l2–l∞ Filtering for 2-D Markov jump systems in FM model under asynchronous switching

Wang Xinwei,Zhang Ying,Zhang Rui
DOI: https://doi.org/10.1109/ChiCC.2015.7259894
2015-01-01
Abstract:This paper is devoted to the design of l2-l∞ filters for a class of 2-D Markov jump systems under asynchronous switching. The asynchronous behavior is considered to be random mismatch between system modes and candidate filters. The mismatch character is described by a suitable stochastic variable satisfying the Bernoulli random binary distribution law. A stochastic parameter-dependent approach is developed to solve the addressed filtering problem. The resulting asynchronous filters can ensure that, for any admissible random mismatch, the filtering error system is stochastically stable and has a prescribed l2-l∞ performance index. Connection with the classical mode-dependent and mode-independent filtering on the l2-l∞ performance is thoroughly discussed. A numerical example is provided to show the validity and potential of the results derived.
What problem does this paper attempt to address?