27 results about "Time scale decomposition" patented technology

The invention discloses a compressed air energy storage system operation state evaluation method based on principal component clustering analysis. The method comprises the steps that system operation data are collected and preprocessed; performing multi-time scale decomposition and feature extraction on the preprocessed data set; performing nonlinear time-varying adaptive principal component analysis on the multi-time scale feature set, and extracting main features of a system operation state; carrying out clustering analysis on the dimension reduction feature space by adopting a PCA-clustering iterative optimization mechanism; establishing an evaluation model of each functional area of the system based on thermodynamic constraints, and calculating performance index values of the functional areas in combination with the principal component transformation matrix and an operation mode clustering result; dynamically adjusting the weight of each evaluation index; and generating a CAES comprehensive operation state evaluation result. According to the method, the problems of poor PCA adaptability, difficulty in multi-time scale data fusion, low abnormal mode detection sensitivity and the like under the nonlinear time-varying characteristic of the CAES system are solved.

The invention relates to the technical field of integrated circuit reliability, and discloses an MOS device service life prediction method and system and electronic equipment, and the method comprises the steps: constructing a continuous time dynamic model, and expressing the degradation process of an MOS device; constructing a device relation graph structure, and capturing spatial correlation between devices; executing multi-time scale decomposition, and analyzing time sequence characteristics of device degradation; a self-adaptive numerical solution algorithm is applied to realize degradation track prediction; physical constraint regularization is implemented, and the physical rationality of the model is improved. By fusing the graph neural differential equation and the multi-time scale decomposition technology, the problems of insufficient discrete time modeling, neglect of a synergistic effect between devices, difficulty in multi-scale characteristic processing and the like in the prior art are solved, and high-precision prediction of the service life of the MOS device is realized.

The invention relates to a train delay prediction method based on multiple time scales, and the method comprises the steps: carrying out the multi-source data fusion of train operation data, and obtaining the space-time embedding representation of a data tensor; wherein the space-time embedding representation comprises time embedding representation, space embedding representation and feature embedding representation; identifying periodicity in a time sequence for the space-time embedding representation through fast Fourier transform, and dividing information of a plurality of time scales; capturing space-time correlation for the space-time embedded representation, and generating space-time correlation feature representation; based on the information of multiple time scales, performing multi-time scale decomposition on the space-time correlation feature representation, and performing channel attention learning on the decomposed information to obtain aggregated multi-time scale sequence information; and based on the aggregated multi-time scale sequence information, obtaining a late prediction result. According to the method, accurate delay prediction of the train under the multi-station and complex railway network conditions can be realized.

The invention discloses an adaptive graph learning traffic flow prediction system and method based on multi-dimensional time, and belongs to the technical field of traffic systems, the system is formed by stacking a plurality of space-time convolution blocks through residual connection, and each block comprises a multi-dimensional time scale feature fusion module, a mask matrix adaptive graph learning module and a gating time sequence convolution module. Wherein the multi-dimensional time scale feature fusion module extracts hour-day-week time scale features and splices the features; the mask matrix adaptive graph learning module constructs an adaptive graph learning structure in combination with the space-time dynamic incidence matrix and the anti-symmetric structure mask matrix; and the gating time sequence convolution module extracts time sequence features and outputs a prediction result. According to the prediction method, time scale decomposition, attention score calculation and other operations are performed on input data, a result is obtained through gating operation, and stable transmission of features is ensured through residual connection. Through deep decoupling and dynamic fusion of spatio-temporal features, prediction precision and robustness are improved, and the method has high practical application value and industrial popularization potential.

The invention discloses an intelligent new energy power data monitoring analysis method and system, and relates to the technical field of power systems, and the method comprises the steps: collecting the time sequence data of a plurality of monitoring parameters of new energy power generation equipment, and determining the normal operation state region of each monitoring parameter; performing multi-time-scale decomposition on the time sequence data, and extracting a fast-varying component and a slow-varying component of each monitoring parameter; constructing a transient coupling matrix, a trend incidence matrix and a cross-scale conduction matrix, and performing adaptive fusion to obtain an adaptive coupling matrix; establishing a parameter evolution model according to the adaptive coupling matrix, and generating an evolution trajectory of each monitoring parameter; and comparing the evolution trajectory with a normal operation state area, when deviation occurs, tracing abnormal source parameters according to the adaptive coupling matrix, and outputting a monitoring analysis result. According to the method, signal features of multiple time scales can be captured at the same time, the dynamic coupling relation among multiple parameters is accurately described, and self-adaptive monitoring analysis and abnormal source rapid positioning are achieved.

The invention discloses a prefabricated part quality optimization method and system based on digital twinning, and relates to the technical field of digital twinning. The method comprises the following steps: constructing a sample performance data set; determining a plurality of prediction time scales of the component, and recombining the data set to construct a component index prediction agent model group containing the corresponding time scales; and acquiring a real-time digital twinborn model of the component, establishing a design optimization problem in combination with the proxy model group, and updating the digital twinborn model by solving the optimization problem to realize quality optimization. According to the method, through multi-time scale decomposition and cascade knowledge distillation training, the precision and efficiency of long-term performance prediction are improved while the high-fidelity data requirement is reduced, meanwhile, the prediction reliability and the adaptive evolution ability of the model are ensured in combination with multi-model cross check and a trigger type feedback mechanism, and the prediction efficiency is improved. Accurate prediction and reliable lightweight design of the full life cycle performance of the prefabricated part are achieved.

The application provides a direct current fault arc detection method and device, which combines wavelet packet decomposition and reconstruction, intrinsic time scale decomposition and reconstruction and classical modal decomposition and reconstruction three kinds of feature extraction technologies to process an electrical signal sequence to be analyzed, and respectively obtains first, second and third electrical characteristic signal sequences. The method can reveal the fault characteristics of the electrical signal from multiple dimensions, realizes feature fusion, effectively improves the accuracy and robustness of fault detection, reduces false positives and false negatives, has strong adaptability, is suitable for various electrical signal analysis, and provides accurate and reliable technical support for arc fault detection of a direct current power supply system.

The invention discloses a hydrogen energy scheduling network optimization method and system, and belongs to the field of hydrogen energy scheduling, and the method comprises the steps: obtaining the structural parameters of an electrolytic cell, and building a millisecond-level dynamic model of an electrolytic hydrogen production process in combination with electrolyte characteristics and electrode dynamic data; according to the electrolytic hydrogen production dynamic model, a time scale decomposition method is adopted to reduce the dimension of the millisecond-level dynamic model into a second-level simplified model so as to meet the preset high-precision requirement that the relative error is smaller than 1%; thermodynamic parameters are extracted from the characteristics of the hydrogen storage material, and a dynamic response model of the hydrogen storage tank is established in combination with hydrogen charging and discharging dynamic data; and for the hydrogen storage tank dynamic response model, extracting a thermodynamic and dynamic coupling relationship, and separating thermodynamic and dynamic variables by adopting a variable replacement decoupling algorithm to obtain a decoupled hydrogen storage tank model.

The present invention discloses a short-term prediction method for railway traction load, which belongs to the field of power data processing technology. The present invention includes data cleaning, data primary decomposition, data secondary decomposition, sample entropy clustering, model training, and data prediction. The present invention combines the intrinsic time scale decomposition algorithm, the robust local mean decomposition algorithm, sample entropy clustering, the time convolution network, and the bidirectional gated recurrent network to establish a short-term prediction method for traction load based on a combination of secondary decomposition reconstruction and deep learning double-layer optimization. This method can reduce the complexity of high-frequency components. At the same time, it effectively solves the defect that existing research cannot deeply capture the real-time dynamic characteristics of the load, thereby improving the accuracy of traction load prediction.

The application discloses a kind of wind generator bearing fault diagnosis method and system, it is related to wind power generation equipment fault diagnosis technical field.The method includes with specific sensor acquisition bearing vibration data and pre-processing;It is denoised by enhanced combination differential morphological filter ECGMF;Intrinsic time scale decomposition ITD algorithm is used to decompose signal after denoising, extract component instantaneous amplitude and spectrum analysis obtains fault characteristic;Deep support vector machine DSVM model is constructed in combination with support vector machine SVM and neural network;Characteristic vector is input into deep support vector machine DSVM model training and judging fault output result.The system contains vibration data acquisition, enhanced combination differential morphological filtering, intrinsic time scale decomposition, deep support vector machine and diagnostic result output module, corresponding to realize the function of each step above, through each link cooperation, effectively extract fault characteristic and accurately diagnose, can promote the accuracy and efficiency of wind generator bearing fault diagnosis.

The application provides a kind of exoskeleton control method and system based on iterative learning and extended state observer, comprising: obtaining lower limb exoskeleton system parameters and motion state data, establishing nonlinear dynamics equation and unifying total disturbance term, obtaining multi-scale dynamics model by multi-time scale decomposition;Based on the model, an extended state observer is constructed to estimate the system state and total disturbance, and the disturbance is fed back to the control input;A game model between man and machine is constructed, parameters are identified online and an iterative learning control update law is designed to obtain an adaptive iterative learning control law;Design fast and slow dynamic controller, integrate related control information through hierarchical control architecture, and output multi-time scale model predictive control quantity.The application improves the coordination, comfort and energy efficiency ratio of human-computer interaction, and can provide stable and natural assistance experience in various complex environments, meeting the needs of rehabilitation training and daily assistance.

The application discloses a prefabricated component quality optimization method and system based on digital twinning, and relates to the technical field of digital twinning. The method comprises the following steps: constructing a sample performance data set; determining a plurality of prediction time scales of the component, and reorganizing the data set to construct a component index prediction agent model group containing the corresponding time scales; obtaining a real-time digital twinning model of the component, combining the agent model group to establish a design optimization problem, and updating the digital twinning model by solving the optimization problem to realize quality optimization. Through multi-time scale decomposition and cascading knowledge distillation training, the application reduces the demand for high-fidelity data, improves the accuracy and efficiency of long-term performance prediction, combines multi-model cross-validation and trigger feedback mechanism to ensure the prediction reliability and the self-adaptive evolution ability of the model, and realizes accurate prediction and reliable lightweight design of the whole life cycle performance of the prefabricated component.

The invention relates to the technical field of artificial intelligence, in particular to an energy storage capacity configuration method and system based on multi-scale data and adaptive optimization. According to the method, historical new energy output data and load data are collected and cleaned, the data are decomposed into short-term, medium-term and long-term components according to the time scale, the difference between each scale component and a local mean value is calculated, a weight attenuation coefficient is combined, and an adaptive weight coefficient is obtained through exponential function mapping normalization; and scaling the corresponding scale component based on the time window extreme value and combining the adaptive weight coefficient to obtain multi-scale normalized data. And calculating load power shortage power, new energy power abandoning power and an energy storage life loss factor based on the normalized data, and performing weighted combination on the three to construct a comprehensive objective function. An improved whale optimization algorithm is utilized, population initialization is performed through multi-scale data statistical information, control parameters are dynamically updated, a search strategy is selected based on a population distribution state, and accurate optimization configuration of the energy storage capacity is realized.

The invention discloses a cooling and heating load prediction method based on multi-time scale decomposition and fusion, and the method comprises the steps: inputting historical cooling and heating load time sequence data into a multi-level prediction framework, obtaining a plurality of subsequences, obtaining the statistical index data of the subsequences through a self-adaptive mixed decomposition algorithm, and carrying out the prediction of the cooling and heating loads according to the statistical index data. Selecting matched expert prediction models for the subsequences from a preset heterogeneous model library, and obtaining corresponding external features; establishing training constraint conditions of the multi-level prediction framework; and taking minimization of the final reconstruction prediction error as a target function, training a multi-level prediction framework through a deep learning back propagation algorithm, and inputting independent prediction results of the expert prediction model on the corresponding subsequences and external characteristics into the trained framework together to obtain a final reconstruction prediction load value. According to the method, the capability of identifying the abnormal operation condition is improved, and the economical efficiency and stability of regional energy supply are better guaranteed.

The invention relates to the technical field of data mining application, and discloses a multi-time-scale hybrid energy storage capacity configuration method and system and a storage medium. The method comprises the following steps: acquiring data according to a wind power plant construction state and preprocessing to obtain standard wind power output power data; acquiring a local power grid load curve or scheduling demand data, and calculating a difference value to generate an energy storage reference power curve; splitting the power signal by adopting an SGMD decomposition algorithm, and identifying multi-time scale power fluctuation to generate feature data; setting constraints based on the feature data, and optimizing energy storage capacity configuration by using an immune hybrid particle swarm algorithm; and performing full-life-cycle simulation on an optimization result, verifying an effect and compliance, and evaluating economy to determine an optimal scheme. Through multi-time scale decomposition and closed-loop optimization, the problems that existing energy storage configuration is poor in adaptability and insufficient in economical efficiency are solved, and wind power integration stability and configuration economical efficiency are improved.

The application relates to the technical field of power system energy storage and automatic control, and discloses a capacity optimization method, system and device of a multi-time-scale hybrid energy storage system, which comprises the following steps: based on a multi-time-scale decomposition result of new energy power fluctuation, an initial capacity optimization model is established and solved to obtain initial capacity configuration parameters and multi-time-scale division parameters; based on the initial parameters, the hybrid energy storage system is controlled to participate in grid frequency modulation, and operation performance and equipment state data are collected; performance feedback indexes are generated according to the data, and adjustable parameters of the capacity optimization model are dynamically corrected based on the indexes, so that the model is solved again, in the process, the division parameters and the capacity configuration parameters are treated as associated variables to obtain updated parameters; the updated parameters are applied to the system and returned to the frequency modulation operation step to form a closed-loop optimization. The application realizes the collaborative dynamic optimization of energy storage capacity and division strategy, and improves the economy, equipment life and operation robustness of the system.

The invention relates to a precipitation diagnosis method based on multivariable scale decomposition. The method comprises the following steps: firstly, obtaining high temporal-spatial resolution reanalysis data lattice point data and precipitation measurement satellite observation data; preprocessing rainfall measurement satellite observation data to generate a rainfall daily change data set; on the basis of the high-temporal-spatial-resolution reanalysis data, water vapor revenue and expenditure items are constructed, and a water vapor revenue and expenditure daily change data set is generated; based on the rainfall daily change data set and the water vapor revenue and expenditure daily change data set, establishing a water vapor revenue and expenditure equation of a daily time scale whole-layer integral containing a nonlinear term; finally, time scale decomposition is carried out on multivariable related to the key water vapor income and expenditure nonlinear term, contributions of the water vapor income and expenditure term acted by different time scale variables are calculated, and relative contributions of the different time scale variables in the water vapor income and expenditure nonlinear term are calculated; according to the method, water vapor income and expenditure equation diagnosis and a multivariable time scale decomposition method are combined, and the relative importance of influence of a multi-scale climate mode on rainfall daily change can be quantitatively revealed.

The invention discloses a composite fault diagnosis method for an aero-engine rolling bearing, and the method comprises the steps: collecting a fault vibration signal of the rolling bearing, decomposing the fault vibration information through a natural time scale decomposition method, and obtaining a plurality of component signals; according to the Euclidean distance between the evaluation vector of each component signal and the evaluation vector of the fault vibration signal, selecting two component signals with the maximum Euclidean distance and the minimum Euclidean distance from the component signals; determining the optimal order of Wiener filtering according to the kurtosis, and inputting the component signal into Wiener filtering with the optimal order for filtering processing to obtain a filtered signal; and processing the filtering signal to obtain a signal frequency spectrum, comparing the signal frequency spectrum with the fault characteristic frequency of the bearing, and identifying the fault type corresponding to the fault vibration signal. According to the invention, noise interference can be effectively eliminated, and fault feature information in the original signal can be highlighted, so that accurate fault diagnosis of the rotating machinery can be realized.

The present invention discloses a method for evaluating the operating state of a compressed air energy storage system based on principal component clustering analysis, including: collecting system operation data and preprocessing it; performing multi-time scale decomposition and feature extraction on the preprocessed data set; performing non-linear time-varying adaptive principal component analysis on the multi-time scale feature set to extract the main features of the system operating state; adopting a PCA-clustering iterative optimization mechanism to perform clustering analysis on the dimensionality-reduced feature space; establishing an evaluation model for each functional area of the system based on thermodynamic constraints, and combining the principal component transformation matrix and the clustering results of the operation mode to calculate the performance index values of the functional area; dynamically adjusting the weights of each evaluation index; and generating a comprehensive operating state evaluation result of the CAES. The present invention solves the problems of poor adaptability of PCA under the non-linear time-varying characteristics of the CAES system, difficulty in multi-time scale data fusion, and low sensitivity in abnormal mode detection.

This invention discloses a dynamic power allocation method for hybrid energy storage based on multi-timescale decomposition, belonging to the field of hybrid energy storage control technology. The method is applied to a hybrid energy storage system composed of lithium battery units, supercapacitor units, and a bidirectional converter, sequentially performing signal acquisition, adaptive variational mode decomposition, physical frequency division boundary calculation, dynamic frequency division boundary smoothing correction, and ramp rate limiting and state-of-charge recovery collaborative compensation. Specifically, the number of modes is determined online based on the energy proportion of newly added mode components; the physical frequency division boundary is synthesized using the upper limit of the lithium battery ramp rate and the current closed-loop bandwidth of the bidirectional converter; the frequency division boundary is smoothed using a hyperbolic tangent function based on the supercapacitor's state-of-charge deviation; and the ramp rate limiting difference and the state-of-charge recovery compensation are jointly injected into the supercapacitor channel. This invention can be deployed in real time and is portable across devices, suppressing supercapacitor saturation, reducing lithium battery stress, and ensuring zero steady-state error tracking of total power.

The application relates to the technical field of power systems, and particularly discloses a wind-solar-water-storage coordinated output modeling method based on multi-time-scale feature decomposition, which comprises the following steps: S1, acquisition and preprocessing of multi-source output and meteorological data; S2, adaptive multi-time-scale decomposition based on meteorological driving, which comprises a mixed decomposition framework of EWT, a meteorological driving mechanism and VMD; the framework realizes adaptive multi-scale decomposition on the multi-energy power signal through three-level nesting of frequency domain identification, meteorological adjustment and time-frequency decomposition; S3, construction of an energy coordination mapping matrix; and S4, coordinated output modeling and reconstruction. The application has the advantages that joint modeling and complementary feature quantization of wind power, photovoltaic power, hydropower and energy storage systems are realized, and support is provided for optimal configuration and safe and efficient operation of a wind-solar-water-storage integrated base.

The application relates to the technical field of data mining application, and discloses a multi-time-scale hybrid energy storage capacity configuration method and system and a storage medium. The method comprises the following steps: collecting data according to a wind power plant construction state and preprocessing to obtain standard wind power output power data; obtaining local power grid load curve or dispatching demand data, calculating a difference value to generate an energy storage reference power curve; splitting the power signal by using an SGMD decomposition algorithm, identifying multi-time-scale power waves to generate feature data; setting constraints based on the feature data, optimizing the energy storage capacity configuration by using an immune hybrid particle swarm algorithm; performing full life cycle simulation on the optimization result, verifying the effect and compliance and evaluating the economy to determine the best scheme. The application solves the problems of poor adaptability and insufficient economy of the existing energy storage configuration through multi-time-scale decomposition and closed-loop optimization, and improves the wind power grid stability and configuration economy.

The invention provides a waveform characteristic visualization method and system based on high and low temperature test, current time sequence data of a chip under high and low temperature test working conditions are collected through a Hall effect sensor, and after sliding window filtering and resampling preprocessing, a continuous and smooth current data fitting waveform is generated through local polynomial fitting; dividing the fitting waveform according to a time window, extracting wave crest and wave trough seed points, and obtaining a pulse width and period characteristic data set; an inherent time scale decomposition algorithm is called to strip waveform PR components, and a PR component density degree data set is obtained through discretization processing; fusing the time domain and frequency domain features to calculate a disorder coefficient, a trend coefficient and a warpage index, and constructing a multi-dimensional feature data set; through feature mapping, normalization processing, drawing of a time domain distribution curve and a thermodynamic diagram visualization chart, and generation of a waveform feature visualization data set, extraction and coupling of current waveform features are realized, and visual and reliable data are provided for chip high and low temperature test working condition analysis.

The application discloses a kind of intelligent new energy electric power data monitoring analysis method and system, it is related to electric power system technical field, including the time series data of the multiple monitoring parameters of new energy power generation equipment, determine the normal operating state area of each monitoring parameter;The time series data is decomposed in multiple time scales, extract the fast variable component and slow variable component of each monitoring parameter;Transient coupling matrix, trend correlation matrix and cross-scale conduction matrix are constructed and adaptively fused, to obtain adaptive coupling matrix;According to adaptive coupling matrix, establish parameter evolution model, generate the evolution track of each monitoring parameter;Evolution track is compared with normal operating state area, when deviating, according to adaptive coupling matrix, trace back abnormal source parameter, and output monitoring analysis result.The application can capture the signal characteristics of multiple time scales simultaneously, accurately depict the dynamic coupling relationship between multiple parameters, realize adaptive monitoring analysis and abnormal source fast positioning.

The invention discloses a short-term prediction method for railway traction load, and belongs to the technical field of power data processing. The method comprises the steps of data cleaning, data primary decomposition, data secondary decomposition, sample entropy clustering, model training and data prediction. According to the invention, an intrinsic time scale decomposition algorithm, a robustness local mean decomposition algorithm, sample entropy clustering, a time convolution network and a bidirectional gating loop network are combined to establish a traction load short-term prediction method based on double-layer optimization combination of secondary decomposition reconstruction and deep learning. The method can reduce the complexity of high-frequency components. And meanwhile, the defect that the real-time dynamic characteristics of the load cannot be deeply captured in the existing research is effectively overcome, so that the traction load prediction precision is improved.

The invention discloses an agricultural park energy safety collaborative early warning method, and particularly relates to the technical field of agricultural parks, and the method comprises the steps: obtaining a decomposition difference coefficient according to sensitive error information, obtaining a collaborative load coefficient according to coupling fluctuation information, and constructing a modeling reliability model; the modeling robustness of the energy safety collaborative early warning is evaluated through the influence capability of the coupling term on the fluctuation of the time scale decomposition model, a model robustness index is generated, the model performance is graded, the stability of the modeling reliability model is classified, and the modeling reliability of the energy safety collaborative early warning is improved. An early warning verification intervention strategy is generated according to the effectiveness of the grading result and the classification result on the collaborative early warning, and is notified to management personnel, so that the influence capability of the input variable on the output result is effectively evaluated on the premise of not being interfered by the time scale, the situation that the influence of the input variable is underestimated or overestimated in the time scale division process is avoided, and the time scale division efficiency is improved. Therefore, the overall precision of time scale division is optimized, and verification is provided for the effectiveness of early warning collaborative analysis.

The application discloses a configuration method and device of an energy storage system, electronic equipment and a storage medium, and belongs to the technical field of energy management. The method comprises the following steps: acquiring a load time sequence of a target area in a preset period; performing fluctuation analysis on the load time sequence to extract fluctuation characteristic parameters; performing multi-time scale decomposition on the load time sequence according to the fluctuation characteristic parameters to obtain multi-time scale components; determining rated power demand and rated capacity demand of the energy storage system based on the multi-time scale components; and configuring the energy storage system according to the rated power demand and the rated capacity demand. Through fluctuation characteristic extraction and multi-time scale decomposition on load data, the optimal power capacity and energy capacity of the energy storage system are determined, so that a more optimal fluctuation suppression effect and higher configuration rationality are realized, and the accuracy and economy of energy storage configuration are improved.