Fakulta elektrotechniky a informatiky / Faculty of Electrical Engineering and Informatics

Search Results

Nyní se zobrazuje 1 - 2 z 2

Omezený přístup
Predictive Controller Based on Feedforward Neural Network with Rectified Linear Units
(Springer Nature Switzerland AG, 2019) Doležel, Petr; Honc, Daniel; Štursa, Dominik
This paper deals with a nonlinear Model Predictive Control with a special form of the process model. Controller uses for the prediction purposes a locally valid linear sub-models. The sub-models are obtained from a neural model with the rectifier activation function in hidden neurons. Simulation example is given to demonstrate proposed solution - neural model design and predictive controller application.
Otevřený přístup
Predictive Control of Nonlinear Plant Using Piecewise-Linear Neural Model
(IEEE (Institute of Electrical and Electronics Engineers), 2017) Honc, Daniel; Doležel, Petr; Gago, Lumír
A special form of a predictive controller is presented in this paper. Based on previous authors' work, a piecewise-linear neural model of nonlinear plant to be controlled is adopted to local linearization. The linearized model is then used for control action evaluation using a predictive controller. Although the linearization using piecewise-linear neural network is simple and efficient, it provides the model in a nonstandard form. Therefore, the proposed predictive controller is designed in order to handle that nonstandard model without any customization. At the end of the paper, the illustrative example demonstrates the main features of the introduced solution.