Stability conditions for the sequential design of non-diagonal multivariable feedback controllers (Q4783552)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1842033
Language Label Description Also known as
default for all languages
No label defined
    English
    Stability conditions for the sequential design of non-diagonal multivariable feedback controllers
    scientific article; zbMATH DE number 1842033

      Statements

      Stability conditions for the sequential design of non-diagonal multivariable feedback controllers (English)
      0 references
      0 references
      0 references
      18 May 2003
      0 references
      MIMO systems
      0 references
      closed-loop internal stability
      0 references
      unstable Smith-McMillan pole-zero cancellations
      0 references
      feedback loop
      0 references
      sequential design
      0 references
      multivariable control systems
      0 references
      disturbance rejection
      0 references
      tracking
      0 references
      off-diagonal controllers
      0 references
      non-minimum phase system
      0 references
      Consider the multivariable system described by \[ {\mathbf Y}(s) ={\mathbf P}(s){\mathbf U}(s) +{\mathbf D}(s){\mathbf W}(s), \tag{1} \] where \({\mathbf Y}(s)\in S^{n\times 1}\) is the vector of system outputs, \({\mathbf U}(s)\in S^{n\times 1}\) is the vector of system inputs, \({\mathbf W}(s)\in S^{n\times 1}\) is the vector of disturbance inputs, \({\mathbf D}(s)\in S^{n\times n}\) is the disturbance transfer function matrix and \({\mathbf P}(s)\in S^{n\times n}\) is the invertible plant transfer function matrix. Using unity feedback, \({\mathbf U}(S)={\mathbf G}(s)[{\mathbf R}(s)-{\mathbf Y}(s)]\), where \({\mathbf G}(s)\in S^{n\times n}\) is the feedback controller matrix and \({\mathbf R}(s)\in S^{n\times 1}\) is the vector of reference commands, the design objective is to achieve closed-loop stability, rejection of the external disturbance \({\mathbf W}(s)\), and tracking of \({\mathbf R}(s)\). Sufficient conditions to guarantee closed-loop internal stability for system (1) are developed in this paper. The proposed criteria explicitly consider the occurrence of unstable Smith-McMillan pole-zero cancellations within the feedback loop. The sequential design based on the proposed criteria thus ensures that internal stability is achieved. In addition, the proposed stability conditions make it possible to expand the class of systems considered to include MIMO systems with non-minimum phase Smith-McMillan zeros. The sequential design based on the presented stability criteria has several advantages. First, the conditions clearly establish when off-diagonal controllers are required to achieve internal stability. Another advantage of the proposed design technique is that the resulting controller structure is chosen based upon the stability requirements of the closed-loop system. For example, only the off-diagonal controllers that are needed to achieve the pre-specified stability objectives are used, thereby reducing the implementation complexity of the control law. However, one disadvantage of the proposed sequential design approach is that closed-loop stability can only be guaranteed during the design of the final loop according to some of the sufficient conditions. Although design practices that can be utilized to avoid violating these conditions are described, these conditions must still be verified upon completion of the design. The application of the stability criteria is demonstrated in a design example for an unstable, non-minimum phase (\(2\times 2\)) system.
      0 references

      Identifiers