371 results about "Data decomposition" patented technology

The invention discloses a power grid power load intelligent prediction method, system and device and a medium, and relates to the technical field of power distribution network transmission optimization. Power load data is decomposed into a trend component, a periodic component and a random fluctuation component, and then the importance of different components is evaluated by using a feature channel attention layer; the time sequence attention layer captures key moments in each component time sequence and extracts features at the key moments, and then the features extracted based on the importance and the features extracted at the key moments are fused, so that the multi-scale features of the power load are captured; and then inputting the fused features into the dynamic gating residual connection LSTM network for prediction, and in the prediction process, improving the attention degrees of long-term sequences, short-term sequences and fluctuation sequences in different features through a residual modulation function, thereby more accurately capturing the multi-scale features of the power load and obtaining a prediction value of the power load.

The present invention relates to urban rainfall runoff pollution prediction in urban water systems, and provides an urban rainfall runoff pollution prediction method based on integrated rolling decomposition method and deep learning algorithm. A rolling decomposition method is firstly used to decompose rainfall runoff sequence data into different sub-sequences; then decomposition is sequentially performed on added data, and future data is excluded, to prevent information leakage; a recurrent neural network is used to model and predict the sub-sequences; and finally, predicted results of the sub-sequences are summed to obtain the predicted result of rainfall runoff pollution.

The invention relates to the technical field of underground cavern surrounding rock displacement prediction, in particular to a high-stress large underground cavern group surrounding rock displacement prediction method. According to the technical scheme, the method comprises the following steps that data collection and arrangement are carried out, and various types of sensors are arranged at different parts of a high-stress large underground cavern group to collect data; the method comprises the following steps: performing preliminary screening, cleaning and normalization processing on collected data, performing feature engineering and data enhancement work, including time-frequency domain feature extraction, wavelet transform-based data decomposition and reconstruction and a data enhancement strategy, and constructing a hybrid neural network model fusing a convolutional neural network and a gating circulation unit. Through multi-source accurate data acquisition and processing, depth feature engineering, hybrid neural network construction training and scientific evaluation and early warning, the accuracy, reliability and timeliness of high-stress large underground cavern group surrounding rock displacement prediction are remarkably improved, powerful support is provided for engineering safety guarantee, and risks and losses are effectively reduced.

The invention discloses a wind power multi-scale decomposition prediction method. At present, single-point prediction is not comprehensive and accurate enough, and cannot adapt to quantitative accurate requirements of a wind power plant and a power grid dispatching mechanism in risk management. The method comprises the following steps of: forming an original wind power sequence from actually acquired wind power data, sequentially performing feature selection and data decomposition processing to form multi-scale modal data, and constructing a depth prediction model according to the multi-scale modal data; a probability prediction interval determination process is completed in the residual error distribution mode depth prediction model through adaptive bandwidth kernel density estimation; after actually obtained wind power data form an original wind power sequence, an initial model is established, feature selection processing is performed on the initial model, that is, weighted marginal contribution is calculated for each feature of the initial model according to all involved feature subsets by using an SHAP algorithm based on a Shapley value in a game theory, and the weighted marginal contribution of each feature of the initial model is calculated; and completing a feature data acquisition process of accurately quantifying interdependence and interaction effect between features.

The invention relates to the technical field of information platform construction, in particular to a construction method and system of a sports information intelligent service platform. The method comprises the following steps: collecting user motion image data and user health basic data; key body node position information in the movement process of the user is extracted; generating a user motion track curve based on the key node position information, and calculating a deviation angle and a deviation distance with a standard track template to obtain motion correctness data; constructing a user fitness risk assessment model according to the action correctness data and the user health basic data, analyzing the difference degree with a preset safety threshold, and generating layered health guidance data; and decomposing the layered health guidance data into quantifiable progress units, and generating progress tracking data. According to the method, the user motion image data and the health basic data are combined, the evaluation model capable of identifying individualized risk factors is constructed, and more accurate safety guidance is realized.

The invention discloses a multi-modal data fused shield tunneling machine cutter state fault prediction system, which comprises a state acquisition module for acquiring multi-modal state data, decomposing the multi-modal state data into time and space related characteristics, and obtaining optimal state data through low-rank matrix decomposition and compression; the anomaly simulation module is used for collecting fault state data, constructing an improved lightweight weight migration network to perform cross-domain feature extraction to obtain a cross-domain common feature vector, and obtaining common features through a feature alignment algorithm based on adversarial training on the basis of the cross-domain common feature vector and the fault state data; and the fault prediction module is used for performing fault prediction through a space-time cross attention dynamic migration network on the basis of the obtained optimal state data and common characteristics, so that long sequence characteristics can be captured when the cutter is abnormal, the calculation amount can be reduced when the cutter is normal, and the real-time performance and the accuracy of fault prediction are improved.

The invention relates to the technical field of data security detection, in particular to a data security detection method for signal transmission software. And obtaining residual time sequence data after decomposition of the plurality of signal-to-noise ratio time sequence data. Because the electromagnetic interference can cause the residual points to present trend characteristics, the value distribution and change trend of the residual points are analyzed, and outliers and trend characteristic values are obtained. The distribution of the outliers subjected to electromagnetic interference is random, so that a local distribution characteristic value is obtained based on the position distribution condition of the outliers. And combining the two indexes to obtain an electromagnetic attribute characteristic value. Furthermore, as the abnormity generated by the network attack is more correlated, the correlation condition of the outliers in the residual time sequence data of the plurality of channels is analyzed and combined with the electromagnetic attribute characteristic value to obtain an electromagnetic interference degree value, and finally, the outliers generated by electromagnetic interference are removed according to the electromagnetic interference degree value to obtain network abnormal points. The detection precision of the network abnormal points is improved, and the safety detection effect is ensured.

A system and method for compression performs analysis of incoming audio or video data, and selects a manifold based on the analysis of the data. A deep learning model is then trained for the manifold. The data is broken down into components and entropy maximization algorithms are utilized for each component before compression commences. Finally, the system translates the compressed data into a standard file format.

The invention relates to the technical field of water conservancy data processing, in particular to a GIS-based water conservancy project management system, which comprises a non-stationary data decomposition module, a nested autoregression modeling module, a space-time dynamic reconstruction module, a risk quantitative mapping module and a risk decision optimization module. According to the method, long-term changes and short-term fluctuations are quantified and synthesized term by term through hydrological time series data in trend fitting and fluctuation amplitude extraction, the precision and timeliness of time series prediction are improved, dynamic integration of data of multiple monitoring points is achieved in combination with geographic information matching and spatial difference value distribution analysis, and the accuracy and timeliness of time series prediction are improved. According to the method, the spatial distribution of water resources is accurately predicted, the correlation degree between regions is quantified, risk index quantification is combined with a spatial interpolation technology, so that risk region division is more detailed, the pertinence of emergency management is optimized, and the scientificity and execution efficiency of risk decision are further improved based on scheduling data analysis and influence factor quantification.

The invention discloses a multi-modal information interface complexity evaluation method and system, and the method comprises the steps: collecting multi-modal interface elements which comprise visual data, auditory data and interaction data; decomposing the interface complexity of the multi-modal interface, and dividing the interface complexity into information complexity, visual complexity, auditory complexity and cognitive complexity; respectively calculating an index of information complexity, an index of visual complexity, an index of auditory complexity and an index of cognitive complexity; respectively carrying out normalization processing on all the calculated indexes; allocating a weight to each index; and comprehensive score calculation: calculating a comprehensive score through the normalized index value and the weight corresponding to each index. According to the method, the multi-modal complexity is decomposed into four dimensions of information, vision, auditory sense and cognition, comprehensive analysis is carried out on multi-modal information fusion, a dynamic optimization process is embedded, the evaluation efficiency and the interface availability are remarkably improved, and the practicability is better.

The invention relates to the technical field of data center dynamic modeling, and discloses a twin model dynamic modeling method for a data center DCIM platform, which comprises the following steps: constructing a discrete state space model containing a thermal coupling matrix and a system matrix, collecting real-time power and temperature time sequence data, and calculating a cross-correlation function to lock hot air dynamic transmission lag time; calculating cut-off frequency based on physical attributes of the cabinet and decomposing data into high and low frequency components by using a complementary filter; according to the method, the model parameters are made to return to a physical source through a frequency domain decoupling mechanism, the problem of aliasing of airflow coupling and structural thermal inertia parameters in a traditional single-scale identification method is solved, and the method is suitable for large-scale identification. And the physical authenticity and prediction robustness of the twin model under a complex working condition are improved.

A system and method for compression performs analysis of incoming audio or video data, and selects a manifold based on the analysis of the data. A deep learning model is then trained for the manifold. The data is broken down into components and entropy maximization algorithms are utilized for each component before compression commences. Finally, the system translates the compressed data into a standard file format.

The invention discloses a dance movement digital acquisition method based on motion capture and a teaching system, and solves the problems of subjective evaluation of traditional teaching, neglect of posture difference in the prior art and the like. The motion capture module collects posture and motion data of a dance actor, decomposes the data into basic motions and extracts features to form a standard sequence, generates a personalized teaching evaluation sequence in combination with a student posture feature difference, collects student motions in real time and compares the student motions with the standard, and generates a targeted teaching strategy according to the difference. The system comprises a cloud data server, a motion capture module and the like, and can store standard data, analyze motion characteristics and perform interactive feedback. According to the invention, through accurate capture, personalized evaluation and real-time feedback, the teaching accuracy and efficiency are improved; and data support is provided for dance innovation.

The invention discloses a multi-scale space-time fusion water quality prediction and anti-counterfeiting method based on a dynamic graph neural network. Comprising the following steps: 1) collecting water quality index hour data of a plurality of monitoring stations in a drainage basin; 2) decomposing the data into a plurality of intrinsic mode functions through variational mode decomposition; 3) constructing a dynamic graph neural network spatial feature extraction module, and generating a discrete dynamic graph structure; 4) constructing a multi-scale time feature extraction module, and synchronously capturing short-term fluctuation and long-term trend; 5) designing a residual fusion mechanism to integrate the spatial-temporal characteristics, and outputting a water quality prediction result through a full connection layer; and 6) calculating a path distance between the input data and a prediction result through a dynamic time warping algorithm, and comparing residual distribution by combining K-S to realize authenticity discrimination of the input data. The method can fully excavate the spatial and temporal characteristics of the basin water quality under the condition that the geographical spatial distribution of the sites is unknown, improves the prediction precision, carries out the authenticity recognition of the water quality data of an unknown source, and prevents the data from being tampered.

The invention relates to the technical field of natural language processing and deep learning, and relates to a self-speculation decoding method and system based on layered quantization KV cache. The method comprises the following steps: performing hierarchical quantization processing on KV cache of a target model, and decomposing original cache data into a high-bit quantization part and a low-bit quantization part; the draft model generates a candidate token sequence based on the KV cache of the high-order quantization part, and stores the candidate token sequence into an asynchronous queue; the target model obtains a candidate token sequence, combines the high-order quantization part and the low-order quantization part into a full-precision cache, and performs parallel verification on the candidate token sequence; and receiving the valid candidate token or regenerating a replacement token by the target model, and updating the hierarchical quantization KV cache. By optimizing the management and quantification strategy of the KV cache, the reasoning process of the large-scale language model is remarkably accelerated, and meanwhile the generation quality is kept.

The invention relates to the technical field of mine safety, in particular to a rock burst prediction method based on dynamic frequency domain decomposition, relative event coding and dynamic sparse attention. According to the method, data decomposition of Fourier transform, adaptive frequency domain noise reduction, relative event coding and a dynamic sparse attention mechanism are combined, key features in micro-seismic data can be effectively extracted, and the accuracy and robustness of rock burst prediction are improved. The method is excellent in performance in rock burst prediction, the problems of noise interference, error accumulation and the like can be effectively solved, and reliable technical support is provided for accurate early warning of rock burst.

The invention discloses a security knowledge graph construction method and system based on behavior tracks, and relates to the technical field of network security. The method comprises the following steps: carrying out operation link aggregation according to user operation records to obtain a standardized behavior track chain; calling and accessing heterogeneous behavior trajectory data by taking the standardized behavior trajectory chain as a constraint to obtain multi-source behavior trajectory data; decomposing the standardized behavior track chain to obtain a link node time sequence operation relation and M link nodes; performing time sequence dependence modeling based on the graph convolutional network to generate a security knowledge graph; decomposing the multi-source behavior track data according to the M link nodes to obtain M pieces of multi-source behavior node data; and taking the M multi-source behavior node data as training data, and executing local parameter adaptive optimization of the security knowledge graph through a meta-learning framework. The technical problem of insufficient adaptability of the knowledge security graph in the prior art is solved, and the technical effect of dynamic adaptive optimization of the security knowledge graph is realized.

The invention provides a fall detection algorithm model for a wearable device of a stroke patient, and relates to the technical field of health monitoring and posture detection, and the technical key points are as follows: a construction method of the model is as follows: S1: obtaining posture information of a wearer by using an inertial sensor IMU in the wearable device, after data collection is completed, performing low-pass filtering processing on original data to obtain sensor data; s2, decomposing the inertial sensor data obtained at the t moment into a trend component, a season component and a residual component by using LOESS according to an STL (Standard Template Library) cyclic trend decomposition method, and then carrying out normalization processing on the data; and S3, performing time sequence modeling on each inertial sensor component by using LSTM (Long Short Term Memory). According to the fall detection algorithm model for the wearable device of the stroke patient, collaborative optimization of multi-physical-quantity coupling feature decoupling and time sequence dynamic modeling is realized, and medical-level fall monitoring performance is realized under resource constraints of the wearable device.

The invention discloses a wave height short-term prediction system and method based on variational mode decomposition and wind and wave time correlation, and belongs to the technical field of ocean engineering.The method comprises the steps that firstly, original storm time series data are decomposed into IMF components with different center frequencies through variational mode decomposition; secondly, performing data cleaning on the decomposed IMF component based on a Pearson's correlation coefficient; then combining two wind speed IMF components with the strongest correlation with the wave height IMF component based on correlation analysis to form a wind speed wave height component combination, and determining an input time length through an autocorrelation coefficient ACF of wave height data and a cross correlation coefficient CCF of the wind speed IMF components and the wave height IMF components; and finally, according to the formed wind speed wave height component combination and the determined input time length, carrying out short-term prediction on the wave height based on a long short-term memory neural network. According to the system and the method provided by the invention, the influence of noise on prediction is reduced, and the prediction precision is improved while the prediction efficiency is improved.

The invention discloses a thermal load prediction method and system based on multi-stage fusion optimization, and the method employs a gray wolf optimization algorithm to replace a conventional empirical method to optimize key parameters of variational mode decomposition, and automatically determines an optimal mode number and a penalty factor through multi-objective optimization. The obtained parameter combination is introduced into a variational mode decomposition algorithm, and then the variational mode decomposition algorithm is utilized to decompose thermal load historical data into a plurality of stable mode components, so that the complexity and noise interference of the data are reduced, and the decomposition quality of variational mode decomposition is remarkably improved; then, the hyper-parameters of the long-short-term memory network are optimized by combining a sparrow search algorithm, so that an optimal hyper-parameter combination is found, and finally, a thermal load data set is predicted through a long-short-term memory model under the obtained optimal hyper-parameter combination.

The invention belongs to the technical field of spaceflight measurement and control and artificial intelligence, and particularly relates to a high-precision low-orbit satellite forecasting method and system, and the method comprises the steps: obtaining the observation data of a low-orbit satellite, and the parameter data of the earth center gravitational force and non-spherical gravitational force perturbation; the method comprises the following steps: decomposing observation data of a low-orbit satellite into long-term trend data, medium-term fluctuation data and short-term change data; constructing a long-term dynamic model, and fitting the long-term trend data and the long-term dynamic model to obtain a long-term fitting result; constructing a medium-term dynamic model, and fitting the medium-term fluctuation data and the long-term dynamic model to obtain a medium-term fitting result; constructing a short-term dynamic model, and fitting the short-term change data with the short-term dynamic model to obtain a short-term fitting result; fusing the fitting results to obtain a satellite orbit state fitting result; and updating parameters of all the dynamic models, and calculating orbit forecast information of the low-orbit satellite. And the orbit forecasting precision of the low-orbit satellite is effectively improved.

The invention discloses a power carbon emission measurement and calculation data anomaly identification method and system. The method comprises the steps of obtaining power carbon emission original data and performing preprocessing; the method comprises the following steps: decomposing data into a plurality of internal mode functions and residual sequences by adopting an adaptive variable scale weighted noise enhancement CEEMDAN method, and effectively separating a time sequence trend from high-frequency fluctuation; constructing an improved box plot model based on the residual sequence, and adaptively adjusting an abnormal boundary by introducing a median deviation coefficient and a dynamic abnormal detection coefficient; and identifying abnormal data points by using the optimized boundary, and performing correction or elimination processing. According to the method, the problem that a traditional method is poor in adaptability to non-stationary and multi-scale electrical carbon data is solved, normal periodic fluctuation is prevented from being misjudged to be abnormal, the accuracy and robustness of anomaly detection are remarkably improved, and a reliable data basis is provided for accurate measurement and calculation of carbon emission of an electric power system.

The invention discloses an energy storage system fault prediction method and system based on multi-modal data fusion, and relates to the technical field of energy storage systems. Comprising the following steps: acquiring multi-modal data, and preprocessing the data to improve the data quality; performing feature extraction on the multi-modal data through a feature extraction model to obtain a feature matrix; performing factorization on the extracted feature matrix, establishing a multi-modal data fusion model, fusing core factor matrixes obtained after decomposition, establishing a prediction model, substituting fused feature vectors for fault prediction, and outputting a prediction result. According to the method, factorization is carried out on the extracted feature matrix through a data decomposition method, the feature decomposition matrix is obtained, core factor matrixes obtained after decomposition are fused through the multi-modal data fusion model, the fused feature vector can further express the mutual relation between multi-modal numerical values, and the fusion efficiency is improved. And the fault prediction accuracy of the energy storage system is improved.

The invention relates to the technical field of photovoltaic power prediction, in particular to a photovoltaic power parallel prediction method, which comprises the following steps: acquiring a photovoltaic power generation power sequence according to a preset time interval, and performing data decomposition on the photovoltaic power generation power sequence based on an REMD-OVMD model to obtain a plurality of intrinsic mode functions (IMF); taking a plurality of IMFs as subsequences, and performing feature extraction on each subsequence in a channel dimension and a time dimension by using a time sequence prediction model FreTS based on FreMLP; based on the fused features of all the subsequences, photovoltaic power prediction is carried out by using a pre-constructed Dline-Transformer parallel structure prediction model, and the sum of photovoltaic power prediction values corresponding to all the subsequences is used as a photovoltaic power prediction result; and learning and predicting an error sequence of a photovoltaic power prediction result by using a TCN model, optimizing network parameters through back propagation, and finally obtaining a corrected photovoltaic power prediction sequence. According to the method, a parallel prediction method is adopted, and a plurality of sub-models can be operated at the same time, so that the calculation time is shortened, and the real-time performance is improved.

The invention discloses a method, a system and a related device for controlling a motor based on an adaptive algorithm, which are used for improving the accuracy of motor control. The method comprises the following steps: acquiring initial data of a motor; decomposing the initial data into intrinsic mode function components, and performing time-frequency transformation on the intrinsic mode function components; dynamic time-frequency features of the intrinsic mode function component are extracted, cross-band weighted fusion is carried out on the dynamic time-frequency features through a preset attention weighting mechanism, and a multi-dimensional feature vector is generated; inputting the multi-dimensional feature vector into a preset deep learning model, and constructing a target prediction model in combination with a loss function and an adaptive momentum optimization algorithm; transmitting the target prediction model to an analysis module of the motor to obtain a load change trend and a confidence coefficient; dynamically adjusting the proportion of the weight of the PID controller according to the load change trend and the confidence coefficient, and optimizing the parameters of the unit of which the proportion is adjusted through a gradient descent method; and dynamically regulating and controlling the motor based on the PID controller.

The invention discloses a reservoir flood forecasting method based on physical constraint and space-time double-flow coupling. The reservoir flood forecasting method comprises the following steps: S1, acquiring multi-source hydrometeorological data; s2, decomposing the multi-source hydro meteorological data into a historical state sequence and a future driving sequence; s3, performing feature extraction on the historical state sequence data through the physical enhanced long-short term memory network of the historical inertial feature extraction branch to obtain historical inertial features, and performing feature extraction on a future driving sequence through the time domain convolutional network of the future forced feature extraction branch to obtain future forced features; s4, performing weighted fusion on the historical inertial features and the future forced features to generate fusion features; and S5, inputting the fused features into a decoder to obtain a predicted water level increment, and superposing the predicted water level increment to the current water level to obtain a predicted value. The prediction timeliness is improved, the prediction precision of the water recession stage is improved, and the physical consistency of the prediction result is enhanced.

The invention discloses a multi-element hydrological information prediction method and system. The method comprises the following steps: acquiring water level, flow velocity, flow and temperature time sequence data of a target water area through a high-frequency acoustic tomography system, and constructing a multivariate hydrological data set containing sampling duration and hydrological parameter vectors at all moments; preprocessing the collected multivariate hydrological data set to generate preprocessed data, and extracting timestamp features of a sequence; and decomposing the preprocessed data into a plurality of intrinsic mode functions and the like by adopting a variational mode decomposition method. The problems of gradient disappearance and local dependence of a traditional RNN model are solved, redundant noise interference is effectively eliminated while the multi-time-scale characteristic of hydrological variables is reserved, the prediction generalization ability under the complex coupling relation is improved, and high-precision technical support is provided for dynamic planning of water resources and disaster early warning.

The invention relates to the field of computer systems based on specific calculation models, and discloses an intelligent water meter data driving leakage identification early warning system based on a neural network model, comprising a learnable decoupling module, parameters of which are determined by a collaborative adversarial training rule, and which is used for decomposing data into a baseline data stream and a transient data stream; the system is provided with a baseline channel and a transient channel in parallel; according to the system and the method, a behavior mode monitor, a transient mode analyzer and an abnormal arbitration module are arranged, through information cooperation of the three modules, accurate identification of different leakage states and normal mode drifting is achieved, through the architecture, the fundamental problem that a calculation model is shielded by a high-amplitude normal water use signal is solved, and the system and the method have the advantages of being high in practicability and high in practicability. According to the method, high sensitivity of the system to tiny baseline leakage is recovered, and meanwhile, active adaptation to user habit changes and effective compensation for a high-frequency leakage blind area are achieved through cross-channel information collaboration.

The invention discloses a landslide disaster early warning method based on an LSTM-SARIMA mixed data driving model. Global and local noise elimination is performed on radar displacement monitoring data by adopting a two-stage noise reduction method combining moving average and wavelet transform, so that the signal-to-noise ratio is effectively improved; the method comprises the following steps: decomposing slope displacement data into a trend term dominated by gravity and a periodic term influenced by the environment by utilizing a Hodry-Precott (HP) filtering method, and realizing differentiated analysis of a multi-factor action mechanism; a mixed prediction model combining a long short-term memory network (LSTM) and a seasonal autoregressive integrated moving average model (SARIMA) is provided, modeling and prediction are carried out on decomposed displacement components respectively, and results are fused to obtain a high-precision total displacement prediction value; an improved T-t curve is constructed based on a predicted displacement-time curve, a tangent angle index for landslide early warning is proposed according to a Saito three-stage theory, and quantitative and interpretable early warning criterion setting is realized.

The invention discloses a landslide early warning method based on multi-source fusion and an intelligent algorithm. The landslide early warning method comprises the following steps: acquiring meteorological and hydrological monitoring data, remote sensing image data and landslide displacement monitoring data of a target landslide area; performing data decomposition on the landslide displacement monitoring data to obtain a trend term and a periodic term; extracting ground monitoring data composed of a plurality of factors and response time of the plurality of factors from the meteorological and hydrological monitoring data; performing fusion processing on the remote sensing image data and the ground monitoring data to obtain key disaster-causing factors; constructing a prediction model used for predicting the landslide displacement rate; triggering an early warning signal based on the predicted value of the landslide displacement rate; according to the method, the landslide multi-scale response rule and the lag effect can be identified more accurately, and the dynamic prediction and early warning capability of the landslide is improved.