26 results about "Fuzzy data" patented technology

The invention relates to the technical field of artificial intelligence, discloses a multi-modal data-driven classroom teaching quality comprehensive evaluation method and system, and aims to solve the problems of low evaluation precision, high privacy risk and the like caused by difficult multi-modal data fusion standardization, insufficient analysis result interpretability and fuzzy data acquisition compliance in the prior art. The method comprises the following steps: executing privacy-protected multi-modal data acquisition and standardization on an edge device; constructing a cross-modal causal time sequence alignment model to generate a unified teaching event sequence; semantic embedding is generated in combination with the teaching knowledge graph and the graph attention network; fusing causal and semantic information through a causal graph attention network, and outputting a teaching quality comprehensive state vector; when the indexes are abnormal, a causal path is traced back, and an interpretable diagnosis report and a teaching strategy are generated. The system comprises corresponding functional units, and realizes closed-loop evaluation from data acquisition to strategy generation. According to the method, the accuracy, compliance and interpretability of evaluation are effectively improved.

The invention provides a bimodal image restoration method based on a diffusion model. The bimodal image restoration method comprises the following steps: S1, acquiring a bimodal fuzzy data set; s2, encoding and fusing the multi-modal data; s3, decoding processing is carried out; and S4, submerged space diffusion recovery. The invention aims to solve the problems of low efficiency, poor reconstruction quality and single data source of the existing image reconstruction technology in a complex scene.

Provided are a training method and device of a target detection model, and a target detection method. The training method comprises: obtaining a training sample set, comprising first and second training samples, the first and second training samples respectively comprising clear data and fuzzy data; inputting the first training sample into the target detection model; determining a first loss value of the target detection model according to an output value of the target detection model and the second training sample, using a first loss function; inputting the first loss value into a Gaussian mixture model; constructing a fuzziness matrix using the Gaussian mixture model, the fuzziness matrix being used to represent the fuzziness of the data in the first training sample and the second training sample; constructing a fuzzy loss function according to the fuzziness matrix, the fuzzy loss function being constructed to be able to distinguish between the clear data and the fuzzy data in the first training sample and the second training sample; constructing a second loss function according to the fuzzy loss function; and retraining the target detection model using the first and second training sample sets and the second loss function.

The technical scheme of the invention discloses an NFT-based data trusted space management and utilization method, which is characterized in that corresponding service data trusted spaces are constructed for service data, different types of service data correspond to different service data trusted spaces, and data are isolated among different service data trusted spaces. According to the method, the NFT smart contract is constructed for different business scenes and business data, so that data isolation and independent management of the different business scenes are realized, credible spatialization management of scene data is realized, and the problems of fuzzy data permission boundary, unauthorized access and the like in a traditional centralized management and unified account book mode are effectively avoided. According to the method, the NFT is innovatively introduced as a management unit of the business data, right confirmation, right control and transaction circulation of the business data are unified into the NFT smart contract for management, and right confirmation and dynamic control of the data are better realized.

The invention relates to the technical field of corner reflector adaptive adjustment, in particular to a corner reflector adaptive adjustment method, system and device based on multi-source radar remote sensing data. Comprising the following steps: S1, acquiring an installation position of a corner reflector in a terrain lifting complex scene of a power transmission line, extracting terrain parameters according to the installation position, and extracting an interference type according to the terrain parameters; the method accurately adapts to the specific requirements of the terrain uplift complex scene of the power transmission line, achieves the accurate recognition of the terrain uplift complex scene, the comprehensive extraction of terrain parameters and the directional judgment of interference types through the access of power transmission line management end data and a digital elevation model, combines the complementary advantages of multi-source radar data, and achieves the accurate recognition of the terrain uplift complex scene. Quantitative inversion and prediction of interference parameters are completed, the problems of fuzzy interference identification and insufficient data support in complex terrains in the prior art are effectively solved, and the stability of radar echo signals and the accuracy of pole and tower state monitoring data are remarkably improved.

The application discloses a fuzzy decision tree construction method and device, equipment and a computer storage medium. In the case that a construction data set is acquired, a first mapping reduction model is used to perform mapping reduction processing on data in the construction data set, a plurality of fuzzy partitions in the construction data set are obtained, a second mapping reduction model is used to perform mapping reduction processing on data of each fuzzy partition in the plurality of fuzzy partitions, a plurality of fuzzy sub-partitions of each fuzzy partition are obtained, a root node set and a sub-node set are determined according to the obtained fuzzy partitions and fuzzy sub-partitions, and finally, a fuzzy decision tree of the above construction data set is constructed based on the root node set and the sub-node set. In this way, through twice mapping reduction processing, a fuzzy decision tree classifier can be distributed constructed from big data based on distributed parallel computing, data classification processing of a data set with high dimension and large data quantity is realized, and the efficiency of inductive classification of related fuzzy data is improved.

The invention relates to the field of road cavity detection, and discloses a road cavity and underground disease body collapse hidden danger detection device and detection method.The road cavity and underground disease body collapse hidden danger detection device comprises a movable bottom plate, a supporting plate is installed on the movable bottom plate, a rotating disc is movably installed in the middle of the supporting plate and rotates in the supporting plate, and a transmission shaft is installed at the bottom of the rotating disc; a first gear is fixedly installed on the transmission shaft, a rotating mechanism is arranged on one side of the first gear, a hydraulic rod is installed on the rotating disc, a detection arm is installed at the top end of the hydraulic rod, an electromagnetic wave transmitting module is installed at one end of the detection arm, and the electromagnetic wave transmitting module comprises an electromagnetic wave transmitting body and a broadband transmitting antenna. Through the arrangement of the electromagnetic wave transmitting module, the receiving sensor and the data preprocessor, a set of cooperative linkage underground anomaly detection system is constructed, and the problems of large signal interference, fuzzy data analysis, anomaly positioning deviation and the like in the traditional detection technology are effectively solved.

The invention discloses a regional crowdsourcing ionospheric scintillation monitoring and early warning method and system based on low-cost equipment, and belongs to the field of ionospheric scintillation monitoring, and the method comprises the steps: recording reference station observation data and ISMR data as an accurate data set; recording observation data acquired by the low-cost equipment as a fuzzy data set; performing flicker intensity assignment of ROTI and GFTI by using ISMR data in the accurate data set; interpolation is carried out on the flicker index of the position where the puncture point coordinates in the fuzzy data set are located by adopting common Kriging interpolation; subtraction is carried out on the interpolation result and the index of the corresponding position in the fuzzy data set, and an abnormal difference value in the subtraction result is calculated; and dividing the obtained ionospheric scintillation detection result according to a scintillation generation range, and analyzing a possible scintillation generation position so as to carry out early warning. A large amount of low-cost equipment observation data in the area is introduced to fill the blank area of professional equipment monitoring data, and the monitoring data density and the resolution of the monitoring result are improved.

In a computer-implemented method for providing fuzzy data to a user, firstly, a user request to access data is received; then, an authorization level associated with the received request is identified. Next, fuzzy data corresponding to the received request is accessed within a protected enclave. The accessed data is fuzzified using a fuzzing algorithm that generates a fuzziness level compatible with the identified authorization level. Finally, the accessed fuzzy data is provided to the user from the protected enclave. Related system and computer program products are also disclosed.

The invention discloses a food detection sample management method and system based on big data analysis, particularly relates to the technical field of food detection data processing, and is used for solving the problem that in an existing sample management big data platform, unstructured text semantics are fuzzy, and the intelligent retrieval and knowledge mining ability is insufficient due to the lack of deep semantic association among data. The method comprises the following steps of: acquiring sample full-process data containing structured and unstructured data, judging a semantic fuzzy condition of the unstructured data by adopting a text similarity algorithm, and performing sequence analysis and semantic completion on a fuzzy text by utilizing a hidden Markov model in combination with space-time correlation characteristics so as to generate a semantic clarification result; and then a semantic association rule is established and optimized based on the semantic clarification result and the structured data, and finally a food detection sample information knowledge graph is constructed according to the optimization rule, so that intelligent semantic retrieval and deep knowledge mining of the sample information are realized.

The invention discloses an intelligent path planning method based on a multi-resolution region segmentation model, and the method comprises the steps: firstly carrying out the obstacle recognition and spatial feature extraction based on a collected environment map image, and constructing an obstacle position coordinate database; secondly, based on the obstacle position coordinate database, a recursive quadtree method is adopted to carry out self-adaptive division on a planning area, and multi-resolution representation of the environment is established; and finally, simultaneously constructing two search trees from a starting point and an ending point, dynamically adjusting a sampling strategy based on leaf area size, obstacle density and depth information, performing path search, and outputting a path planning result. According to the method, the problems that traditional environment modeling information is fuzzy and data support is insufficient are solved, and unification of high efficiency, safety and integrity of path planning is achieved.

The application discloses a food detection sample management method and system based on big data analysis, and particularly relates to the technical field of food detection data processing, and is used for solving the problem of insufficient intelligent search and knowledge mining ability caused by unstructured text semantic ambiguity and lack of deep semantic association between data in the existing sample management big data platform; the problem is solved by the following steps: obtaining sample full-process data containing structured and unstructured data, using a text similarity algorithm to determine the semantic ambiguity of unstructured data, using a hidden Markov model combined with space-time correlation characteristics to perform sequence analysis and semantic completion on the ambiguous text to generate a semantic clarification result, then establishing and optimizing semantic association rules based on the semantic clarification result and the structured data, and finally constructing a food detection sample information knowledge graph according to the optimized rules, so as to realize intelligent semantic search and deep knowledge mining of sample information.

The invention provides an unstructured data feature multimode extraction and identification method, which comprises the following steps of: setting a data type judgment module in a system, after unstructured data is accessed to the system, firstly carrying out preliminary analysis on the data, judging the data type according to the format and content features of the data, and judging the data type according to the content features of the data; carrying out an identification method based on content, natural language processing and semantic analysis on the plain text; and for the picture, a content identification method based on machine learning is adopted. According to the method, by designing a multimode feature extraction strategy, accurate recognition of data with various formats and semantic fuzziness is achieved, the capability of capturing text context semantic information is enhanced, the processing speed and precision of complex background and low-resolution pictures are improved, a flexible technical combination scheme is constructed, and the method is suitable for large-scale popularization and application. The feature extraction requirements of different types of unstructured data are met, so that the overall data processing efficiency and accuracy are improved, and the ever-increasing unstructured data processing requirements in the industries of finance, telecommunication, medical treatment and the like are effectively met.

The invention discloses an underground water level early warning method and system. The method comprises the following steps: acquiring an underground water level time sequence data set of all stations in a target area; a hierarchical clustering algorithm based on dynamic time warping is adopted to recognize spatial heterogeneity of underground water level changes, and all stations are classified; for each type of stations, acquiring the groundwater level mean value and the water level change rate of each type of stations as input variables, and standardizing the input variables; performing fuzzification processing on the standardized input variables through a Gaussian membership function to generate multiple pieces of input variable fuzzy data; performing fuzzy reasoning on the multiple pieces of input variable fuzzy data according to a preset fuzzy reasoning rule to obtain multiple pieces of output variable fuzzy data; the multiple pieces of output variable fuzzy data are converted into underground water level risk coefficients through a gravity center method; and determining the underground water level early warning level of the corresponding type of station according to the underground water level risk coefficient of each type of station.

The application discloses an event-image deblurring method based on an LSTM, and comprises the following steps: constructing an event-image deblurring network model for processing a blurred image; training the event-image deblurring network by using training set data in a blurred data set, obtaining a trained event-image deblurring network model; inputting test set data in the blurred data set into the trained event-image deblurring network model, and obtaining a clear frame sequence. The advantages of time sequence information of events are fully used, the deblurring performance is improved, the generalization of the network is improved, one model can be used to complete deblurring output of any time resolution, and the time resolution of the output clear frame is controlled by the time resolution of the input event. Meanwhile, a bidirectional coding-decoding structure is used to enhance the deblurring effect of the network, and the method has the following advantages: one model needs to be trained, and any time resolution output can be achieved, so that the problem of limited generalization of the model is solved.

The present application relates to the technical field of power system planning evaluation, and more particularly to a method and system for dynamic evaluation of energy storage based on scenario joint debugging and risk early warning, the method comprising: constructing an evaluation index system; obtaining original index data of the energy storage technology to be evaluated under standard working conditions and performing dimensionless processing; calculating the first weight of each evaluation index using a grey correlation degree analysis method, calculating the second weight of each evaluation index using an entropy weight method, and obtaining a basic fusion weight by fusion; determining a scenario correction coefficient based on the demand proportion of grid frequency modulation and ramp service, and performing a first dynamic correction on the original index data; determining an environmental risk correction coefficient based on an extreme temperature environment, and performing a second dynamic correction on the original index data; and performing weighted calculation on the original index data and the basic fusion weight, and outputting a comprehensive utility score. Through the present application, the problems of lack of scene adaptability and environmental risk early warning and the problem of fuzzy data quantization in the existing energy storage evaluation method are effectively solved.

The embodiment of the application provides a kind of method and system for identifying air target based on fuzzy inference, it is related to target identification technical field.The method for identifying air target based on fuzzy inference includes: obtaining the multiple flight data corresponding to target aircraft at multiple times;According to the fuzzy processing of multiple flight data by preset flight parameter, generate multiple input variable fuzzy data;According to the multiple input variable fuzzy data obtained by preset intuitionistic fuzzy inference model and multiple output variable fuzzy data;The comprehensive identification result is obtained by fusing multiple output variable fuzzy data.The method for identifying air target based on fuzzy inference can realize the technical effect of the accuracy of air target identification.

The invention relates to a network data security protection method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring network transmission data to be identified and filtered; performing preliminary screening on the network transmission data through the black and white list database to obtain intercepted alternative network data and first intercepted data; performing secondary screening on suspected data in the alternative network data through a regular rule base to obtain intercepted candidate network data and second intercepted data; identifying and intercepting fuzzy data in the candidate network data through a semantic understanding model to obtain target network data and third intercepted data; and according to a processing strategy matched with the first intercepted data, the second intercepted data and the third intercepted data, executing an alarm operation corresponding to the processing strategy. Therefore, full-scene coverage from explicit violation to implicit violation is achieved, the missed judgment rate and the misjudgment rate are reduced, redundant data are filtered through black and white list parallel screening, and the processing efficiency and the recognition precision in a high-concurrency scene are balanced.

This invention discloses a BIM-based tunnel boring machine (TBM) monitoring system based on fuzzy data query and nearest neighbor algorithm. The key technical features include: a data acquisition module, a data preprocessing module, a fuzzy data query module, a nearest neighbor algorithm module, path prediction, and a BIM platform integration module. The data acquisition module is used to collect real-time tunneling data of the TBM, including tunneling speed, position, and direction. The data preprocessing module is used to preprocess the collected tunneling data, including noise reduction and outlier handling, to improve the accuracy and reliability of the data. Through modular design and the combination of advanced algorithms, this system achieves comprehensive monitoring of the TBM tunneling process. The modules work closely together, from data acquisition and preprocessing to fuzzy query, prediction, and visualization, providing accurate and efficient monitoring and decision support for TBM construction, thus improving the safety, efficiency, and intelligence level of TBM projects.

The invention provides a commercial vehicle side angle radar detection method, a storage medium and electronic equipment. The method comprises the following steps: initializing an angle radar and acquiring basic information of a vehicle body; based on the basic information of the vehicle body, the angular radar performs partition scanning detection to obtain radar detection data; wherein the subareas comprise a side blind area, an adjacent lane area and a vehicle door periphery early warning area; processing the radar detection data to obtain target feature data; and performing target identification on the target feature data, and performing risk assessment and risk early warning based on a target identification result. According to the invention, the technical problems of fuzzy division, simple data processing, low target identification precision and single early warning mechanism of the side detection area of the commercial vehicle in the prior art are effectively solved.

The embodiment of the invention provides an image deblurring method based on motion prior information guidance, and relates to the technical field of image processing. The image deblurring method based on motion prior information guidance comprises the following steps: S1, predicting an offset (px, py) and a weight (w) from an input blurred image X through a motion estimator, S2, generating low-layer feature embedding of the input blurred image X through a convolutional layer; s3, a diffusion stage; s4, a motion estimator; s5, capturing motion information along the vertical and horizontal directions; and S6, a motion injection module, wherein prior features related to motion are utilized in the Transform block. The overall distribution of the blurred image is learned by a prior enhanced Transform; a priori is fused into the attention module to enhance the representation of motion details. By combining with the generative model, these designs can effectively capture and utilize the motion pattern responsible for obscuring the workpiece. Finally, an excellent image deblurring effect is achieved under verification of various fuzzy data sets.

The invention relates to the technical field of neural network system structures, in particular to a fuzzy AI model training method for a data processing accelerator, which comprises the following steps: acquiring fuzzy data, sending the fuzzy data into the accelerator, tracking a first dragging point of a signal to merge an interference position set, identifying offset nodes along a path, and serially connecting to construct a sequence; and observing round-trip conflict tracks in the channels to recombine signal segments, stripping steering positions, extending non-offset linear paths, connecting the channels to the end of the process in sequence, positioning feature mapping segments, and identifying sample relationships to obtain a fuzzy training result. According to the method, the initial disturbance position is recognized by capturing signal response changes, path offset nodes are continuously tracked, dynamic connection is achieved, a clear track sequence of a conflict area is constructed to improve the connection stability, guide path fragments are recombined to define the channel propulsion sequence, and propulsion disorder caused by multi-path staggering is avoided; logic smooth conduction is achieved, mapping association is enhanced, and it is ensured that a fuzzy data guiding result has integrity and path consistency.

This invention discloses a method and apparatus for recommending job postings based on career prediction, relating to the field of online recruitment technology. The method includes: acquiring user profile data of job seekers; predicting the next virtual work experience based on the job seeker's resume information; vectorizing the next virtual work experience to obtain vectorized representations of multiple virtual work experiences, and adding them to the user profile data; performing vector retrieval on the vectorized representations of all job posting information to obtain vectorized representations of several job postings that match the vectorized representations of the virtual work experiences; filtering and sorting the candidate set of job postings according to screening criteria, and using the sorted job postings as the job posting recommendation results for the job seeker. This invention significantly overcomes the strong dependence on explicit user needs, resume completeness, and behavioral data, effectively solving the problems of fuzzy job seeker needs, data sparsity, and cold start.

The invention relates to a virtual power plant operation strategy determination method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: for each candidate operation strategy related to a virtual power plant, obtaining corresponding index data from each corresponding strategy sample, determining an entropy value of each index based on each index data, and determining an objective weight of the candidate operation strategy under each index based on the entropy value of each index; based on the triangular fuzzy data corresponding to the candidate operation strategies, determining subjective weights of the candidate operation strategies under the indexes; for each candidate operation strategy, based on the subjective weight and the objective weight under each index, determining the fusion weight under each index, and based on each index data and the fusion weight under each index, determining the comprehensive satisfaction degree of each strategy sample; and determining an optimized operation strategy of the virtual power plant based on the plurality of comprehensive satisfaction degrees corresponding to each candidate operation strategy, thereby improving the effectiveness of the operation strategy of the virtual power plant.

The application relates to the technical field of data processing, in particular to a semantic enhancement storage and distributed retrieval system for unstructured data. The system collects target unstructured data through a data collection module, encodes the target unstructured data through a semantic analysis module, monitors fuzzy data through a semantic enhancement module, stores the data in fragments through a semantic storage module, and acquires a distributed parallel retrieval scheme through a distributed retrieval module. The application collects target unstructured data, performs semantic analysis, semantic enhancement and data storage on the target unstructured data according to the target unstructured data, and acquires a distributed parallel retrieval scheme in real time, so that the target unstructured data semantics are enhanced, and the target unstructured data is efficiently stored and retrieved.

The invention discloses a channel flow measurement method, which belongs to the technical field of flow measurement, and comprises the following steps: S1, selecting and finishing a measurement section; s2, arranging a water level gauge; s3, mounting and positioning the water cutter; s4, water cut-off operation; s5, data acquisition; and S6, flow calculation. According to the method for measuring the channel flow through the scheme, the complete process of section finishing, water breaking and blocking, data collection and accurate calculation is constructed, the measurement basis is standardized from the source, errors caused by irregular sections and fuzzy basis are avoided, high-precision professional equipment is not needed, the measurement process steps are clear, logic coherence is achieved, and the measurement accuracy is high from section preparation to flow calculation. The operation of each link is simple and easy to understand, complex professional skill training is not needed, the whole operation can be completed by a single person, the precision is reliable, the truth value acquisition capability is high, a traceable calibration path is provided for flow measurement, and the problem of blank existing calibration systems can be solved.