Method for establishing time series model for predicting multi-phase mixed effect and based on empirical mode decomposition (EMD)

Abstract

The invention discloses a method for establishing a time series model for predicting a multi-phase mixed effect and based on empirical mode decomposition (EMD), which is mainly applied to judging a fluid mixing effect and theoretically guiding the design of experiments in chemical engineering experiments. The method comprises the following steps: (1) acquiring a real-time multi-phase mixing pattern; (2) calculating the zeroth-dimensional betti number of the real-time pattern; (3) acquiring parts of related data by the method in (2) so as to obtain a corresponding time series x (t); (4) finding out all maximum value points and minimal value points of the x (t), and fitting the maximum value points and minimal value points into upper envelope and lower envelope of the data x (t) by a cubic spline function; (5) when h1k meets requirements of SD, causing c1 to be equal to h1k to acquire a first IMF component c1 of the x (t), thus being capable of obtaining a new data sequence r1 of which the high frequency components are removed; (6) repeating step (4) and step (5) until the number of the extreme points of the last data sequence rn is 2; (7) summing each data sequence correspondingly to obtain a final predicted value series (5); and (8) checking predicted errors. The method of the invention is a reliable and practical prediction method for judging the mixed effect in the chemical engineering experiments.

Description

technical field [0001] The invention belongs to the technical field of chemical engineering, in particular to a method for detecting and predicting the mixing effect of all fluids in the chemical industry. Background technique [0002] Time series analysis is the theory and method of establishing mathematical models through curve fitting and parameter estimation based on time series data obtained from systematic observation. It is generally performed using curve fitting and parameter estimation methods such as nonlinear least squares. Time series analysis is commonly used in national economic macro control, regional comprehensive development planning, enterprise management, market potential forecasting, weather forecasting, hydrological forecasting, earthquake precursor forecasting, crop disease and pest disaster forecasting, environmental pollution control, ecological balance, astronomy and oceanography study etc. The basic steps of time series modeling are: ① Obtain the ...

AI Technical Summary

Problems solved by technology

The ARIMA model in the time series forecasting method needs to be differentiated many times to achieve stability, and the forecasting effect is not good

However, the ordinary fitting method has poor real-time performance, and the long-term prediction is even worse. Even the modified fitting method cannot reach a satisfactory level.

However, phase reconstruction needs to calculate a lot of parameters, and the prediction effect is not good. The neural network needs a training set, and the prediction error of the chaotic time series based on the initial value sensitivity is too large, and the verification of the whole process is time-consuming and laborious. , and it is not cost-effective, and cannot effectively control abnormal operations and avoid accidents

Images