Modeling of the nonlinear sequences based on the modified Genetic Algorithm
The GA (Genetic Algorithm) is a technique to mimic some of the processes observed in natural evolution [\textit{P. Grassberger} and \textit{I. Procaccia}, Physica D 9, 189-208 (1983; Zbl 0593.58024)]. The aim of this paper is to propose a new modeling method based on the modified GA which finds rapidly the structure and coefficients of the functions for nonlinear sequences. The modified GA approach was tested to model nonlinear sequences produced by the logistic map function and Henon map function [\textit{R. C. Hilborn}, Chaos and nonlinear dynamics: An introduction for scientists and engineers (1994; Zbl 0804.58002)]. A deterministic modeling function for nonlinear sequences was assumed as a difference equation, which is composed of time-delayed variables and constants and expressed as \(X_n= f(a_1,a_2,\dots, a_m; X_{n-1},\dots,X_{n-k})\), e.g., \(X_n=aX_{n-1}(1-X_{n-1})\) or \(X_n= 1+bX_{n-2}- aX_{n-1}^2\). The simulation results show that the proposed method was efficient to find a modeling function for the nonlinear sequences with a minimum error. Although there were discrepancies between coefficients of the input and estimated function, their statistical and chaotic properties of the two sequences remain very close, respectively. Therefore the proposed modeling method can be used as a powerful method to model and approximate the nonlinear and chaotic behaviour of nonlinear sequences.
- Obtaining functional form for chaotic time series evolution using genetic algorithm
- Chaotic system modelling using a neural network with optimized structure
- Evolving cellular automata rules for multiple-step-ahead prediction of complex binary sequences
- A quadratic empirical model formulation for dynamical systems using a genetic algorithm
- Reconstructing the nonlinear dynamical systems by evolutionary computation techniques
- Comparative evaluation of genetic algorithm and backpropagation for training neural networks
- GENETIC PROGRAMMING BASED APPROACH FOR MODELING TIME SERIES DATA OF REAL SYSTEMS
- A nonlinear model for predicting ECG R-R interval variation based on the evolutionary computation approach
- A quadratic empirical model formulation for dynamical systems using a genetic algorithm
- Obtaining functional form for chaotic time series evolution using genetic algorithm
- A simulation research of the nonlinear Lowry model based on a genetic algorithm
- scientific article; zbMATH DE number 7338578 (Why is no real title available?)
This page was built for publication: Modeling of the nonlinear sequences based on the modified Genetic Algorithm
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4239787)