Open AccessStability-Based Parameter Selection for Data-Driven Model-Free Adaptive Controllers
Author(s) -
Kai Deng,
Chunhua Yang
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1302/3/032055
Subject(s) - stability (learning theory) , basis (linear algebra) , selection (genetic algorithm) , control theory (sociology) , computer science , range (aeronautics) , adaptive control , estimation theory , function (biology) , mathematical optimization , algorithm , control (management) , mathematics , engineering , artificial intelligence , machine learning , geometry , evolutionary biology , biology , aerospace engineering
The algorithm of model-free adaptive control normally includes a control algorithm and a pseudo-partial derivative estimation algorithm. The selection of each parameter in the algorithm directly affects the control performance and stability of the system. Two artificial additional parameters have been introduced for the algorithm, but the corresponding theoretical basis needs to be improved. Although the range of artificial additional parameters is given in the control algorithm, the stability condition of the system cannot be guaranteed. In this paper, the theoretical basis for introducing two additional parameters is provided by modifying the cost function. Furthermore, the selection conditions of control parameters are given such that the controlled system stable and not entirely dependent on the trial-and-error. The simulation results are given to confirm the theoretical findings.