Empirical mode decomposition endpoint effect inhibition method based on grey GM (1,1) forecasting model

Abstract

The invention discloses an empirical mode decomposition endpoint effect inhibition method based on a grey GM (1,1) forecasting model, and belongs to the field of signal processing. The method comprises the following steps: by combining a grey GM (1,1) forecasting model theory, establishing a relevant grey GM (1,1) forecasting model by using searched extreme value points (or data sequences) in an empirical mode decomposition 'screening' process, respectively forecasting a maximum value and a minimum value at a data endpoint, and forecasting and expanding original data sequences. Therefore, the influences of an endpoint effect to decomposition results in the empirical mode decomposition 'screening' process can be effectively reduced; and moreover, a small data volume required for forecasting modeling can be achieved, and the empirical mode decomposition endpoint effect inhibition method disclosed by the invention is especially suitable for improving the empirical mode decomposition precision of small sample data.

Description

technical field [0001] The invention relates to the field of signal processing, in particular to a gray GM (1,1) prediction model-based method for suppressing end-point effects of empirical mode decomposition. Background technique [0002] Due to the existence of nonlinear devices such as amplifiers, modulators, demodulators, limiters, mixers, and switching circuits or pulse circuits, the data required to establish the electromagnetic compatibility model of the system or equipment basically presents non-linear components. Due to the linear characteristics, it is very difficult to directly establish the electromagnetic compatibility model of the system or equipment through this kind of data. How to process the test data, improve the modeling accuracy and reduce the difficulty of modeling has become the focus of people's research. [0003] The Empirical Mode Decomposition (EMD) method has unique advantages in dealing with nonlinear and non-stationary data. Through EMD decompos...

AI Technical Summary

Problems solved by technology

[0004] However, the empirical mode decomposition is an "empirical" algorithm, and there is no strict theoretical model in mathematics, and there are still some deficiencies in many aspects.

In the "screening" process of empirical mode decomposition, it is necessary to determine the extreme points in the data multiple times to fit the upper and lower envelopes, and it is impossible to determine whether the data at both ends of the data sample are extreme points, resulting in the formation of The upper and lower envelopes will diverge at both ends of the data. During the screening process, it is necessary to fit the upper and lower envelopes many times. Such divergence will gradually pollute the entire data from both ends of the data "inward". , leading to large distortion of the decomposition results, so the "endpoint effect" problem of empirical mode decomposition is the most prominent

Images