A robust MP-ADRC-based strategy for uncertain minimum phase systems

关键词:  Modified-plant ADRC · Uncertain systems · Minimum phase plants · Robust control · Extended state observer

Key words:  Modified-plant ADRC · Uncertain systems · Minimum phase plants · Robust control · Extended state observer

摘要:

This paper proposes an extension of the Modified-Plant ADRC (MP-ADRC) strategy to broaden its application to minimum phase dynamical systems. The main features of the MP-ADRC method are the inclusion of a constant gain in series with the plant output error and a linear filter in parallel with the overall error system. These structural changes do not influence the input/output dynamics of the original plant, but are intentionally introduced to modify the dynamics to be estimated by the extended state observer (ESO) and, thus, promote an increase in the robustness of the method. Some advantages can also be attributed to the proposed methodology, such as (i) the design procedures of both the controller and the ESO only require knowledge of the sign (±) of the plant input channel coefficient (or control gain); (ii) the plant control input is generated directly by a single ESO state variable. Despite the advantages and the characteristics of MP-ADRC mentioned earlier, closed-loop stability cannot be guaranteed when it is applied to dynamical systems that have finite zeros. To overcome this difficulty, this work introduces an extension in the MP-ADRC method. It basically consists of rewriting the minimum phase plant dynamics according to its relative order, and then follows with the design of the ESO by conveniently increasing the number of ESO state variables. The simulation results are also presented to illustrate the application of the proposed method.

Abstract:

This paper proposes an extension of the Modified-Plant ADRC (MP-ADRC) strategy to broaden its application to minimum phase dynamical systems. The main features of the MP-ADRC method are the inclusion of a constant gain in series with the plant output error and a linear filter in parallel with the overall error system. These structural changes do not influence the input/output dynamics of the original plant, but are intentionally introduced to modify the dynamics to be estimated by the extended state observer (ESO) and, thus, promote an increase in the robustness of the method. Some advantages can also be attributed to the proposed methodology, such as (i) the design procedures of both the controller and the ESO only require knowledge of the sign (±) of the plant input channel coefficient (or control gain); (ii) the plant control input is generated directly by a single ESO state variable. Despite the advantages and the characteristics of MP-ADRC mentioned earlier, closed-loop stability cannot be guaranteed when it is applied to dynamical systems that have finite zeros. To overcome this difficulty, this work introduces an extension in the MP-ADRC method. It basically consists of rewriting the minimum phase plant dynamics according to its relative order, and then follows with the design of the ESO by conveniently increasing the number of ESO state variables. The simulation results are also presented to illustrate the application of the proposed method.