TITLE:
Stability Analysis of a Neurocontroller with Kalman Estimator for the Inverted Pendulum Case
AUTHORS:
Julián Antonio Pucheta, Cristian Rodríguez Rivero, Carlos Alberto Salas, Martín Herrera, Sergio Oscar Laboret
KEYWORDS:
Optimal Control, Neurocontroller, Inverted Pendulum, State Estimation, Stability
JOURNAL NAME:
Applied Mathematics,
Vol.8 No.11,
November
24,
2017
ABSTRACT:
In this paper, a practical analysis of stability by simulation for the effect of incorporating
a Kalman estimator in the control loop of the inverted pendulum
with a neurocontroller is presented. The neurocontroller is calculated by approximate
optimal control, without considering the Kalman estimator in the
loop following the Theorem of the separation. The results are compared with
a time-varying linear controller, which in noiseless conditions in the state or
in the measurement has an acceptable performance, but when it is under
noise conditions its operation closes into a state space range more limited
than the one proposed here.