Title :
Time-Variation Nonlinear System Identification Based on Bayesian-Gaussian Neural Network
Author :
Liu, Yijian ; Peng, Chen
Author_Institution :
Sch. of Electr. & Autom. Eng., Nanjing Normal Univ., Nanjing, China
Abstract :
A Bayesian-Gaussian neural network (BGNN) method for nonlinear time variation system identification is proposed in this article. In the redefined BGNN training algorithms, the threshold matrix parameters are optimized by the swarm intelligence optimization algorithm(s) off-line and the sliding window data method are adopted for the BGNN on-line prediction. Some typical time-variation nonlinear systems are been used for the validation of the BGNN modeling effectiveness.
Keywords :
learning (artificial intelligence); neurocontrollers; nonlinear control systems; optimisation; time-varying systems; BGNN training algorithm; Bayesian-Gaussian neural network; nonlinear system identification; swarm intelligence optimization algorithm; time-variation system identification; Artificial neural networks; Automation; Bayesian methods; Computer networks; Finite impulse response filter; Network topology; Neural networks; Nonlinear systems; Optimization methods; Particle swarm optimization;
Conference_Titel :
Natural Computation, 2009. ICNC '09. Fifth International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-0-7695-3736-8
DOI :
10.1109/ICNC.2009.187
