Application of a Novel Data Mining Method Based on Wavelet Analysis and Chaotic Neural Network on Satellite Clock Bias Prediction

2014 ◽  
Vol 513-517 ◽  
pp. 1144-1149
Author(s):  
Yue Hu ◽  
Dao Ping Tang

A novel four-stage data mining method for clock bias prediction based on wavelet analysis and chaotic neural networks is proposed. The basic ideas, prediction models and steps of clock bias prediction based on wavelet analysis and chaotic neural network are discussed respectively. And then, to validate the feasibility and validity of the proposed method, make a careful precision analysis for satellite clock bias prediction with the performance parameters of GPS satellite clock, and make comparison and analysis with Grey system model and neural network model. The results of simulation shows that the prediction precision of the novel four-stage model based on wavelet analysis and chaotic neural networks is more better, can afford high precise satellite clock bias prediction for real-time GPS precise point positioning.

2020 ◽  
Vol 38 (4) ◽  
pp. 3717-3725
Author(s):  
Jingyong Zhou ◽  
Yuan Guo ◽  
Yu Sun ◽  
Kai Wu

2001 ◽  
Vol 11 (06) ◽  
pp. 1631-1643 ◽  
Author(s):  
HIROYUKI KITAJIMA ◽  
TETSUYA YOSHINAGA ◽  
KAZUYUKI AIHARA ◽  
HIROSHI KAWAKAMI

We investigate a noninvertible map describing burst firing in a chaotic neural network model with ring structure. Since each neuron interacts with many other neurons in biological neural systems, it is important to consider global dynamics of networks composed of nonlinear neurons in order to clarify not only mechanisms of emergence of the burst firing but also its possible functional roles. We analyze parameter regions in which burst firing can be observed, and show that dynamics of strange attractors with burst firing is related to the generation of a homoclinic-like situation and vanishing of an invariant closed curve of the map.


2018 ◽  
Vol 22 (3) ◽  
pp. 225-242 ◽  
Author(s):  
K. Mathan ◽  
Priyan Malarvizhi Kumar ◽  
Parthasarathy Panchatcharam ◽  
Gunasekaran Manogaran ◽  
R. Varadharajan

2013 ◽  
Vol 380-384 ◽  
pp. 1860-1863
Author(s):  
Ping Zhang Gou ◽  
Yong Zhong Tang

This study proposed a novel relational database data mining method based on the artificial neural network. It analyzed the disadvantages of the existed data mining methods and then introduced the novel algorithm. This algorithm discovered the implicit knowledge by training the data samples in the database. This study introduced the artificial neural network method training model and algorithm, and tested the method by an example.


Sign in / Sign up

Export Citation Format

Share Document