Title :
Piecewise-Linear Neural Network: A Tool for Modeling of the Processes to Be Controlled
Author :
Doleel, Petr ; Rozsival, Pavel ; Marika, Martin
Author_Institution :
Fac. of Electr. Eng. & Inf., Univ. of Pardubice, Pardubice, Czech Republic
Abstract :
The article introduces a new technique for process modeling. On the basis of the fact that once nonlinear problem is modeled by piecewise-linear model, it can be solved by many efficient techniques, the result of introduced technique provides a set of linear equations. Each of these equations is valid in some region of state space and together, they approximate whole nonlinear process.
Keywords :
neurocontrollers; nonlinear control systems; piecewise linear techniques; state-space methods; linear equations; nonlinear problem; piecewise-linear model; piecewise-linear neural network; process modeling; Artificial neural networks; Mathematical model; Neurons; Polynomials; Process control; Training; artificial neural network; piecewise-linear activation function; process modeling;
Conference_Titel :
Computational Science and Computational Intelligence (CSCI), 2014 International Conference on
Conference_Location :
Las Vegas, NV
DOI :
10.1109/CSCI.2014.76
