiness in implementation and the like, and can well solve the precision positioning problem of the vehicles in the complex environments.

The invention discloses a gesture recognition method based on a surface electromyography signal, which utilizes a sensor to collect the surface electromyography signal data. Seven feature extraction combinations were used to extract features from surface electromyography data. Three different classifiers are used to classify and compare the data after feature extraction. The average accuracy of the classification results of each group of surface electromyography data is calculated, and the combination of the feature extraction value algorithm corresponding to the maximum value in the average accuracy of the classification results and the pattern recognition classifier is taken as the optimal combination to determine what kind of gesture the surface electromyography signal belongs to. The embodiment of the invention adopts seven feature extraction combination algorithms and three classifiers for classification, thereby effectively improving the rapidity and stability of gesture recognition and bringing better user experience when applied to dexterous hand control.

The invention relates to a method for monitoring and calculating a subsynchronous oscillation mode from an electric generating set in real-time when a power house and an electric power system generate subsynchronous oscillation (SSO) and discloses a method for obtaining a shafting torsional oscillation mode of a steamer electric generating set through combining a main filter, a starting filter and a narrow band wave-resistance filter. The method works out the change of a rotary palstance (dw) of an electric generating shafting through a rotary speed pulse of a rotary speed sensor arranged on the electric generating set shafting and works out an instant value of each mode of the electric generating set shafting torsional oscillation with linear phaseshift through the combination algorithm of a broad band filter combined by the main filter and the narrow band wave-resistance filter; a dw signal is combined with the narrow band wave-resistance filter by passing through the starting filter to be a mode starting condition. The method can be used for restraining the subsynchronous oscillation (SSO) steamer electric generating set into a subsynchronous oscillation damp controller (SSODC).

The invention discloses a dual-model fault diagnosis method based on dynamic weighing. The method comprises the steps that sensor vibration signals and fault recording texts collected by a motor driveend under the normal state and various fault states are selected; then,the sensor vibration signals and the fault recording texts are learned,then,a dynamic weighing combination algorithm is adoptedfor giving a weight to a model,sub-model SVM multi-classification voting results are combined,and a final classification result is obtained. Joint diagnosis for bearing fault data and bearing fault texts can be achieved. Non-balanced processing and valuable information extraction and classification are performed on equipment operation data,manual recording experience knowledges are effectively combined for text data mining,compared with a single diagnosis model,the method can remarkably improve the fault diagnosis precision,a better performance evaluation index is obtained,and the good theoretic and application value is obtained.

A base station device, a method thereof and a communication system thereof are disclosed. The base station comprises: a covariance matrix obtaining unit that forms a covariance matrix of an interference vector and a noise vector from received signals from multiple user equipments; a ratio calculation and comparison unit that calculates a relative ratio between the interference and the noise in the received signals based on the covariance matrix and compares the ratio with a predetermined threshold; an adjusting unit that reduces values of non-diagonal elements in the covariance matrix when the ratio is less than the predetermined threshold; an equalizer unit that performs equalization based on the covariance matrix by utilizing an Interference Rejection Combining algorithm when the ratio is greater than or equal to the predetermined threshold, and equalizes the received signals based on the adjusted covariance matrix by utilizing a Minimum Mean Square Error algorithm when the ratio is less than the predetermined threshold. By using the uniform receiver architecture, the advantages of the two algorithms are achieved in a simple structure, and further, its performance is superior to a MMSE receiver and an IRC receiver.

A diversity-receiving terminal of digital TV single-frequency network consists of multiple directional antenna enabling to receive TV broadcast band signal, multiple receiving-component enabling to independently receive TV broadcast signal and enabling to generate different diversity indication signal according to different diversity selection principle, a diversity selection component for carrying out diversity selection according to different diversity combination algorithm and a video decoding component.

The invention provides a method and a system for detecting pedestrians in video streams. The method comprises the steps of obtaining multiple paths of video streams; performing moving target detection on each path of video stream; segmenting out a moving target region of each picture; splicing all pictures into a single large picture through a divide-and-conquer-based combination algorithm; obtaining a spliced image; and performing neural network prediction on the spliced image through a convolutional neural network to obtain a detection result. In the whole process, a moving target detection algorithm and a convolutional neural network pedestrian detection algorithm are efficiently combined together through the divide-and-conquer region combination algorithm, so that the pedestrians in the video streams can be efficiently and accurately detected.

The invention discloses a high resolution remote sensing image analysis tree automatic extraction method based on space consistency and relates to the remote sensing image processing technology. The method includes the following steps: S1) building various complex target image representational sets; S2) adopting an over-segmentation algorithm to obtain different image objects; S3) conducting statistics on multi-dimensional characteristics of the objects such as color, grain, shape and the like; S4) acquiring class attribute of the objects combined with an implied probability semantic model; S5) selecting consistent object based on space consistency; and S6) calculating a similar matrix of the consistent objects and finally generating an analysis tree structure of the image scene through a split combination algorithm. The method highlights important target information in the scene by analyzing space geological arrangement, acquires an analysis tree of the scene, is favorable for improving semantic expressing accuracy and improves target recognition and scene classification robustness.