|
Adaptivní prediktivní LQ řízení s omezením
Böhm, Josef
Controller design for real systems must consider levels of signals in the system. This contrasts with the most design techniques that are based on linearity assumption. The Model predictive control (MPC) is a technique able to consider desired levels of signals. Paper deals with the situation in LQ controller design. It is shown that the latest results in MPC control can be also considered as an LQ approach.
|
| |
|
Klasifikace obrazu - návrh klasifikátoru
Klimešová, Dana ; Ocelíková, E.
This paper presents overview of the mostly used classification approaches and shows the possibility of the approach based on the maximum entropy theory that can optimally describe the spatial data distribution and gives actual error estimation. The approach is highly recommended in case we are not able to test the assumptions of given method.
|
|
Factorizace 2-D polynom
Ježek, Jan ; Zezula, P.
Many algorithms are solving the problem of 1-D factorization. One of them was used and generalized to 2-D. This simple step brings new different difficulties up. Nevertheless, by no means could be the matter of 2-D factorization announced as completely investigated. Just a case with explicitly defined causality in both variables is offered in this paper. The numerical implementation of the algorithm below is the matter of ongoing labor.
|
| |
|
Studie prediktivních řídicích algoritmů pro paralení struktury robotů
Belda, Květoslav ; Böhm, Josef ; Valášek, M.
This paper focuses on several possible variants of discrete predictive algorithms for design of control actions applied to the parallel robot structures; i.e. absolute, incremental, incremental with nonlinear simulative prediction and predictive algorithm, used for the planning of quadratically-optimal trajectories. The control design is based on mathematical models describing the mechanical parallel robotic structures. All algorithms are illustrated by simulation in MATLAB - SIMULINK environment.
|