Title :
Gaussian Process adaptive control of nonlinear system base on online algorithm
Author_Institution :
Coll. of Autom. Sci. & Eng., South China Univ. of Technol., Guangzhou, China
Abstract :
It takes massive computation time to find the optimal hyper parameters of Gaussian Process. That can not be applied to the online training in real time applications or time-variant data source. The online algorithms proposed by other researchers are high computationally intensive. This manuscript presents natural gradient online algorithm for GP regression. GP may be used as a universal function approximator. Then an adaptive GP controller is designed in the state feedback control for a class nonlinear system. In order to demonstrate the availability of this adaptive GP controller, a simulation of the inverted pendulum system is given. The results of simulation demonstrate this GP online algorithm is very effective and the GP controller can achieve a satisfactory performance.
Keywords :
Gaussian processes; adaptive control; control system synthesis; function approximation; gradient methods; nonlinear control systems; pendulums; regression analysis; state feedback; GP online algorithm; GP regression; Gaussian process adaptive control; Gaussian process optimal hyperparameters; adaptive GP controller design; inverted pendulum system; natural gradient online algorithm; nonlinear system; state feedback control; universal function approximator; Adaptive systems; Algorithm design and analysis; Approximation algorithms; Estimation; Gaussian processes; Nonlinear systems; Training; Gaussian process; adaptive control; online algorithm;
Conference_Titel :
Control Conference (CCC), 2014 33rd Chinese
Conference_Location :
Nanjing
DOI :
10.1109/ChiCC.2014.6896478
