適応型モデル追従制御系の誤差特性改善に関する一考察 : 特にBalestrino適応制御系について

概要

論文の詳細を見る
The increased performance on an adaptive model-following control system (AMFC) for robotic manipulators, which is developed by Balestrino et al via Popov's hyperstability theory, has been discussed from the viewpoint of especially improving the steady error characteristics e (∞). Using a 1-degree of freedom DC servo motor system, the followings have been obtained both from the simulation and experimental points of view. The steady error e (∞) seen in the AMFC where the factors (p and q) characterizing the discontinous controlling input signals for robust compensation is larger, can be satisfactorily explained from the realation (the size of error ‖e‖^2 ∝c_1p+c_2q), with positive constant c_i. With the saturation function (sat (υ)) replacement in place of the discontinous input function (sgn (υ)), the steady error e (∞) can be greatly improved, in addition to the elimination effect of the chattering during the sliding mode, and in the function sat (υ) there exists an optimal gradient to minimize the steady error e (∞). The robustness against the system parameter variations can also be well explained based on the optimal gradient in the function sat (υ).
日本シミュレーション学会の論文
1992-12-15