Premium
Self‐falsification in multimodel unfalsified adaptive switching control
Author(s) -
Nouri Manzar Mojtaba,
KhakiSedigh Ali
Publication year - 2017
Publication title -
international journal of adaptive control and signal processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.73
H-Index - 66
eISSN - 1099-1115
pISSN - 0890-6327
DOI - 10.1002/acs.2796
Subject(s) - control theory (sociology) , benchmark (surveying) , monotonic function , reset (finance) , controller (irrigation) , constraint (computer aided design) , sequence (biology) , process (computing) , function (biology) , engineering , nonlinear system , stability (learning theory) , control (management) , computer science , control engineering , mathematics , artificial intelligence , mathematical analysis , financial economics , genetics , biology , operating system , geodesy , quantum mechanics , evolutionary biology , machine learning , agronomy , mechanical engineering , physics , economics , geography
Summary This paper addresses a multimodel unfalsified adaptive switching control with finite fixed time window cost function by utilizing a self‐falsification strategy. A closed‐loop stability proof is provided, and it is shown that the forgetting factor employed with finite fixed windowed cost function improves the closed‐loop performance. Furthermore, it is shown that the unfalsified adaptive control with nonmonotone cost function is unable to select the appropriate controller, and a new reset strategy is proposed to resolve this problem. The γ sequence monotonicity in the linear increasing cost‐level algorithm causes a performance deterioration, and a γ sequence reset is introduced for performance enhancement. Effectiveness of the proposed method is investigated for a nonlinear pH neutralization process and the 2‐cart benchmark example.