Smooth time-varying pure feedback control for chained non-holonomic systems with exponential convergent rate

The feedback stabilisation problem of non-holonomic chained systems and a novel feedback design scheme is proposed, which renders a smooth, time-varying, aperiodic, pure feedback control with exponential convergence rates. There are three main advantages with the proposed design. (i) In general, time-varying designs are mostly periodic and render asymptotic stability, whereas the proposed approach is aperiodic and have exponential convergent rates. (ii) A novel state scaling transformation is proposed. It shows that even though u(1) vanishes in regulation problems, intrinsic controllability of chained systems can be regained by judiciously designing the input u(1) and by applying the state scaling transformations. (iii) A class of memory functions is introduced into the control design, the controller dependency on the system's initial conditions in our previous work is removed and the control is a pure feedback. Moreover, the design is shown to be inversely optimal. Simulations and comparisons are conducted to verify the effectiveness of the proposed approach.