Title :
New switched filtering method for recurrent neural networks
Author_Institution :
Fac. of the Dept. of Automotive Eng., Seoul Nat. Univ. of Sci. & Technol., Seoul, South Korea
Abstract :
In this paper, we propose a new robust filtering method for switched neural networks via input/output-to-state stability (IOSS) approach. This robust filtering method guarantees that the filtering error system is asymptotically stable and input/output-to-state stable for the external disturbance. The unknown gain matrix of the proposed filter can be obtained by solving a set of linear matrix inequalities (LMIs), which can be easily facilitated by using some standard numerical packages.
Keywords :
linear matrix inequalities; recurrent neural nets; stability; external disturbance; filtering error system; gain matrix; input/output-to-state stability; linear matrix inequalities; recurrent neural networks; robust filtering method; switched filtering method; switched neural networks; Asymptotic stability; Biological neural networks; Filtering; Linear matrix inequalities; Stability analysis; Switches; input/output-to-state stability; robust filtering; switched neural networks;
Conference_Titel :
Uncertainty Reasoning and Knowledge Engineering (URKE), 2011 International Conference on
Conference_Location :
Bali
Print_ISBN :
978-1-4244-9985-4
Electronic_ISBN :
978-1-4244-9984-7
DOI :
10.1109/URKE.2011.6007842