Time sequence prediction method based on chaotic optimization neural network model

Abstract

The invention discloses a time sequence prediction method based on a chaotic optimization neural network model, and the method comprises the steps: firstly obtaining urban historical daily water demand time sequence data online from a database, carrying out the data preprocessing, and carrying out the chaotic feature recognition; determining the structure of the chaos optimization BP neural network by using the embedded dimension of the model input data reconstruction phase space, and carrying out model training at the same time; finding a global optimal value of the BP neural network weight through chaotic optimization search in the model training process; after model training is finished, carrying out chaotic parameter control on an output time sequence predicted value through a parameter control method; and finally, realizing prediction of the urban daily water demand time sequence. The urban daily water demand time series prediction method based on the chaotic optimization neural network model needs few training data samples, is high in convergence rate, can easily reach a global minimum value, is good in prediction result overall error index, and shows good comprehensive prediction performance.

Description

technical field [0001] The invention belongs to the technical field of model prediction control, and in particular relates to a time series prediction method based on a chaos optimization neural network model. Background technique [0002] The rapid economic development has brought about the continuous development of urbanization and industrial production scale, which also makes the contradiction between supply and demand of urban water supply system become more prominent, especially in summer, the phenomenon of urban water supply shortage is becoming more and more common. The time series prediction of urban daily water demand has become an important research field in the discipline of modern water supply dispatching system. Since the time series growth of urban daily water demand is affected by many factors such as economic development, industrial structure, residents’ income level, climate, etc., using conventional mathematical methods to establish models not only requires...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a time series prediction method based on the chaos optimization neural network model, which solves the problem that the existing urban daily water demand time series prediction model method requires many training samples, large prediction errors, slow algorithm learning convergence speed, and network problems. Difficult to determine the structure and other issues

Images