Neural network modeling method and system

Abstract

The invention discloses a neural network modeling method and a neural network modeling system. The method comprises: a data preprocessing step of acquiring historical data in actual processes and screening sample data out; a neural network initializing step of initializing parameters of a neural network model; an output calculation step of calculating the output of the neural network model according to the model and the sample input; a gain calculation step of calculating the gain of the output of the model relative to the input according to the parameters of the model and the sample data; and a model training step of instructing the neural network model to perform parameter iterative learning according to an error between the output of the model and the sample data and the input output gain of the model until the model meeting accuracy requirement and gain restriction is obtained. Through the method and the system, gain relationship between the input and the output of the neural network model can be definitely represented and the established model can be more consistent with actual conditions through the gain restriction.

Description

technical field [0001] The invention relates to the technical field of intelligent control, in particular to a neural network modeling method and system. Background technique [0002] With the development of modern process industry production to be complex and large-scale, the simple conventional control method based on feedback can no longer meet the production requirements of modern enterprises, and the design of various advanced control algorithms for the system based on the process model has gradually become a general and effective method . Among them, obtaining the precise mathematical model of the controlled object has become the primary task of the current control system design and the implementation of advanced control projects. [0003] There are mainly two types of system modeling methods, one is mechanism modeling, that is, to establish a system model based on the internal mechanism of the process itself, but most of the actual industrial production process mecha...

AI Technical Summary

Problems solved by technology

However, the BP algorithm has defects such as long training time, slow convergence speed, and easy to fall into local minimum points.

Hush has strictly proved that the parameter learning of a multi-layer neural network is a kind of NP-hard problem. A series of studies have shown that in the training process of the neural network, there are phase transitions and critical points. At the critical point, the optimization problem The computational complexity of will increase greatly, and the optimization process will become very difficult

[0010] To sum up, for the problem of nonlinear system modeling difficulties in the current industrial production, it is necessary to design a nonlinear neural network model that is consistent with actual production experience and can reasonably reflect the relationship between the input and output of the system (ie gain) and stable and efficient. The neural network learning method is of great significance

Images