The LQG Control Algorithms for Nonlinear Dynamic Systems
Katalin György · Procedia Manufacturing · 2019
In this paper I will discuss some theoretical and practical aspects of the linear–quadratic–Gaussian (LQG) control algorithms implementation. The LQG control problem is one important optimal control problem and it can be used for systems where additive Gaussian noise appears on the system (output and state variables) or where the variable states are not complete measurable. The LQG control is a combination between one state estimator algorithm and one linear–quadratic regulator (LQR) algorithm. If the system to be controlled can be characterized by a linear system model, then the linear Kalman filter can be used. If the system is nonlinear, the Extended Kalman Filter (EKF) can be applied or other substitute methods such as the Unscented Kalman filter (UKF) algorithm.