20results about How to "Increase data dimension" patented technology

The invention provides a geological disaster monitoring multi-dimensional visual early warning system and method based on big data, and relates to the technical field of geological disaster monitoring, and the method comprises the steps: collecting ground surface displacement, slope stability and underground water level in real time; analyzing and processing the collected data to generate a surface risk coefficient, a slope risk coefficient and a water level risk coefficient; establishing a risk level evaluation model of which the input quantity is an earth surface risk coefficient, a slope risk coefficient and a water level risk coefficient and the output quantity is a geological risk level, obtaining historical earth surface displacement parameters, slope stability parameters and underground water level parameters, and defining corresponding geological risk level coefficients; and comparing the risk grade coefficient analyzed by the evaluation model with a threshold value, and sending out a corresponding early warning signal. According to the method, the accuracy of disaster early warning and risk assessment can be improved, model parameters are verified by using historical disaster data in big data, and the accuracy and reliability of a risk grade evaluation model are improved, so that the geological disaster risk is predicted more accurately.

The present disclosure provides an object recommendation method, an object recommendation device, an electronic device and a storage medium, which can be applied to the technical field of big data. The object recommendation method comprises: determining M first objective indexes, N second objective indexes and a target variable relationship matched with an application scene, the target variable relationship being used to determine a to-be-recommended object, the first objective indexes being used to represent transaction characteristics of a user, the second objective indexes being used to represent social relationship characteristics of the user, M being greater than or equal to 1, and N being greater than or equal to 1; obtaining a recommendation data set of a target user, the recommendation data set comprising first objective data corresponding to the first objective indexes and second objective data corresponding to the second objective indexes; and determining the to-be-recommended object according to the M first objective data, the N second objective data and the target variable relationship.

The invention is suitable for the technical field of laser additive manufacturing, and provides a laser additive multi-mode monitoring device which comprises a first mounting plate assembly, a second mounting plate assembly, a powder bed monitoring module, an OT camera monitoring module, a structured light camera module, an infrared monitoring module, a printing substrate and a supporting shell. The supporting shell is of an external bearing structure, and the first mounting plate assembly and the second mounting plate assembly are transversely arranged along preset mounting positions on the top of the supporting shell and are fixedly connected with the supporting shell through detachable fixing pieces. According to the method, multi-modal data is fully covered, and a complete process data link is constructed; acquisition precision and stability are obviously improved, and requirements of severe working conditions are met. And the process optimization of data closed-loop driving is realized, and the forming quality and the large-scale production capacity are improved.

The invention relates to a Raman spectrum data preprocessing method and a Raman spectrum data preprocessing system for pathogenic microorganisms, which are applied to the field of spectrum data processing, and the method comprises the following steps: obtaining Raman spectrum data of the pathogenic microorganisms, the Raman spectrum data comprising spectrum intensity data of Raman spectrum; performing normalization processing on the spectral intensity data to obtain normalized spectral intensity data; marking the normalized spectral intensity data in a preset rectangular coordinate system, and obtaining position information of the normalized spectral intensity data in the rectangular coordinate system; performing polar coordinate representation on the normalized spectral intensity data according to the position information, and obtaining a spectral data angle according to an included angle between the spectral intensity data after polar coordinate representation and a polar axis; and performing matrix representation on the spectral data angle to obtain and output matrix spectral data. The method has the technical effects that the feature extraction difficulty is reduced, and the possibility of losing effective information is reduced.

The present application relates to a distribution line protection method, device and electronic equipment, and to the field of circuit protection technology. The method includes: obtaining a voltage waveform data array on both sides of the distribution line, and a current waveform data array on both sides of the distribution line; filtering the current waveform data array to obtain a filtered current waveform data array; determining the voltage qualification rate corresponding to the voltage waveform data array, and determining the waveform similarity corresponding to the filtered current waveform data array when the voltage qualification rate at both ends of the distribution line meets the discrimination range; determining the section discrimination criterion of the distribution line based on the waveform similarity and the discrimination threshold, and determining the protection strategy of the distribution line based on the section discrimination criterion and the overcurrent protection criterion of the distribution line; the discrimination threshold is determined after simulating the waveform similarity of the distribution line under different fault conditions based on the distribution network simulation model. The use of this method can improve the safety of the distribution line.

The invention discloses an intelligent prediction model implementation method for deformation of an adjacent existing sensitive structure, and the method comprises the steps: S1, updating construction information and ground wall deflection information according to different prediction tasks; s2, the construction information and the ground wall deflection information are preprocessed; s3, constructing an intelligent prediction model based on a long short-term memory network and a convolutional neural network; and S4, inputting the data information into the intelligent prediction model so as to generate a prediction result of the ground wall deflection curve. According to the invention, continuous prediction of a plurality of future time steps is realized, and the prediction precision and stability are improved.

The invention relates to an intelligent omni-directional logistics vehicle based on multi-source heterogeneous information fusion and a control method. The logistics vehicle comprises a logistics vehicle body, an integrated sensor box module, a visual module, a control module and a driving module, wherein the integrated sensor box module, the visual module, the control module and the driving module are arranged on the logistics vehicle body. The integrated sensor box module, the vision module, the control module and the driving module are all in communication connection with the control module; the control module obtains image information collected by the monocular camera and the depth camera for mapping, plans a driving route and finds an optimal solution of a cargo circulation path; the control module obtains data of the left sensor integration box and the right sensor integration box for integration analysis, judges the conditions of obstacles on two wings and in front of the side, and assists camera information in obstacle avoidance. And the control module acquires data of the inertial measurement unit, and performs visual SLAM (Simultaneous Localization and Mapping) processing by fusing image information. Compared with the prior art, high-precision mapping is carried out, goods are dispatched efficiently, and the obstacle avoidance and path planning capability of the logistics vehicle is improved.

The invention discloses an enterprise humanized management optimization method and system based on machine learning. The method comprises the steps of data classification and collection, employee satisfaction prediction, employee loss risk prediction, employee personalized development prediction and enterprise humanized management optimization. The invention relates to the technical field of enterprise management machine learning, in particular to an enterprise humanized management optimization method and system based on machine learning, and the scheme adopts three predictions of employee satisfaction prediction, loss risk prediction and personalized development suggestion prediction, and improves the data dimension and result category of enterprise personalized management; employee satisfaction prediction is carried out by adopting a graph convolution gating circulation unit network combining dimension graph generation and three demand dimension data parallelism; employee loss risk prediction is carried out by using a bidirectional long-short-term memory neural network combined with a bidirectional encoder; and a graph convolution gating circulation unit network of a bidirectional encoder is adopted to carry out employee personalized development prediction.

The invention relates to an eye movement recognition method based on Transform global semantic coding, and relates to the field of eye movement recognition. Comprising the steps of collecting corresponding eye movement data according to different task requirements, and constructing an eye movement recognition data set; the eye movement data set comprises an image and eye movement data corresponding to the watched image; according to the fixation point position recorded in the eye movement data, performing fixation point mask processing on the related image, and taking the image subjected to mask processing and the image not subjected to mask processing as the input of a Transform model; the method comprises the following steps: carrying out pre-training on a Transform model; extracting deep features from all fixation point positions of each image by using a pre-trained Transform model, and predicting an image-based category probability based on the deep features; and integrating the category probabilities based on the images to obtain a comprehensive classification result of the subjects. The problems that an existing eye movement recognition method is insufficient in deep semantic information mining and cannot capture the global dependency relationship are effectively solved.

This invention discloses an environmental health analysis method and system based on egg-laying hen farming, relating to the field of livestock environmental monitoring technology, including methods based on NH3 concentration and VOCs. S Concentration and NO X Two-dimensional empirical wavelet transform is performed on the concentration to calculate texture intensity, gradient direction, and local entropy. Spatial texture features are obtained by horizontal stacking based on NH3 concentration, HRV standard deviation, and VOC. S Concentration and NO X Concentration is calculated, time-domain features are calculated, and frequency-domain features are calculated based on heart rate and HRV standard deviation. Weighted summation yields the SPS joint features. This invention improves data dimensionality and monitoring comprehensiveness by collecting multimodal data and calculating spatial texture features, thereby enhancing the accuracy of environmental dynamic analysis. Through SPCA dimensionality reduction and weighted risk scoring models, it provides accurate quantitative evidence for the environmental health status of laying hens, improving the health management level and production efficiency of laying hen farming.

The present invention provides a method for generating an interpretable language for automobile abnormality detection results, comprising: acquiring abnormality detection database data, abnormal event description data, abnormality detection cause explanation and vehicle body signal real-time data and analyzing and processing them, obtaining abnormal data description, result description, cause description, state description and abnormal ID and inputting them into a preset first model for training, and obtaining abnormal context information through the trained first model; inputting first question information for the abnormality into the trained first model to obtain first context information of the first question; inputting the first question information and the first context information into a second model, training the second model, and obtaining first explanation information of the first question through the trained second model; associating and storing the abnormality, the abnormality ID and the first explanation information; receiving the second question information; the second question information includes the abnormality ID; and querying and obtaining target explanation information of the abnormality ID according to the abnormality ID.

The invention relates to the technical field of enterprise management, in particular to an enterprise portrait classification system based on a block chain technology, and the system comprises a distribution module which is used for creating block chain distributed digital identities and distributing the digital identities to enterprise users; the interaction platform is used for verifying the digital identity information of the enterprise user, allowing the enterprise user to log in and requesting the state information of the enterprise user; the request module is used for controlling the interaction platform to send an enterprise user state information request to an enterprise user logged in the platform; and the encryption transmission module is used for monitoring the enterprise user state information in the request module to solve whether the enterprise user state information is agreed by a corresponding online enterprise user in an interaction platform, and under the agreed state of the enterprise user, the block chain technology is utilized to create a block chain distributed digital identity for the user, and asymmetric encryption and data chaining modes are utilized to improve the encryption efficiency of the user. And data flow records are solidified, and credible and reliable data traceability is realized. And a solid foundation is laid for improving the data quality of the system.

The invention provides a file uploading method and device, a terminal and a storage medium. The method is applied to the terminal and comprises the following steps: in response to a touch event of a user on a screen of the terminal, determining a target window corresponding to the touch event from at least one window managed by the terminal; determining a target view element corresponding to the touch event from at least one view element of the target window; obtaining an identifier of a target application program corresponding to the target window, and obtaining an identifier of a target view element; recording the touch event, an identifier of a target application program corresponding to the target window and an identifier of the target view element in a pre-created log file in an associated manner; and uploading the log file to a server corresponding to the terminal. Based on the scheme, the touch event, the identifier of the target application program corresponding to the target window and the identifier of the target view element can be uploaded to the server in the form of the log file, and meanwhile, the effects of reducing the development workload and improving the terminal performance are achieved.

The invention discloses an intelligent social chat method based on a large model, and the method comprises the following steps: 1, collecting chat data, preference data and feedback data of a user, and forming training data; 2, inputting the training data into a recurrent neural network machine learning model for training to obtain a chat assistant model; 3, after the user receives the chat content, the chat assistant model generates a plurality of candidate reply contents according to the chat content and the context; and step 4, after the user selects the corresponding reply content, adding the chat content and the selected reply content to the chat data. Through the deep learning and natural language processing technology, the language habit and emotional state of the user can be analyzed in real time, so that the system can more accurately grasp the intention and demand of the user, thereby generating a reply better meeting the expectation of the user, improving the chatting efficiency, and enhancing the emotional connection between the user and the social contact object.

The utility model discloses a magnesia carbon brick performance test bench, which is characterized in that a magnesia carbon brick is a magnesia carbon brick body; comprising a first testing box, a second fixing block arranged above the first testing box, moving mechanisms fixedly connected to the two sides of the second fixing block, a lifting mechanism fixedly connected to the upper end of the second fixing block, a bearing mechanism fixedly connected to the lifting end of the lifting mechanism and a magnesia carbon brick body placed in the bearing mechanism. The acid-alkali liquid soaking mechanism is arranged outside the first testing box, the pressure mechanism is arranged on the side wall of the acid-alkali liquid soaking mechanism, and a camera is fixedly connected to the side wall of the second fixing block. The utility model has the advantages of integration and high efficiency.

The embodiments of the present application disclose a method, device, medium and electronic device for health monitoring of protective facilities. The method includes: feature matching the target image monitoring data of the protective facilities to be predicted with the candidate image monitoring data in the protective facilities corrosion sample database, and selecting similar image monitoring data for the target image monitoring data from the candidate image monitoring data; based on the association relationship between the image monitoring data and the sensor monitoring data, selecting at least two sensor monitoring data associated with the similar image monitoring data from the protective facilities corrosion sample database as target sensor monitoring data; using the target image monitoring data and the target sensor monitoring data as input data of the protective facilities remaining life prediction model to obtain the remaining life of the protective facilities to be predicted. Executing the technical solution provided in the embodiments of the present application can improve the accuracy of the remaining life prediction of protective facilities.

The present invention discloses a method and system for predicting e-commerce consumer preferences based on machine learning, and the method includes data collection, data reconstruction, data preprocessing, consumer preference inference and e-commerce consumer preference prediction. The present invention belongs to the field of e-commerce analysis technology, and specifically refers to a method and system for predicting e-commerce consumer preferences based on machine learning. The scheme connects text comment data and reconstructs heterogeneous data, thereby realizing simultaneous processing of numerical purchase record data and text comment data, and improving the data dimension of consumer preference prediction; adopts text filtering and word segmentation processing methods to preprocess text, and optimizes the data representation of text through word vector conversion, thereby improving data quality and providing good data support for subsequent modeling and prediction; adopts a multi-view latent Dirichlet allocation model to infer consumer preferences, thereby improving the interpretability of the model and thus improving the feasibility and effectiveness of analyzing consumer purchase motivations.

The invention relates to the technical field of artificial intelligence, and provides a multi-scale correlation time sequence prediction method, system, device, product and medium, and the method comprises the steps: obtaining an input sequence, splitting the input sequence to obtain a plurality of single-dimensional input sequences, and carrying out the calculation to obtain a stationary time sequence; obtaining a position coding enhancement vector, performing cutting to obtain a plurality of slice time sequences, and obtaining a time sequence vector through the slice time sequences and the position coding enhancement vector; constructing a time encoder of a multi-head attention mechanism to obtain a plurality of time encoding sequences; constructing a stage conversion layer to obtain a plurality of feature time sequences; constructing a variable encoder of a multi-head attention mechanism, and calculating to obtain a plurality of variable encoding sequences; and obtaining a plurality of single-dimensional prediction results through recombination and expansion projection processing, obtaining a plurality of target single-dimensional prediction results through the plurality of single-dimensional prediction results, and fusing the target single-dimensional prediction results to obtain a target prediction result. According to the invention, the precision of time sequence prediction is effectively improved.

The invention discloses a photovoltaic prediction method and system based on historical data, and relates to the technical field of photovoltaic power generation, and the method comprises the steps: obtaining photovoltaic data, grid point meteorological data and station meteorological data; inputting the grid point meteorological data into a weather discriminator to obtain a weather feature sequence; inputting the photovoltaic data and the weather feature sequence into a tower-shaped wavelet feature prediction network to obtain a grid point photovoltaic prediction feature matrix; fusing the grid point meteorological data with the station meteorological data to obtain a grid station data fusion feature matrix; and combining the grid point photovoltaic prediction feature matrix with the grid station data fusion feature matrix to obtain a three-dimensional grid point prediction result. According to the method, the photovoltaic grid point data is regarded as a multi-dimensional time sequence to be subjected to refined decomposition and prediction, the complex characteristics of photovoltaic resource changes at different spatial positions are effectively captured, multi-source grid point data and station data in historical data are fused, and more comprehensive and accurate prediction analysis of photovoltaic power generation output is achieved.

The invention discloses a control strategy sample generation method, device and equipment and a readable storage medium, and the method can obtain a system operation point, and a strategy agent and an adversarial agent which are subjected to interactive training, and the adversarial agent is obtained through training by taking the generation probability of a control extreme operation point as a target; processing the system operation point by using a confrontation agent to generate a confrontation operation state; and inputting the confrontation operation state into a strategy agent, generating a control strategy matched with the confrontation operation state, and forming a control strategy sample by the confrontation operation state and the control strategy. On the basis, various confrontation operation states and even extreme operation states can be generated by utilizing the confrontation intelligent agent, and the strategy intelligent agent is guided to provide matched control strategies for different operation states, so that the extreme property and the universality of the confrontation operation states and the matched control strategies are ensured. Therefore, the method can generate various control strategy samples which guarantee the extreme performance and the universality.