Higher-order differential feedback control of a flexible-joint manipulator

Abstract

This paper presents a new control scheme for a flexible-joint manipulator using a higher-order differential feedback controller (HODFC). Two higher-order differential operators were designed and used to perform observations of both the reference input and the output of the manipulator, together with the requisite state derivatives. An error-based state-space model was then derived from the observed states. A pole-placement procedure with filtering was then used to drive the system error to zero. Practical controller implementation was carried out using the dSPACE real-time prototyping system. For the comparative validation of the performance of the HODFC with respect to a classical proportional-integral and proportional-integral-derivative (PID) controller, several experiments were undertaken. In these experiments, the step input, sine waves, kane trajectories, and external disturbances were applied to the controlled flexible-joint manipulator. The results showed that the HODFC controller eliminated disturbances within one second of occurrence, and produced superior kane trajectory tracking. Moreover, based on the root-mean-square tracking error criterion, the HODFC was observed to track both the sine and kane function trajectories with one-fourth the tracking error obtained with classical PID control.

Trajectory and Vibration Control of a Single-Link Flexible-Joint Manipulator Using a Distributed Higher-Order Differential Feedback Controller

“…The variable y, which could be used with its derivatives to achieve an alternative representation of a given dynamic system, is said to be the flat output of the system [21]. In Ref.…”

“…3, a recollection of the classical HODFC used in Ref. [21] is presented. The mathematical formulation of the structure of the distributed HODFC controller then follows.…”

“…Figure 3, used in Ref. [21], summarizes the structure of the HODFC controller. The classical HODFC consists of two higherorder differential (HOD) structures, feedback gains, and a filter with a very high forward gain.…”

“…These pseudomodel based control strategies have enabled a compromise between the cost of real-time controller implementation and the need for highly accurate models in model-based control. A pseudomodel technique recently applied to the control of FJMs, the higher-order differential feedback controller (HODFC), was reported to have achieved very good trajectory control [21]. The approach that forms the body of the current contribution is a further exploitation of the HODFC structure for achieving further performance improvements in the control of the manipulator.…”

Intelligent proportional integral control of a polar axis solar tracker

“…Whilst, such AI-based control strategies have been shown to sustain good tracking performances in the presence of system variations, they often require specialist skills for controller tuning. Recently, some simple control structures, the intelligent proportional integral controller (iPI) and the intelligent proportional integral derivative controller (iPID), based on a simple approximation of the system model have been proposed [15][16][17][18]. The use of approximate system models for the design of these iPI and iPID controllers suggest that, these so-called intelligent controllers may be insensitive to model parameter variations.…”