Title :
Time series predictive wavelet neural network control method
Author :
Liu Jing-wei ; Wang Pu ; Liu Hong ; Yang Lei
Author_Institution :
Coll. of Comput. Sci., Beijing Univ. of Technol., Beijing, China
fDate :
May 31 2014-June 2 2014
Abstract :
Vector time series autoregressive moving average predictive and wavelet neural networks PID method is proposed in order to solve the requirement of optimal parameters real-time calculation in various control fields (engineering systems, economics systems, etc.), especially to solve the instability and poor control performance problem of control parameters online-tuning in engineering practice. Based on additional wavelet neural networks and vector autoregressive moving average predict method and comparative experiments of relative methods, better feasibility, reliability, speed, lower static error, more flexible parameter adjustment ability are verified.
Keywords :
autoregressive moving average processes; neurocontrollers; optimal control; predictive control; three-term control; time series; vectors; wavelet neural nets; control parameters online-tuning; engineering practice; optimal parameters real-time calculation; predictive neural networks PID method; time series predictive wavelet neural network control method; vector autoregressive moving average predict method; vector time series autoregressive moving average; wavelet neural networks PID method; Adaptive systems; Delay systems; MATLAB; Neural networks; Simulation; index control; online-tuning; predict control; vector time series; wavelet neural network;
Conference_Titel :
Control and Decision Conference (2014 CCDC), The 26th Chinese
Conference_Location :
Changsha
Print_ISBN :
978-1-4799-3707-3
DOI :
10.1109/CCDC.2014.6852172
