A versatile controller design software for unstable systems
Introduction
Methods for design of stable linear plants are well established [1]. It is of no surprise that such approaches have also helped advancing the design of stable nonlinear systems [2]. However, there are many good practical problems that are inherently unstable. For example, the attitude control of a space craft is such a problem [3]. There are a number of techniques for control system design of unstable systems [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20]. The most attractive approaches (such as H∞ approach) are based on one form of factorization [21]; however, still research is in progress to systematize those approaches [22], [23]. One other approach that has received considerable attention is based on the Youla parameterization; however, systematic controller design procedures for model-matching in unstable plants using this approach have received limited attention [4]. The primary contribution of this paper is the developed software that accomplishes the controller design for unstable linear SISO plants. A demonstration example problem is presented, and the results are compared with other works reported in the open literature [4], [24]. The prior related software assumes that plants are stable [25]; and hence coprime factors are selected trivially as noted in the following section. However, if the plant is unstable, then a stable factorization involves selecting nontrivial stable factors; the corresponding optimum location of these factors must be determined. In this work, the prior software [25] is extended to allow determination of optimum pole locations of stable factors.
Section snippets
Background
In this paper, all rational functions are represented by capital letters with an indication that they depend on the complex frequency s or jω, polynomials are represented by small letters and an indication that they are functions of complex frequency s, all other small and capital letters are either constants or have an indication that are functions of time; variables may have subscripts and superscripts for clarity. Matrices and vectors are in bold letters.
The set of all controllers that
Controller design algorithm
The controller design algorithm is composed of six steps as described below.
- Step 1:
Derive a transfer function representing the desired reference linear model that exhibits the desired performance measures such as settling time, phase margin, gain margin and velocity error constant.
- Step 2:
Select the desired region of the s-plane that you would want to place the poles of the closed-loop system. It is recommended, to draw a vertical line in the s-plane at the pole of the transfer function of Step 1 that is
Demonstration example problem
Consider the following difficult to control plant [4], [24]:The design should achieve a velocity error constant equal to 10 s−1, a damping ratio of 0.5, and the natural frequency should be 1 rad/s.
To begin with, the following reference linear model is identified [29]:Following the above task, the initial locations of the coprime factor poles are determined. Arbitrarily, the initial locations of the coprime factor poles are all set to −0.5. Using the Matlab
Summary and conclusions
The goal of this research was to develop a new systematic approach for design of controllers for unstable linear plants. This goal is met. The approach is given in terms of a new controller design algorithm, and it is based on a two-layer optimization. In the inner layer optimization and for a given set of values for the pole locations of the lucid coprime factors, a sub-optimal design is obtained and the objective function is evaluated. In the outer layer optimization, a SIMPLEX optimization
References (30)
A PI-PD controller design for control of unstable and integrating processes
ISA Trans
(2003)- et al.
Linear unstable plants with saturating actuators: robust stabilization by a time varying sliding surface
Automatica
(2007) - et al.
Robustness issues in H2 optimal control of unstable plants
Syst Control Lett
(2006) - et al.
Anti-windup control design for exponentially unstable LTI systems with actuator saturation
Syst Control Lett
(2004) - et al.
Robust prediction-based control for unstable delay systems: application to the yaw control of a mini-helicopter
Automatica
(2004) - et al.
Extended H∞ control–H∞ control with unstable weights
Automatica
(2000) A new software tool for design of linear compensators
Adv Eng Software
(2008)Computation of lucid factors for Bezout identity
Adv Eng Software
(2003)Modern control systems
(1989)- et al.
A systematic single-range controller synthesis procedure for nonlinear and multivariable liquid propellant engines
Aerosp Sci Technol
(2006)