92 results about "Fuzzy data" patented technology

The invention discloses an alarm monitoring strategy analysis method based on a data mining technology and belongs to the field of communication alarm monitoring. Through data mining, valuable knowledge can be extracted from enormous fuzzy data comprising noise, alarm data of a communication network is enormous and dynamic and is rich in knowledge, the alarm data of the communication network is mined without reliance on a network structure, and dynamic change of the network can be adapted. Data mining comprises the following steps: obtaining and preprocessing the alarm data; converting the obtained alarm data into a set of transaction items, and carrying out frequent item set data mining by use of a correlation algorithm; and through combination with actual application, performing compression, merging and classification processing on a result of a frequent item set. According to the invention, through merging and converting alarms, the multiple alarms are merged into one alarm with a larger information amount so as to replace the multiple alarms, and assistance is provided for network management personnel in analyzing fault information and rapidly positioning faults.

The invention relates to a big data technology, and discloses a second-level caching method, which comprises the following steps: obtaining an original data set obtained from a webpage, and dividing the original data set into high-frequency data, common data and fuzzy data; creating a first-level cache in a local process, and storing the high-frequency data into the first-level cache; building a common second-level cache outside the local process, and storing the common data into the common second-level cache; constructing a fuzzy secondary cache outside a local process, and storing the fuzzydata into the fuzzy secondary cache; and receiving a query command input by a client, performing data query on the query command according to the first-level cache, the common second-level cache and the fuzzy second-level cache, and returning a query result to the client. The invention further provides a second-level caching device and a computer readable storage medium. According to the second-level caching method, efficient storage and query of data are realized.

The invention discloses a product early-fault root cause recognition method based on fuzzy data processing. The method comprises the following steps of 1, constructing a product early-fault root cause relevance tree layer model; 2, constructing a potential fault root cause data model; 3, collecting product service life period quality and reliability data; 4, on the basis of the fault relevance tree layer model, a process target node and data fuzziness analysis are determined, and then node influence factors and fuzzy values are determined; 5, constructing a product early-fault root cause fuzzy data envelopment analysis model; 6, estimating an efficiency evaluation value of a fault relevance tree node; 7, fault relevance node relative weights are divided, and node priorities are ranked; 8, results are analyzed, and fault root cause recognition is completed. Development of the early fault root cause recognition technology under the early fault mechanism recognition cognition fuzzy environment is broken through, prevention measures are adopted for product design, technological design stage and other early fault forming stages, afterward treatment is changed into beforehand prevention, and the new idea is provided for early fault prevention and rectification.

The invention discloses a video multi-target fuzzy data correlation method and device. The method comprises the steps of carrying out online target motion detection on a current video frame, and obtaining a possible moving object as an observation result; calculating the shielding degree among the prediction results of different targets in the current video frame; according to the shielding degree, respectively judging whether the shielding is generated between each prediction result and other prediction results or not; if no shielding is generated between the prediction result and any other prediction result, performing the first data correlation on the prediction result and the observation result; if the shielding is generated between the prediction result and any other prediction result, performing the second data correlation on the prediction result and the observation result, wherein the first data association is different from the first data association. By means of the above mode, the accuracy of multi-target tracking in a complex environment can be improved.

The invention, which relates to the technical field of image processing, discloses a definition enhancement method and apparatus for a dynamic video image. The method comprises: YUV data of a current pixel are obtained and normalization processing is carried out on the YUV data of the current pixel; neighborhood fuzzy data and image texture data of a Y component after normalization are calculated; sharpening enhancement is carried out by using the neighborhood fuzzy data and image texture data of the Y component and an overall contrast ratio of the Y component after sharpening enhancement is adjusted; and RGB data of the current pixel are calculated by using the adjusted Y component and a UV component of the current pixel and the RGB data of the current pixel are outputted. According to the technical scheme, high-efficiency definition enhancement is realized by improving aberration of different objects in the image rapidly; and the continuous processing demand of the dynamic video image can be satisfied.

The invention relates to a multi-sensor fire detection method based on hierarchical fuzzy fusion. According to the method, when a fire is detected, five fire parameters including a flame signal, the temperature change rate, the temperature, the carbon monoxide concentration and the smoke concentration are processed; a hierarchical fuzzy data fusion model composed of two-stage fuzzy reasoning models is built; the first stage takes the temperature change rate and the temperature as inputs; the second stage takes a first-stage judgment result, the carbon monoxide concentration, the smoke concentration and the flame signal as inputs; an output is a probability that the fire occurs; and accurate fire judgment can be carried out by comparing the value with a set threshold value. Through fuzzy fusion of multi-sensor fire information and a hierarchical mechanism of the two-stage fuzzy reasoning models, the number of fuzzy rule bases of a system is remarkably reduced, and the robustness, rapidity and accuracy of fire detection are improved.

The present invention provides a wireless local area network handover method based on fuzzy rules, the method comprises: the mobile stationS1: receives the beacon frames sent in predefined time interval from the current associated access point and the adjacent access point during a preset time period, obtains and stores signal strength of two access points through beacon frames;S2: performs fuzzy processing to the value of signal strength of the current access point and the adjacent access point and the change rate of signal strength respectively, then obtains fuzzy data characterizing levels of signal strength and the change rate of signal strength;S3: performs fuzzy reasoning taking the fuzzy data as the input according to a preset fuzzy rules, obtains a reasoning output variables which contain handover modes of the mobile station, and determine the target access point according to the reasoning outputs;S4: the mobile station authenticates with the target access point;S5: the mobile station sends the re-association request frame to the target access point after passing through the authentication; the handover is completed when the establishment of the re-association is finished after the mobile station receives the re-association response frame.According to the method of the present invention, the mobile station can automatically adjust the handover mechanism according to signal strength of the current access point and the adjacent point and change rate of signal strength and the better handover performance can therefore be achieved.

The invention belongs to the technical field of digital image intelligent processing, and particularly relates to an image defocusing blurring method based on deep learning. The method comprises the following steps: constructing a defocusing blurred data set by shooting or adding random blurring and the like, so that each group of data comprises a clear image as an original image and a plurality of blurred images as blurred images corresponding to the clear image; training a defocusing fuzzy deep neural network; recovering a blurred object out of a focal plane from the image through a deep neural network by using a non-alignment loss function; a non-pixel-level aligned deblurred data set is shot in a real scene, and a deep neural network is trained through a non-alignment loss function. Experimental results show that the out-of-focus blurred image shot in a real scene can be effectively recovered, and the proposed data set can effectively train the out-of-focus blurred network througha non-alignment loss function. The method can be used for camera zooming, robot vision systems and the like.

The invention relates to a new method which combines the fuzzy adapting of wavelet transform and curvelet transform in the image noise removal. The noise removal is one of important research programs in the image processing; however, the existing noise removal method can not completely solve the conflict between the noise removal and the edge preserving. The invention aims at providing an image noise removal method with the combination of the wavelet and curvelet fuzzy adapting on the basis of the defect of the prior art. The method of the invention establishes a flatness membership function of a sub-block to fuzzy express the edge information content in the sub-block and takes the membership function as the weight factor to carry out the data fusion to each sub-block by adopting the results from the noise removal with the wavelet transform and the curvelet transform. The method of the invention has the advantages that the data fusion substitutes the compulsory smoothing processing of the adapting combination method to solve the problem of blocking effect more thoroughly and retain more edge details; the advantages of the noise removal with the wavelet and the curvelet are flexibly integrated by the fuzzy data fusion so as to further improve the quality of noise removal.

A method for clustering data elements stored in a data storage system includes reading data elements from the data storage system. Clusters of data elements are formed with each data element being a member of at least one cluster. At least one data element is associated with two or more clusters. Membership of the data element belonging to respective ones of the two or more clusters is represented by a measure of ambiguity. Information is stored in the data storage system to represent the formed clusters.

The invention discloses a five-degree-of-freedom magnetic levitation electric spindle rotor displacement self-detection system and method. The system is composed of a fuzzy support vector machine displacement prediction module, two linear closed-loop controllers and two force / current converters. The fuzzy support vector machine displacement prediction module is composed of four fuzzy support vector machine radial displacement prediction modules and a fuzzy support vector machine axial displacement prediction module. Each of the radial displacement prediction module and the axial displacement prediction module is composed of a training sample set module, a data preprocessing module, a fuzzy data module, an optimal performance parameter determination module and a fuzzy support vector machinetraining module. The fuzzification data module fuzzifies the training sample set by using a fuzzy membership function; the optimal performance parameter determination module optimizes a penalty parameter and a kernel width by using a simplified particle swarm optimization algorithm, and obtains a group of penalty parameter and kernel width with the best performance index; and the system structureis simplified, and the control performance of the magnetic bearing is improved.

The invention discloses a window-based probability query method for fuzzy data in the high-dimensional environment, which includes: compressing information of the fuzzy region and information of the probability distribution function of each object by means of meshing, column charts and wavelet transformation; storing all compressed information of the object into an index file; inquiring by firstly calculating the upper limit of the probability that the object turns to the inquiry result according to all compressed information of the object and then pruning the unqualified object according to the upper limit of probability of each object so as to acquire a candidate answer set; and finally judging whether the candidate object is the real inquiry result or not according to the uncompressed information of each candidate object in the candidate answer set. On the basis of existing research and implement achievements on database and information retrieval and expansion and fusion of the existing compression methods, window-based probability query capacity for fuzzy data can be realized conveniently and quickly, dependence on the dimensionality of the fuzzy data is omitted, and the best performance can be achieved by the window-based probability query method.

An authentication method includes an enrolment stage comprising: receiving fuzzy data from a noisy authentication factor and fixed authentication data; generating a secret string independently from the received fuzzy data and the received fixed authentication data; deriving metadata from the fuzzy data and the secret string and helper data from the secret string and metadata; encrypting the helper data using the fixed authentication data as encryption key; outputting the encrypted helper data as public data, and an authentication stage including receiving the public data output during the enrolment stage, decrypting the received public data using the fixed authentication data as decryption key, recovering the helper data and the metadata from the decrypted public data, reproducing the secret string using the further fuzzy data and the recovered metadata, validating the reproduced secret string using the recovered helper data, and releasing the reproduced secret string if the validating yields a positive outcome.

Systems and methods for obscuring data from a data source include devices and processes that may objectively measure the information loss for a dataset that is caused by applying a privacy policy, and may select and apply a policy to the dataset based on the measured information loss. The systems and methods may measure the information loss for a large dataset by taking a representative sample from the dataset and applying the policy to the sample in order to quantify the information loss. The quantified information loss can be iteratively used to change the policy in order to meet utility and / or privacy goals, and the system can subsequently apply the changed policy to the dataset.

The embodiment of the invention discloses a vulnerability discovery method and device and electronic equipment. The method comprises the steps that fuzzy configuration data based on a target application program is generated; the target application program is started in a virtual machine arranged in advance, and a sample file for the target application program is read; the format of the sample file for the target application program is analyzed to obtain a data block, the fuzzy configuration data based on the target application program is called, and fuzzy configuration is carried out on the data block to generate fuzzy data; the fuzzy data is written into an input interface of the target application program, the fuzzy data runs in the target application program, and vulnerability discovery is carried out based on the information of a running result of the fuzzy data. By applying the vulnerability discovery method and device and the electronic equipment, the vulnerability discovery efficiency can be improved.

The embodiment of the invention relates to the field of natural language processing, and discloses a voice recognition text error correction method and device, electronic equipment and a storage medium. The method includes: receiving voice information; identifying at least one user intention corresponding to the voice information; according to the identified at least one user intention, selectingall data of the user intention in the cloud data set as a personalized fuzzy data set; carrying out error correction on the text identified according to the voice information by combining the personalized fuzzy data set and the preset basic fuzzy data set. The data volume required for error correction is reduced while the error correction accuracy is ensured through the personalized user intentionof the user, and the error correction efficiency is improved.

The invention discloses a Chinese geographic coding and decoding method. The method comprises the following steps: a GIS (Geographic Information System) including an electronic map, coordinate data and geological information data is called for use, wherein each map unit on the electronic map corresponds to one coordinate data and a group of geographic information data; as for each map unit, a standard Chinese geographic indication is set and stored in a standard data base, the storage position of the standard Chinese geographic indication in the standard data base is mapped in the geographic information data, and the standard Chinese geographic indication comprises separator fields, pointing fields and identification fields; and as for each standard Chinese geographic indication, a fuzzy Chinese geographic indication is set and stored in a fuzzy data base, the storage position of the fuzzy Chinese geographic indication in the fuzzy data base is mapped in the geographic information data, the mapping of the fuzzy Chinese geographic indication and that of the standard Chinese geographic indication are correlated, and the fuzzy Chinese geographic indication comprises a primary fuzzy Chinese geographic indication and a secondary fuzzy Chinese geographic indication.

The invention relates to an evaluation method for arch rib hoisting construction stability based on fuzzy comprehensive evaluation, belonging to the technical field of bridge construction. The method comprises the following steps: S1, determining factors affecting the arch rib hoisting construction stability; S2, dividing the evaluation grades of the arch rib hoisting construction stability according to factor theoretical values and defining critical judgment values of the evaluation grades under the factors; S3, acquiring the actual monitoring values of the factors in a construction process, determining a membership function, and calculating the membership degrees of the stability evaluation grades of the factors; S4, determining the factor weights by means of a subjective and objective combined method; and S5, performing fuzzy comprehensive evaluation on the arch rib hoisting construction stability according to the factor weights and the membership degrees. According to the method provided by the invention, the safety problem in the arch rib construction process is quantitatively evaluated, so that the construction safety and the construction monitoring effectiveness are guaranteed; by treating a fuzzy evaluation object through a precise digital means, quantitative evaluation on fuzzy data is performed.

The invention discloses a method and a system for detecting the quality defect of software based on intelligent dynamic fuzzy detection. The method comprises the following steps of: determining software to be detected and defining a detection range so as to invoke a corresponding detection strategy; constructing fuzzy detection data used for detection according to the detection strategy; executing the defect detection of the software to be detected by utilizing the fuzzy detection data; monitoring the process of carrying out the defect detection on the software to be detected; if discovering abnormity by monitoring, carrying out state recording on the process of the fuzzy detection of the software to be detected and feeding back a recording result to a strategy editor; carrying out strategy editing and regulation automatically by the strategy editor according to the recording result which is fed back so as to form a novel detection strategy and repeating the operation of the step 2 to the step 6 according to the novel detection strategy; and carrying out defect positioning according to the detection result. The system comprises a strategy editor, a detection strategy library, an intelligent fuzzy data generator, a detection engine, a software state monitor, a defect positioning module and a result generation module.

The invention discloses an optical aberration blur removing method based on deep learning. The method comprises the following steps: 1) obtaining a point spread function of the optical system with aberration; 2.1) selecting a high-resolution image, and performing energy domain transformation to obtain an energy domain image; 2.2) performing block convolution on the energy domain image by using thecalculated correction point diffusion matrix to obtain an energy domain simulation fuzzy graph; 2.3) performing numerical domain transformation on the energy domain simulation blurred image to obtainan aberration blurred image, and forming an aberration blurred data set; 3) based on the aberration fuzzy data set, training an aberration correction neural network; and 4) correcting an image shot by the aberration optical system developed and produced by using the optical parameters through the aberration correction neural network obtained by training in the step 3) to obtain a corrected image.When the method is used, optical parameters of a camera (camera head) are operated by adopting the method disclosed by the invention, and image blurring caused by aberration of an optical system canbe well eliminated.

The invention discloses a mutual constraint based fuzzy data classification method which is used for information category mode setup and data category analysis. The mutual constraint based fuzzy data classification method is characterized by including steps: using an elasticity based four-point center and border line algorithm to construct a category rule (quintuple mode); using a constraint-based inching classification algorithm and an optimized classification rule algorithm based on self-training to optimize and adjust the category rule (quintuple mode). The method has special sample (currently unknown category) detection capability, is suitable for classification analysis and excavation of pervasive data, and is well adaptable to outliers, category topological irregularity and 'acute border' problems. Further, the method is applicable to classification and analysis of data sets which are large in data volume and cannot be read in a memory by one time, and has functions of autonomous category adjustment and identification. Compared with an existing method, the mutual constraint based fuzzy data classification method has the advantages that average recognition rate is up to 99.47%, average false alarm rate is only 5.2%, and operating speed is slightly lower than a traditional algorithm.

The distributed new energy related multi-source heterogeneous data processing method comprises the following steps: M1, synthesizing big data reading and analyzing data attributes; M2, establishing attribute index and classification of data using machine learning mode; M3, carry out independent data integrity assessment on that read-in data to eliminate incomplete, redundant and fuzzy data; M4, the data is evaluated for the first time, the noise is eliminated, the noise is reorganized and filtered, and the data is finally reduced; M5, based on the matrix recovery theory, the structured low rank representation model and the structured sparse constraint and low rank constraint M6 are established to separate the data errors and to compensate the error segmentation defects. M8, security management of fused data to prevent illegal data loss. This method can reduce the repeated investment of manpower and equipment in the distributed new energy business, and reduce the indirect investment cost.

The invention relates to a target track fuzzy data fusion method based on video monitoring and is applied to actual police work. Police officers cannot acquire GPS position data through active uploading of abnormal objects, and only acquires the positioning information through a monitoring sensor network. Most of behaviors generated by suspected objects of various cases in different stages such ascollusion, implementation, hiding and the like have subjective invisibility, so that the data acquired by the monitoring network is fuzzy data, and through the target track fuzzy data fusion method based on video monitoring, a fuzzy track can be recovered, and help is brought to police work and target tracking.

The present invention discloses a Lucene full-text retrieval based Chinese word segmentation method. The method comprises: storing a dictionary in a database in the form of one word for each row; caching the dictionary in the database into a server in the form of a tree; inputting text information that needs to be segmented; matching a text with a caching dictionary tree word by word, and outputting a successfully matched longest word; and outputting a word segmentation result. According to the method provided by the present invention, a user can extract useful information from massive fuzzy data for detailed study and summarization, and it is convenient for the user to perform semantic analysis and data analysis, so that a problem in a marketing service can be found in time, thereby improving a power grid marketing service level.

The invention provides a fuzzy data association method in a clutter environment. The fuzzy data association method comprises the following steps: step 1, establishing an interconnection matrix of candidate measurement and targets according to the distribution condition of measurement in a confirmation area; step 2, constructing a statistical distance through an interconnection rule; step 3, constraining the target function by utilizing KL divergence regular information; step 4, calculating the interconnection probability between each candidate measurement and different targets in the observation area through an iterative optimization algorithm; and step 5, updating the target state and the covariance by using probability weighting. The real-time performance of multi-target tracking is greatly improved, the multi-target tracking precision and the effective tracking rate of the method are similar to those of a classical joint probability data association algorithm, and the requirement for effective target tracking can be met. Correspondingly, the invention further provides a multi-target tracking method.

The invention relates to the technical field of quantum machine learning, fuzzy set theories and network confrontation, and provides a quantum fuzzy machine learning confrontation attack model method.The method comprises the steps: S1, constructing a quantum fuzzy data sample of a legal user; S2, constructing an attack strategy by a malicious attacker, and adding the constructed disturbance intoa quantum fuzzy data sample of a legal user to form a quantum fuzzy confrontation sample; S3, enabling the quantum fuzzy machine learning system to perform classification according to the quantum fuzzy adversarial samples and make corresponding correct decisions or error decisions; wherein when the quantum fuzzy adversarial sample is a quantum fuzzy data sample of a legal user, a correct decisionis made; when the quantum fuzzy adversarial sample is a sample constructed by a malicious attacker, an error decision is made, the attack purpose is achieved, the defects of vulnerability and many defects of a quantum fuzzy machine learning algorithm are overcome, and safety and robustness are improved.

The invention discloses a quantum fuzzy machine learning adversarial defense model method. The method comprises the steps of S1, constructing a quantum fuzzy data sample of a legal user; S2, simulating a malicious attacker to construct an attack strategy: adding the constructed disturbance into a quantum fuzzy data sample of a legal user to form a quantum fuzzy countermeasure sample of the malicious attacker; S3, submitting the quantum fuzzy data sample of the legal user and the quantum fuzzy adversarial sample of the malicious attacker to a quantum fuzzy machine learning system for training and learning, and making a correct decision by the quantum fuzzy machine learning system, wherein the quantum fuzzy machine learning system comprises an adversarial defense module, and the adversarialdefense module is an adversarial sample for defending malicious attackers, so that the quantum fuzzy machine learning system makes a correct decision. The model method can effectively resist the attack of a malicious attacker, improve the safety and robustness of a quantum fuzzy machine learning system, and ensure the safe and reliable operation of a quantum fuzzy machine learning algorithm.

The invention discloses an error correction method for a multi-channel HRWS-SAR channel. The error correction method comprises the following steps of 1): performing distance pulse compression on original echo data received by each channel; 2) searching an isolated strong scattering point with the maximum power in the two-dimensional time domain of the echo signal; 3) performing sub-aperture segmentation on the data of each channel; 4) in the azimuth frequency domain, extracting strong scattering point signals in the sub-apertures; 5) splicing the strong scattering point signals in the sub-apertures; and 6) estimating an antenna array flow pattern by using the distance Doppler spectrum with the plurality of strong scattering points being not fuzzy, and completing channel error correction. The method aims at solving the problems that an existing channel error correction method depends on a parameter model, azimuth space-variant errors are difficult to correct accurately, and robustness is poor, isolated strong scattering point echo signals in an imaging scene are automatically extracted from fuzzy data through adoption of a sub-aperture signal processing technology, an unambiguous Doppler spectrum of the echo signals is obtained and used for estimating channel errors, and therefore, accurate correction of the errors is achieved.

The invention provides a distribution network equipment health diagnosis method for multi-source information fusion analysis, which is used for obtaining the health state level of distribution equipment in a distribution cable health diagnosis system, and comprises the following steps: obtaining signals of each influence factor dimension of the distribution network equipment through a sensing layer; the signals of the influence factor dimensions comprise terahertz time domain signals, partial discharge signals and temperature signals; transmitting the signal of each influence factor dimension to a data layer through a transmission layer; completing feature extraction of the signal of each influence factor dimension in the data layer to obtain an influence factor vector of each dimension; and using a fuzzy evaluation method to obtain a comprehensive evaluation result taking the distribution network equipment as an evaluated object. A fuzzy data fusion method is adopted to evaluate the health state of the system, and a high-quality diagnosis system can be obtained during practical application of multi-source information fusion analysis.

The invention discloses a video image processing method suitable for video monitoring equipment. The method comprises the following steps: acquiring an image frame sequence; splitting the image frame sequence into a training set and a test set, and training the neural network model through the training set to obtain a trained neural network; extracting a foreground image, extracting a foreground image of each image frame, sequentially generating a fuzzy sequence of the foreground image of each image frame, sequentially judging whether each image frame has a fuzzy region or not according to the corresponding fuzzy sequence, if so, deleting the image frame with the fuzzy region, and if not, storing the image frame without the fuzzy region; according to the technical scheme, the volume of the image sequence formed by the video or the multiple images can be compressed, so that on the video monitoring equipment, the storage space required by a server side can be greatly compressed, the proportion of effective data can be increased, invalid fuzzy data can be cleared away, and the user experience can be improved.