Adaptive backstepping with a high-order tuner (Q5949059)

Adaptive backstepping and high-order tuners were proposed and investigated actively by several researchers in the 1990s, and they are undoubtedly important results in the history of adaptive control. This work combines those two methodologies to construct adaptive nonlinear control systems. One of the features of high-order tuners is that the time-derivatives of tuning parameters are obtained up to certain orders, and that simplifies the structure of adaptive control schemes. However, the direct application of a high-order tuner to the backstepping procedure sometimes gives rise to the requirement of time-derivatives of states, which are not available for measurement. Thus, in the first half of the paper, the authors propose the modified high-order tuner where time-derivatives of states are not included in the computation of time-derivatives of tuning parameters. Then, those techniques are applied to the construction of adaptive nonlinear control systems via backstepping. The modified high-order tuners are utilized in each iterative step of backstepping procedures to stabilize nonlinear plants in parametric-strict-feedback forms. Owing to the specified forms of high-order tuner algorithms, tuning functions (or virtual adaptation laws) are not included in the synthesis of virtual and actual control signals, but the time-derivatives of tuning parameters are contained directly in each iterative step. Those properties simplify the control schemes compared with the conventional backstepping. The paper shows the alternative version of adaptive backstepping without tuning functions and without overparametrizations.