Nonlinearizable single-input control systems do not admit stationary symmetries (Q1605167)

In the introductory section the authors introduce the notions of symmetry and stationary symmetry of a control system. The main theorem 1 deals with the latter in the case of affine single-input control systems. A ``linear approximation'' \((F,G)\) of the control system is introduced, which plays a role among the assumptions of theorem 1, but \(F\) does not have any geometric meaning. In section 2 a ``smooth feedback transformation'' leads from a given affine control system to another one. In fact, by means of a diffeomorphism of the state space -- a component of the feedback transformation --, two functions and two vector fields the authors define two new vector fields which describe the new system. Then the authors say that the feedback transformation preserves the given system if this one equals the new system, but they do not say what it means for two systems to be equal. The following link between symmetries and feedback transformations (proposition 1) is affected by this omission. In order not to omit defining equivalence, however, the authors would also have had to speak about the intrinsic feature of the control variables. Afterwards the authors write a Taylor series expansion of the given affine system, and obviously this expansion is considered as a new system, since it is denoted by an additional upper index \((\infty)\). The following sections contain the proof of the main theorem, but being based on the `new' system above, they are not understandable.