Active Disturbance Rejection Control of High-Order Flat Underactuated Systems: Mass-Spring Benchmark Problem

Abstract

A custom active disturbance rejection control (ADRC) is proposed for trajectory tracking in uncertain underactuated systems. By utilizing a particular property of the differentially flat plant model, a robust output-feedback controller with a novel cascade of extended state observers (ESO) is introduced. It effectively deals with the problem of governing high-order plants without over-amplification of the measurement noise, typically seen in conventional single high-gain observer-centered control approaches. The proposed solution is based on full utilization of the information already available about the governed system, without necessity for additional measurement devices. In order to be easily implementable, it assumes only limited knowledge of the system model and is expressed in an industry familiar error-based form with a straightforward tuning method. A representable two-mass three-spring benchmark problem is used throughout the paper to convey the proposed idea and evaluate it experimentally on a laboratory testbed.