A Signal Prediction Method Based on Variational Mode Decomposition and Support Vector Regression

Abstract

The invention discloses a signal prediction method based on variational mode decomposition and support vector regression. Firstly, empirical mode decomposition is performed on the collected original signal, and key parameters are adaptively obtained by using empirical mode decomposition. The number of effective mode components is K , using the number of effective modal components K to achieve variational modal decomposition, can effectively eliminate the combination of modal aliasing and boundary effects, can effectively separate the disturbance signal, and realize the adaptive decomposition of the signal, so as to be decomposed into K The eigenmode component of the eigenmode component, and then support vector regression prediction is performed on the signal after the variational mode decomposition, and the time series prediction of the signal is performed by using the support vector regression for nonlinear and non-stationary data, and then reconstructed to obtain For the final prediction signal, the method of the present invention is simple and the obtained signal is stable.

Description

technical field [0001] The invention relates to a time series signal prediction method, in particular to a signal prediction method based on variational mode decomposition and support vector regression. Background technique [0002] At present, the research on forecasting methods of time series signals mainly focuses on single-method forecasting research such as Auto-Regressive Moving Average (ARMA), which is stretched for nonlinear and non-stationary signal processing. The methods for signal decomposition mainly focus on wavelet transform (Wavelet Transform, WT), empirical mode decomposition (Empirical Modes Decomposition, EMD) and other methods. The WT method has better time-frequency resolution and is a multi-scale time-frequency analysis method, but it relies too much on the basis function to achieve adaptive signal decomposition. The EMD method is an adaptive signal decomposition method, which can effectively avoid the influence on the basis function, but the EMD metho...

AI Technical Summary

Problems solved by technology

The WT method has better time-frequency resolution and is a multi-scale time-frequency analysis method, but it is too dependent on the basis function to achieve adaptive decomposition of the signal

The EMD method is an adaptive signal decomposition method, which can effectively avoid the influence on the basis function, but the EMD method is not effective for the disturbance signal due to its lack of complete mathematical theory support and the results are prone to modal aliasing and boundary effects. separated and have an impact on the prediction results

Images