Neural controller for designing of nanopositioning systems

Bazylevych, O.; Hirnyak, Yuriy; Івахів, Орест

doi:10.1109/idaacs.2009.5342979

Cited by 3 publications

(1 citation statement)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Difference equation, which corresponds to the expression (6), at the k-th step of controlling, will look as follows: (7) specifies a single-layer structure (Fig. 1) [5,6] of the four-input neural network with one adder and one linear circuit of activation. The scheme of proportional-integral-differential discrete controller Here , the signal at its output is signal at the input of the first layer adder of the neural network and its weight coefficient, respectively (in this case…”

Section: Neurocontroller Designmentioning

confidence: 99%

Control system of robot movement

Hirnyak

Івахів

Nakonechnyi

et al. 2013

2013 IEEE 7th International Conference on Intelligent Data Acquisition and Advanced Computing Systems (IDAACS)

Self Cite

View full text Add to dashboard Cite

Development of modern technologies is related to an increasing complexity of the objects controled and hence the systems controlling them. In the most cases, automatic control systems consist of different nonlinear elements that significantly limit the capabilities of classical control theory in designing controllers. In recent decades, the methodology of neural networks has been increasingly used lately. These networks may provide easier solutions to various complex control problems. Moreover, their basic elements (i.e., neurons) are nonlinear, and that is why neural networks are essentially nonlinear systems which can perform various control tasks. In this paper are investigated some new proposed structures, i.e., with so-called two separate inputs which allow to improve the dynamic characteristics of control system as well.

show abstract

Section: Neurocontroller Designmentioning

confidence: 99%