32 results about "Vine copula" patented technology

According to the method, on the basis of the transfer entropy of a classical causal analysis method, a Copula function in statistics is combined and popularized to the transfer entropy in a high-dimensional form, and a regular vine Copula function is used for estimation, so that an effective new R-Vine Copula multivariable transfer entropy method is provided. Firstly, compared with an existing algorithm on a series of simulation data, the multivariable causal analysis performance of the method is verified. Secondly, the method is applied to multi-channel electromyographic signal analysis when different wrist movement tasks are executed. According to the application of the coupling relation between muscles, when the neuromuscular system executes different wrist movement tasks such as wrist extension and wrist radial deviation, two community structures which are related to a joint area and have a stable causal coupling relation are formed. The R-Vine Copula multivariable transfer entropy accurately deduces the complex causal coupling relationship, and has a good application value.

The invention discloses a method for determining a probability optimal power flow based on a vine copula function. The method comprises the following steps that the minimum power generation cost serves as a target function, and a probability optimal power flow model containing the multi-wind electric field is built; a single wind power plant wind speed distribution function is established according to the non-parameter nuclear density fitting wind speed; a single wind power plant wind speed distribution function and a vine copula function are utilized, and a multi-wind power plant wind speed combined distribution model is built; M scenarios of loads and wind power and the probability of occurrence of each situation can be determined based on a multi-wind power plant wind speed combined distribution model and by means of a Rosenblat transformation and a three-point estimation method; according to the probability optimal power flow model of the multi-wind power plant, the optimal power flow under the M scenes can be calculated by utilizing an internal point method; the probability optimal power flow is obtained according to the optimal power flow under the m scenarios and the probability of each situation. According to the invention, the diversified correlation structure of the wind speed is effectively described, and the calculation efficiency is high, so that effective information can be provided for the operation of the electric power system in time.

The invention discloses a method and device for predicting the combined output of multiple wind farms based on a dynamic R-Vine Copula model, belonging to the field of wind power interval prediction in a power system. The method includes: S1: combining the predicted output data of multiple wind farms with The prediction error corresponding to the joint output is used as the first input data; S2: Input the first input data into the dynamic marginal distribution function model established based on the ARIMA-GARCH model, so that the dynamic marginal distribution function model converts the first input data into a cumulative probability sequence ; S3: Input the cumulative probability sequence into the pre-established dynamic R‑Vine Copula model, so that the dynamic R‑Vine Copula model outputs the joint output prediction results corresponding to multiple wind farms under different confidence levels; wherein, the model of the dynamic marginal distribution function model The parameters and model parameters of the dynamic R‑Vine Copula model are calculated and updated on a rolling basis based on the phase space reconstruction method. The present application can improve the ultra-short-term interval prediction accuracy of the combined output of multiple wind farms.

The invention discloses a multi-failure mode correlation reliability calculation method of mechanical parts, which is applied to the technical field of reliability analysis of mechanical products. The present invention considers the correlation between failure modes, uses the Vine Copula function to characterize the correlation between failure modes, and establishes a multi-dimensional joint probability density function of multiple failure modes; considering that the multi-dimensional joint probability density function is difficult to integrate, According to the obtained multi-dimensional joint probability density function of multiple failure modes considering correlation, each failure mode is transformed into independent random responses; and the failure probability of multiple failure modes is calculated based on the simulation method and saddle point approximation. Compared with the failure probability method for calculating multiple failure modes, the correlation between multiple failure modes is quantified, which is closer to the actual engineering situation.

The invention discloses a multi-wind-power-plant combined output prediction method and device based on a dynamic R-Vine Copula model, which belong to the field of wind power interval prediction in an electric power system. The method comprises the steps of S1, taking prediction output data of multiple wind power plants and prediction errors corresponding to combined output as first input data, S2, inputting the first input data into a dynamic edge distribution function model established based on an ARIMA-GARCH model, so that the dynamic edge distribution function model converts the first input data into a cumulative probability sequence, S3, inputting the cumulative probability sequence into a pre-established dynamic R-Vine Copula model so as to enable the dynamic R-Vine Copula model to output joint output prediction results corresponding to multiple wind power plants under different confidence coefficients, wherein the model parameters of the dynamic edge distribution function model and the model parameters of the dynamic R-Vine Copula model are subjected to rolling calculation and updating based on a phase space reconstruction method. Thus, the ultra-short-term interval prediction accuracy of multi-wind-power-plant combined output can be improved.

The present invention provides an adaptive soft measurement method and system based on vine copula quantile regression. The steps are as follows: obtaining data under normal operating conditions, selecting appropriate auxiliary variables; normalizing training data; using D‑vine copula establishes a quantile regression model; calculates the conditional quantile function and the average prediction interval difference of the training sample according to the training sample; collects and standardizes the auxiliary variables of the sample to be predicted online; determines the conditional quantile function; according to the set quantile Calculate each conditional quantile function value; denormalize the conditional quantile function value, and then obtain the median and confidence interval of the final predicted value; calculate the confidence interval difference of the predicted value, and compare whether the confidence interval difference of the predicted value is The average prediction interval difference over the training sample.

The invention discloses a multi-wind farm power modeling, PDF construction, and prediction scene generation method and system, belonging to the field of wind power scene prediction. The invention establishes the probability distribution model of the prediction error, and adds the probability distribution of the prediction error and the point prediction power of each wind farm power as the edge distribution model of each wind farm power. Calculate the cumulative probability of the power data of each wind farm as the input data of the time-varying R-Vine Copula model. By combining the ARIMA-GARCH-t model and the time-varying R-vine Copula model, a joint probability distribution model of high-dimensional wind power data is established. Based on the fitting model parameters of the historical power data of each wind farm, combined with the point forecast power data of each wind farm in the next day, a method for generating multi-wind farm power day-ahead forecast scenarios is proposed. The day-ahead prediction scenario generation model established by the present invention can better fit the time-space correlation characteristics of multi-wind farm power, and improve the accuracy and effectiveness of multi-wind farm power day-ahead prediction scenarios.

The invention discloses an intelligent evaluation method for a high-temperature capacity reduction state of a wind turbine generator, and the method comprises the following steps: collecting multi-dimensional feature variables of the wind turbine generator, and forming a training data set of a vin-Copula Bayesian network model based on the multi-dimensional feature variables; establishing a pair-Copula function according to the training data set so as to determine an optimal Copula function between every two of the multi-dimensional characteristic variables; according to the optimal Copula function, obtaining a correlation coefficient of each group of pair-Copula functions in the optimal Copula function, constructing a relationship tree based on the correlation coefficient, and generatinga vine-Copula Bayesian network model by utilizing the relationship tree; and evaluating the test samples in the state point test sample set by using the vine-Copula Bayesian network model. The invention further discloses an intelligent evaluation system for the high-temperature capacity reduction state of the wind turbine generator. According to the method, the vine-Copula Bayesian network model is utilized to comprehensively consider the state parameters of each dimension of the characteristic variables of the wind turbine generator and the correlation of the state parameters, and the occurrence probability of the high-temperature capacity reduction state of the wind turbine generator can be accurately evaluated.

The invention discloses a hydrological dependent structure modeling method based on mutual information and vine copula. Firstly, mutual information and conditional mutual information are used to measure the correlation and uncertainty of hydrological variables, in combination with the principle of the strongest correlation and the least uncertainty, the structure of vine copula is selected, starting from the first tree, the mutual information of pairs of paired variables is calculated, the pairing mode that maximizes the sum of mutual information is selected as the edge of the tree, the conditional mutual information of possible pairing variables is calculated, and the pairing mode that maximizes the sum of conditional mutual information is selected as tree 2, and the pairing mode is repeated until the structure of the whole tree is determined. Secondly, according to the tree structure, fitting of edge distribution is carried out, a goodness-of-fit test is carried out, starting from tree 1, the AIC criterion is utilized to determine the copula type of the edge, parameters are estimated, a goodness-of-fit test is performed, then the conditional edge distribution of variables is calculated, and the determination of copula type, estimating parameters and testing steps are repeated until all the trees are determined. All trees and edges are connected to complete the modeling of hydrological dependent structures.

The invention provides an industrial production simulation scene generator and a scene generation method based on vine-copulas. The industrial production simulation scene generator comprises a scene data acquisition module, a scene preprocessing module, a Vine-copulas creation module and a scene generation module. In the scene generation method, the existing scene data is acquired from an initialscene database module, by preprocessing of a scene clustering module, a feature standardization module and a feature dimension reduction module, and so on, a scene generation model is created througha vine copula method through a model fitting module, a goodness detection module and a binary copula model library module, and the scene needed is generated through a scene sampling module and an inverse mapping module. The scene generator provided by the invention can be embedded into an emulation system to be used as an organization module of a production simulation system and used for configuration operations of the production simulation system. The scene generated by the method provided by the invention can be used for scene emulation, and also can provide reliable original data for otherexperiments and researches to perform targeted study.

The present invention proposes a soft sensor method and system based on vine copula, including the following steps: selecting suitable auxiliary variables for the soft sensor model according to actual industrial production conditions and expert knowledge; standardizing and monotonically transforming the training data to obtain the transformed The final data that complies with copula modeling; use D‑vine copula for correlation modeling to obtain the joint probability density function of training sample auxiliary variables and target variables; online collection, standardization processing and monotone transformation calculation of auxiliary variables of samples to be predicted; calculation After processing the copula function values ​​of the auxiliary variable of the sample to be predicted and the target variable of all training samples, the weight of each training sample is calculated; according to the weight of the training sample calculated above, the target variable of the training sample is linearly weighted to obtain the sample to be predicted The predicted values ​​normalized by the target variable are then inversely transformed to obtain the final predicted values.

The invention discloses a modeling method for product multi-dimensional correlation degradation failure; it specifically includes: determining the reliability characteristic quantity of the product, collecting and processing the degradation data of each characteristic quantity, establishing the degradation model of each characteristic quantity, estimating each Parameters in the characteristic quantity degradation model, using D‑Vine Copula to establish a multidimensional correlation degradation model, and determining the types and parameter values ​​of all bivariate Copula functions. The present invention uses D-Vine Copula to connect the joint density function of multi-dimensional correlation variables and its edge density function with multiple bivariate Copula functions, which can effectively deal with multi-dimensional degradation processes with complex correlations, and has better adaptability and promotion ability.