Data-Driven Adaptive Control: An Incremental Triangular Dynamic Linearization Approach

Abstract

This brief investigates the data-driven control problem for a class of nonlinear systems. An incremental triangular dynamic linearization (ITDL) model is presented to equivalently describe these systems, and it contains three parameters including a single time-varying parameter, all of which have clear physical meanings. The ITDL model has obvious advantages compared with the existing partial form dynamic linearization (PFDL) model, although it is a special case of the PFDL model in form. An online estimation algorithm is designed for the single time-varying parameter, and then an adaptive control law is designed. The convergence and stability of the closed-loop system are analyzed, and the corresponding explicit conditions are derived. Finally, a comparative simulation study is conducted to illustrate the effectiveness of the proposed method.