44 results about "Granger causality" patented technology

A conjoint analysis method for electroencephalograph and electromyography signals based on autonomous movement and imagination movement comprises the steps of performing system setup, and using a LabVIEW 8.6 to generate square wave pulse signals; respectively collecting electroencephalograph signals and electromyography signals including electroencephalograph signals and electromyography signals in autonomous movement modalities and in imagination movement modalities; performing noise removal pretreatment on collected original data; performing electroencephalograph and electromyography time-domain signal analysis in the autonomous movement and imagination movement modalities on electroencephalograph and electromyography signal time-domain pictures which are performed with noise removal pretreatment in the autonomous movement and imagination movement modalities; performing time-frequency signal analysis on electroencephalograph and electromyography signals performed with noise removal pretreatment and in the autonomous movement and imagination movement modalities based on Morlet wavelet transformation; and performing partial directional coherence analysis, and in particular adopting granger causality to perform the partial directional coherence analysis. The conjoint analysis method provides new evaluation parameters for monitoring recovery auxiliary equipment and assessing organism movement level.

The invention relates to a method and a device for utilizing time sequence correlation to perform IT fault root cause analysis. The method comprises the steps that system logs are acquired; key fields of the system logs are extracted, and statistical analysis is performed on the key fields to obtain time sequence data of the system logs; relevant features of the time sequence data are automatically extracted based on quantitative hypothetic detection; and when an IT fault occurs, the relevant features of the time sequence data are inspected through Granger causality, wherein a quantitative value of the causality between all the relevant features of the time sequence data serves as a basis for evaluating an occurrence cause of the IT fault. The method and the device have the advantages that a fault root cause analysis process is completed automatically by means of machine learning, a user is helped to quickly find out the fault occurrence cause, fault diagnosis time (Mean Time To Diagnose, MTTD) is shortened, and a system is returned to normal in a highest speed.

The invention discloses a flight delay causality network building method based on a Granger test, pertaining to the technical field of delay analyses for civil aviation. The method comprises the following steps: firstly, acquiring a delay time sequence of each airport with the set time period according to historical data of flight of airports; then, taking each airport as a node for the network and testing a Granger causality of each pair of airports, when an airport I causes delay of an airport j by the test and establishing a directed edge from a node i to node j; and ultimately, repeating the above processes till selected airports are added to the network and making flight delay analyses by means of the obtained network. Through the established flight delay causality network, the numberof delay dissemination paths is analyzed for some large airports. Therefore, the delay dissemination range is narrowed, thereby effectively resolving a delay problem. The delays disseminated by an overall system are reduced. Therefore, loss caused by many flight delays is lowered.

The invention provides a method for predicting the key industrial electricity consumption based on an industrial condition index. The method comprises the following steps: (1) obtaining the key industrial condition index and historical electricity consumption data; (2) performing seasonal adjustment and a stationary test on the data; (3) judging whether the industrial condition index and the industrial electricity consumption have a causal relationship or not through a Granger causality test and determining an optimal lag period of the condition index; (4) creating a time sequence ARIMA (autoregressive integrated moving average) model of the key industrial electricity consumption, introducing the key industrial condition index into an original ARIMA model, and creating a regressive model; (5) on the basis of an AIC (Akaike information criterion), screening out an optimal model; (6) performing model popularization and application, and predicting the industrial electricity consumption in the future. The key industrial electricity consumption is taken as a study object, the electricity consumption and the influence of the industrial condition index on the electricity consumption are studied by introducing the industrial condition index, the key industrial electricity consumption is accurately predicted in combination with the time sequence model, and a basis is provided for development and planning of electricity industry in the future.

The invention discloses a brain functional effective connection analyzing method, which is provided due to the fact that: in the technical field of conventional brain functional imaging, when a conditional Granger causality method is adopted to study an effective connection between interested regions of the brain functional magnetic resonance data and electroencephalogram data, errors occurring in the effective connection analysis are caused because of high redundancy existing among multiple channels of signals. The method comprehensively applies the conditional Granger causality of Yonghong Chen and principal component analysis, does not need to consider the coherence problem of estimating the same spectrum in a calculating process by conditional Granger causality, removes a negative value causality spectrum without physical meanings, eliminates the high-degree redundancy existing among multiple channels of signals, and effectively improves the accuracy of effective connection analysis.

The invention discloses an industrial data nonlinear cause and effect analysis method based on a sparse deep neural network. The method comprises the following steps of (1) collecting the process output signals of all control loops in an industrial process to be detected; (2) selecting one process output signal as an output variable and taking each order lag of all variables as an input variable to construct the sparse deep neural network; (3) through successively deleting the input variables, completing Granger causality testing one by one, and acquiring all the Granger causes of the output signal of the process; (4) repeating the steps (2) and (3), and acquiring a causal relationship among the output signals of all the process; and (5) integrating the causal relationship among the outputsignals of all the process, and positioning a fault source location and fault propagation path. In the invention, nonlinear causal analysis can be performed on the control loop signal of the industrial process, and fault source positioning and fault propagation path analysis are completed.

An electroencephalogram source localization method based on granger causality includes the steps: recording scalp electroencephalogram signals of a plurality of leads by an electroencephalogram acquisition device and performing elementary pretreatment; respectively analyzing the granger causality among each lead and the other leads by taking each lead as an observation lead, and performing source localization among the leads according to causality indexes; counting the number of the leads capable of becoming sources, and calculating the possibility index of taking each lead capable of becoming a source as a whole brain source area to realize whole brain source localization. The electroencephalogram source localization method solves the problems of poor electroencephalogram source localization stability and uniqueness, and the possibility of taking each lead as the whole brain source area can be obtained, so that whole brain source localization is realized. Instability and non-uniqueness of solutions of electroencephalogram inverse problems can be improved to a certain extent. The electroencephalogram source localization method can be used for determining a cranial nerve system disease focus portion, localizing a neurosurgery operation and performing electroencephalogram source localization and tracking in recognition tasks, and has an important significance in scientific research and clinical practice.

The invention belongs to the technical field of wind speed prediction, and discloses a wind speed prediction method and system based on a neighborhood gate long-short-term memory network. The Pearsoncorrelation coefficient and the maximum information coefficient are respectively adopted to explore the linear and nonlinear correlation among variables to screen the wind speed correlation factors. On the basis of correlation analysis, the Granger causality test is used to explore the statistical causality of wind speed and wind speed factors. The structure of causality is divided into five types, and all types of causality are unified into an equivalent tree causality structure by the method of 'decomposition-dummy variable-pruning'. Aiming at the causality structure of equivalent tree, a long-term and short-term memory network model based on neighborhood gates is proposed to predict wind speed. The prediction method (NLSTM) of the invention accurately considers the causality between thewind speed and the wind speed factors, effectively improves the prediction accuracy of the wind speed, and plays a vital role in the application of the wind power and the dispatching of the power grid.

The invention provides a 10kV line change relationship evaluation method based on a grey correlation analysis method. The method comprises steps that S1, original data of same period line loss evaluation of the transmission line is collected; S2, the collected original data is pre-processed; S3, the correlation degree is acquired through a grey correlation analysis method; S4, transformers with low correlation degrees are selected after sorting according to the correlation degrees; and S5, a Granger causality test method is utilized to test whether the transformers with the low correlation degrees are causes for line loss power generation, and the transformers with the abnormal line change relationship are determined based on the test result. The method is advantaged in that the human andfinancial resource consumption is reduced, the line change relationship of a power system is effectively and quickly sorted, the number of lines with disqualified line loss is quickly reduced, the causes of abnormal line loss are ascertained, the effect of line loss treatment is effectively improved, and the improvement bottleneck of the same period line loss evaluation index in the prior art is broken through.

The invention relates to a method for locating and recognizing a process multi-loop fluctuation source, in particular to a method for locating and recognizing a process multi-loop variable fluctuation source in the field of the system performance estimation and the fault detection and diagnosis of a process industry. The method screens out disturbed fluctuation variables containing similar frequency components by combining spectral independent component analysis, carries out the Granger causality examination and analysis of the disturbed fluctuation variables and intuitively expresses the cause-effect influence relation among the loop variables by using a cause-effect relation diagram so as to represent propagation paths disturbed by fluctuation, simplifies the cause-effect relation diagram by utilizing apriori process knowledge and filters secondary cause-effect relation branches in an automatic threshold value searching mode to obtain a main propagation path for locating and recognizing a fluctuation source. The invention has the advantages that the accurate locating of the fluctuation source is beneficial to the maintenance and overhaul of subsequent fault loops and is an important link for solving the problem of performance lowering of a device with multi-level loops.

The invention discloses a causal relationship mining method based on deep learning. The method comprises the following steps: firstly, preprocessing data by using technologies such as missing data supplement, data normalization and independent hot coding; then, based on a Keras deep learning framework, carrying out univariate time sequence prediction on the target characteristics by using an LSTM;adjusting the model structure and a series of hyper-parameters to obtain an optimal model; recording the R2 score of the model on the test set; and then, predicting all the alternative features by using the model to obtain R2 scores of the alternative features on the test set, subtracting the two scores to obtain a Granger causal relationship score of the alternative features and the target feature, and obtaining a quantitative number for describing the Granger causal relationship between the alternative features and the target feature. The method is suitable for the problem of influence factor analysis of other time sequences. In conclusion, the Granger causal relationship mining method based on deep learning has the advantages of mining much data in various fields.

The invention discloses a wireless network performance optimization method based on causality diagnosis. The method includes the following steps: a data preselection step which includes obtaining historical data of all base stations in a target city and setting indices of data analysis, selecting a plurality of network communication index variables according to the indices and selecting time series data of the network communication index variables in any one time period; a data analysis step which includes adopting a Granger causality test method to analyze the time series data and outputtingan index directed causality network diagram; and an optimization scheme formulation step in which a user analyzes the factors influencing wireless network performance optimization according to a dependency relationship among the indices presented by the index directed causality network diagram and formulates a corresponding scheme or intervention measures. The invention also discloses electronic equipment and a storage medium. The wireless network performance optimization method based on causality diagnosis effectively discovers and summarize mutual influence rules among the related communication indices influencing the performance of the wireless network from the data, and thus provides more effective data reference for improving the quality of the wireless network.

The invention provides a method for constructing industrial equipment fault relationship based on Granger causality verification, and the method comprises the steps: collecting at least two types of operation data of to-be-tested industrial equipment, and correspondingly forming a piece of time sequence data for each type of operation data according to the time sequence; preprocessing the time series data; verifying the preprocessed pairwise time series data by adopting Granger causality verification; and according to the verification result, constructing a causal relationship pointing graph between two kinds of operation data corresponding to the time series data in pairs to form a fault relationship graph of the to-be-tested industrial equipment. The problem that in fault relationship construction based on industrial equipment at present, accuracy is low and directivity is lacked due to dependence on correlation between the operation data are solved. The scheme determines the causalrelationship between the operation data by utilizing Granger causal relationship verification, the directivity is clear, and the success rate of matching the constructed fault relationship with the actual alarm case is relatively high according to measurement of the actual alarm case.

The invention discloses a PCA and Granger causality based brain network feature extraction method. The method comprises: firstly, performing brain functional region division on a plurality of channel signals; secondly, extracting maximum main component time information of each functional region by utilizing PCA; and finally, calculating causality measure between maximum main components, and taking the causality measure as a feature parameter. According to the method, connection effects among the brain regions are revealed based on a Granger causality theory from a brain functional network effect, so that rich information contained in electroencephalogram signals is represented more comprehensively and further mode classification is facilitated.

The invention relates to a method for analyzing morning-evening rush hour congestion conditions and propagation mechanism of road network based on Granger causality. The method comprises the following steps of: 1) establishing a time sequence of the traffic state of an intersection: modeling road intersections by adopting a dual method, taking a road section as a node, abstracting the intersections as edges between the nodes, preprocessing traffic state data, sorting out a directional intersection all-day saturation time sequence, selecting congestion time in the morning and evening rush hours, and selecting a traffic state time sequence with a set length forwards; 2) constructing a road network congestion propagation relationship network based on Granger causality; 3) searching key nodes based on the average out-degree and the average in-degree of the nodes, introducing a dynamic propagation model, fitting the congestion propagation process of the road network, and constructing a traffic congestion propagation mechanism of the road. The method is simple and beneficial to software implementation, and the result can be applied to traffic state analysis and traffic guidance and control systems.

The invention provides a Granger causality discrimination method based on a quantitative minimum error entropy criterion. According to the method, the coefficient and the order of a regression model are determined by adopting the quantitative minimum error entropy criterion and a Bayesian information criterion, a causality discrimination index is obtained by calculating the error entropy and coefficient, and the causality between two time sequences is determined according to a causality judgment standard. Compared with a traditional Granger causality discrimination method based on a minimum mean square error criterion, the method is more accurate in estimating coefficients of the regression model, the obtained error entropy is smaller, and the causality discrimination index can be more accurately calculated. Due to the adoption of a quantization method, the calculation complexity of the method is remarkably reduced. The method integrates the error entropy and the coefficient when calculating the causality discrimination index, which makes the calculation of the causality discrimination index more accurate and robust. Therefore, the Granger causality discrimination method based on the quantitative minimum error entropy criterion provided by the invention is more easily promoted and used in practical applications.

The invention relates to a method and a system for forecasting the investment demand of a grid. The forecasting method comprises the following steps: 1) selecting variable sequences that have the co-integration relationship and significance and influence the grid investment, and establishing a grid investment demand long-term equilibrium model in advance; 2) solving the grid investment demand long-term equilibrium model to obtain a long-term equilibrium relationship between the grid investment sequence and the corresponding variable sequence; 3) performing the Granger causality test to obtaina causal relationship between the grid investment sequence and the corresponding variable sequence; 4 ) performing the vector error correction test to obtain the corresponding grid investment change amount; and 5) performing logarithm inverse transformation on the grid investment change amount corresponding to each variable sequence, and combining the grid power investment base value of the previous year to obtain the corresponding current-year grid investment short-term adjustment value of each variable sequence. The method and the system can be widely used in the power system grid investmentquantitative evaluation technology.

The invention discloses a medium and long term load prediction method, and the method comprises the steps: finding out an economic factor which has a long term equilibrium relation with electric quantity through a co-integration test in measurement economics based on a Granger causality test and an LSTM; determining economic factors beneficial to electric quantity prediction by using a Granger causality test method so as to reduce the number of input variables of the prediction model; and finally, inputting the economic factor data into the LSTM model for load prediction. A method of combining the Granger causal relationship test and the LSTM time sequence prediction model is introduced into a multivariable system, the medium-and-long-term load prediction model which is not easy to overfit and high in expandability is constructed, and the model is used for medium-and-long-term load prediction. And the predicted medium-and-long-term load has relatively high accuracy.

The invention provides a multi-wavelet-basis function expansion-based accurate identification method of spike-potential time-varying Granger causality (GC), and belongs to the technical field of signal analysis and processing. As shown in FIG.1, the method includes: firstly, using an AIC (Akaike information criterion) method to select optimal memory length corresponding to each neuron; then establishing a generalized L-V (Laguerre-Volterra) model, and using a multi-wavelet-basis function method to expand the same to obtain a time-invariant parameter model; then carrying out sparsification on the expanded-formula model through an OFR algorithm, estimating sparse model parameters, and inversely reconstructing a time-varying kernel function in the generalized L-V model; and finally, carryingout solving of logarithmic likelihood values of a model point process, and calculating final time-varying Granger causality values of the corresponding neurons. Compared with existing SSPPF (stochastic state point process filter)-based time-varying Granger estimation method, the method provided by the invention can better track fast-changing causality relationships, improves time-varying causalityrecognition accuracy, and provides a theoretical calculation framework and a new solution for neuron spike-potential time-varying function connection identification.

The invention discloses a topology sensing method, device and system for a non-cooperative wireless network. The topology sensing method comprises the steps of: acquiring a node serial number and a sending moment of a target network for sending data in a period of time through a signal detection mechanism, and forming a data matrix; carrying out Granger causality hypothesis testing on every two data matrixes, solving a mean value of Granger causality zero distribution by utilizing a time window method, and screening as a threshold value to obtain a potential neighbor set of the nodes; and carrying out conditional Granger causality hypothesis testing according to the data matrix and the potential neighbor set grouping, solving a mean value of conditional Granger causality zero distribution by using a time window method, and screening as a threshold value to obtain a final neighbor set of the nodes, thereby realizing topology perception. According to the topology sensing algorithm based on conditional Granger causality, the communication relation can be inferred according to the signal sending data matrix under the condition that the data packet is not decoded, network topology information can be accurately reasoned, and topology reasoning can be carried out by utilizing sensing information.

The invention discloses a method for mining an air pollution propagation path. The method is oriented to atmospheric pollution multi-dimensional time series data, and provides a method for mining an air pollution propagation path through the air pollution multi-dimensional time series data. Granger causality test is used to mine the data with Granger causality, and then the causality among air pollution pollutants is mined to find the propagation path of air pollution. Compared with other methods, Granger causality test can reveal the intrinsic relationship between different pollutants in airpollution through the complexity of pollutant concentration changes and their interactions, and the pollutants with causality are extracted from the data.

The invention discloses a polygraph method based on multi-channel EEG signal Granger causality. The method comprises the following steps: calculating the GC values between the respective lead EEG signals of the two types of subjects, generating a GC adjacency matrix of the two types of subjects; using the GC values of the electrode pairs with significant differences as classification characteristic, sending the GC values into the machine learning algorithm, performing the model training of machine learning, performing real time extraction on the EEG signals of the testers through the multi-lead EEG electrodes, obtaining the tester's multi-lead EEG signals and saving the signals, and performing respective pre-processing operations on the respective lead EEG signals, calculating the GC values between the tester's respective lead EEG signals, generating the tester's GC adjacency matrix, and sending the GC adjacency matrix as input to a trained classifier to obtain the polygraph result. The scheme is based on brain nerve signals, which greatly reduces the stimulation frequency required for testing, finally, after testing, the accuracy of lie detection is greatly improved.

The invention relates to a method for detecting the influence of transcranial magnetic stimulation on working memory based on Granger causality, which comprises the following steps of: 1, performing transcranial magnetic stimulation on a target object, and implanting a microelectrode into the head of the target object; 2, enabling the target object to execute a T-maze behavioral task, collecting a local field potential signal of a prefrontal cortex layer of the target object through a microelectrode, preprocessing the local field potential signal, and extracting a gamma frequency band; 3, employing a Granger causal algorithm to calculate GC values between the microelectrode channels in the gamma frequency band, by using the GC values as characteristic parameters, constructing a Granger causal network, wherein the size of the GC values and the Granger causal density of the Granger causal network reflect the influence of transcranial magnetic stimulation on the working memory of the target object. According to the method, a local field potential signal of a prefrontal cortex of a target object is collected, and the Granger causality of a neural rhythm is analyzed, so that the working memory and cognitive functions of the target object are reflected.