scientific article; zbMATH DE number 3925675

From MaRDI portal
Publication:3700364





[For the entire collection see Zbl 0568.00025.] For the following optimal control problem and its special case (a quadratic problem): \[ \inf_{\left[ \begin{matrix} Q\\ u\end{matrix} \right]\in V}(\| Q-R\|^ p_ 1+K\| u\|^ p_ 2)\quad^{1/p}\quad or\quad \inf_{\left[ \begin{matrix} Q\\ u\end{matrix} \right]\in V}\max \{\| Q-R\|_ 1,\quad K\| u\|_ 2\} \] where: \(V=\{\left[ \begin{matrix} Q\\ u\end{matrix} \right]\in E_ 1\times E_ 2:\) \(AQ+Bu+f=0\), \(u\in U\subset E_ 2\}\), \(K\geq 0\), \(1\leq p<\infty\), \(A\in {\mathcal L}(E_ 1,E_ 3)\), \(B\in {\mathcal L}(E_ 2,E_ 3)\), \(R\in E_ 1\), \(f\in E_ 3\), \(E_ i\) are normed spaces with \(\| \cdot \|_ i\), the author obtains upper and lower bounds for the performance index and the optimal control. The obtained results which generalize those given in another paper of the author [Abh. Akad. Wiss. DDR, Abt. Math. Naturwiss. Tech. 1981, Nr. 2N, 19-28 (1981; Zbl 0474.49027)] and the author and \textit{S. Kossert} [Z. Anal. Anwend. 1, No.3, 13-21 (1982; Zbl 0531.49013)] are compared with results of \textit{E. Aronoff} and \textit{C. T. Leoudes} [Inst. J. Syst. Sci. 7, 17-25 (1976; Zbl 0317.49024)] and \textit{W. L. Chan} and \textit{L. F. Ho} [J. Math. Anal. Appl. 70, 530-545 (1979; Zbl 0418.49022)]. Upper and lower bounds can be applied for estimating how close a sub- optimal control is to the optimal without knowing the latter and in proving the bang-bang-ness of optimal controls [see Z. Anal. Anwend., loc. cit.]. The results of the paper are applicable for lumped as well as distributed parameter systems.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3700364)