Abstract:Most systems in the field of modern control contain nonlinear parts. The nonlinear system does not take on homogeneous and superpose. When we identify the nonlinear system, traditional method is linearizing the system first, and then using LS to make identification of the system. In order to linearize the system, much information of the system is needed, such as the structure and the order of the system. But the control system is more and more complicated, it is difficult to get much information of the system, so the shortcomings of such method are more and more notable. The artificial neural network has the ability to approximate the nonlinear function in any extent, so it provides a new method of system identification. However, the present neural network has shortcomings, first, the structure will be too large and loose as the system become complicated, on the other hand, the time delay is not arbitrary, it is not according to the fact. To solve such problem, this paper proposes the Universal Learning Network (ULN)([1][2]). This network has characteristic as follows: (1) all of the nodes are connected each other; (2) there are multiple branches between every two nodes; (3) arbitrary time delay can be set on every branch; (4) there is a switching function on every branch. Here, the switching function is used to delete the unnecessary node and branch, to make the network simple. The lean-Ling algorithm of the network uses ordered derivative, and learning based on the gradient decent algorithm. Using ULN to identify the nonlinear dynamical system, it can be proved that this network has excellent learning and general ability, and also make the network compact.

A Novel Linear Recurrent Neural Network for Multivariable System Identification

Neural network system identification based on levenberg-marquardt algorithm

Identification of Nonlinear Systems Using Robust Stable Recurrent High Order Neural Networks

Recurrent Neural Networks for Recursive Identification of Nonlinear Dynamic Process

Robust Identification for Unknown Nonlinear Multivariable Systems Based on Dynamic Neural Networks

A New Neural Network Model And Its Application

Robust On-Line Nonlinear Systems Identification Using Multilayer Dynamic Neural Networks with Two-Time Scales

Complex Nonlinear System Modelling and Parameters Identification by Deep Neural Networks

A New Neural Network Model for the Feedback Stabilization of Nonlinear Systems

Univariate ReLU neural network and its application in nonlinear system identification

Identification of Nonlinear Discrete-Time Multivariable Dynamical Systems by RBF Neural Networks

Quasi-Linear Recurrent Neural Network Based Identification and Predictive Control

Nonlinear system identification using a recurrent network in a Bayesian framework

Efficient Identification Of Rbf Neural-Net Models For Nonlinear Discrete-Time Multivariable Dynamical-Systems

Automatic Structure and Parameter Training Methods for Modeling of Mechanical System by Recurrent Neural Networks

Nonlinear Systems Identification Using Deep Dynamic Neural Networks

Nonlinear systems identification and control via dynamic multitime scales neural networks.

Affection Of The Multi-Branch Number Of Universal Learning Networks On Network Structure

Evolutionary Diagonal Recurrent Neural Network For Nonlinear Dynamic System Identification

A Real-Time Learning Control Approach for Nonlinear Continuous-Time System Using Recurrent Neural Networks

A Novel Time-Delay Recurrent Neural Network and Application for Identifying and Controlling Nonlinear Systems