Control from the angle of the Zermelo axiom of choice (Q1883844)

It has been shown by \textit{A. B. Kharazishvili} [Soobshch. Akad. Nauk Gruz. SSR 94, 533--536 (1979; Zbl 0409.28001)] (not cited here) that the axiom of choice is a necessary and sufficient condition for the existence of sections for a general system seen as a subset of \(X\times Y\) \((X\) is the base set of states and \(Y\) is the set of controls). The linear case requires a non-denumerable axiom of choice, a property which seems to have been omitted by \textit{Lin Yi} and \textit{Ma Yanghao} in [Int. J. Gen. Syst. 18, 143--154 (1990; Zbl 0729.93005)] (also not cited here). Here the author generalizes (in some sense) the situation to general fibered spaces foliated by the control sets. Reviewer's comments: (i) \(u\) should replace \(U\) in the last formula which moreover does not necessarily describe a feedback or is just constant for identical fibers. (ii) The axiom of choice is necessary not only for the existence of feedbacks, as seen in this and the above mentioned publications, but for any kind of control for systems defined on sets not necessarily having the above structures. Control theory certainly evaporates without the axiom of choice and in any control system the symbol denoting the control should involve the choice function (omitted by ignorance or abuse of notation). This is why it is not allowed in general to speak of the ``control parameter'' as is done so often.