Abstract:The application of neural networks to non-linear dynamic system identification tasks has a long history, which consists mostly of autoregressive approaches. Autoregression, the usage of the model outputs of previous time steps, is a method of transferring a system state between time steps, which is not necessary for modeling dynamic systems with modern neural network structures, such as gated recurrent units (GRUs) and Temporal Convolutional Networks (TCNs). We compare the accuracy and execution performance of autoregressive and non-autoregressive implementations of a GRU and TCN on the simulation task of three publicly available system identification benchmarks. Our results show, that the non-autoregressive neural networks are significantly faster and at least as accurate as their autoregressive counterparts. Comparisons with other state-of-the-art black-box system identification methods show, that our implementation of the non-autoregressive GRU is the best performing neural network-based system identification method, and in the benchmarks without extrapolation, the best performing black-box method.

Recurrent Neural Networks and Their Application in System Identification

Recurrent Neural Networks for Recursive Identification of Nonlinear Dynamic Process

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

Neural network system identification based on levenberg-marquardt algorithm

Identification for Non-Linear Systems Based on Particle Swarm Optimization and Recurrent Neural Network

Recurrent Neural Network Baged Identification Modeling Study

Study on the recursive identification algorithm for dynamic systems using multi-layered neural network model

Non-Autoregressive vs Autoregressive Neural Networks for System Identification

Training Elman Neural Network For Dynamic System Identification Using An Adaptive Local Search Algorithm

Design and Application of Recurrent Neural Networks

Evolutionary Elman Neural Network Model and Identification for Non-linear Systems

Identification and Control of Time-Delay System with Recurrent Wavelet Neural Networks

Evolutionary Diagonal Recurrent Neural Network For Nonlinear Dynamic System Identification

High-order neural network structures for identification of dynamical systems

Nonlinear Systems Identification Using Deep Dynamic Neural Networks

Application of Second Order Diagonal Recurrent Neural Network in Nonlinear System Identification

A Novel Linear Recurrent Neural Network for Multivariable System Identification

Dynamic T-S Recurrent Fuzzy Neural Networks and Its Application

Univariate ReLU neural network and its application in nonlinear system identification

The Study of Non-Linear Structure’s Identification using Dynamic Neural Network

Nonlinear System Identification Based On Recurrent Neural Networks With Shared And Specialized Memories