LPVSYNOPTIONS - Create a options object for LPV synthesis and analysis

Contents

Syntax

  opt = lpvsynoptions
  opt = lpvsynOptions(Name1,Value1,Name2,Value2,...)

Description

opt = lpvsynOptions(Name1,Value1,Name2,Value2,...) creates a options object for parameter-varying synthesis and analysis. The lpvsynOptions object is used to specify the parameters of the optimization routines used in the synthesis and analysis functions: lpvsyn, lpvmixsyn, lpvncfsyn, lpvloopshapesyn, lpvestsyn, lpvsfsyn, lpvnorm, lpvstochsyn, lpnvnorm, lpvwcgain.

opt = lpvsynOptions creates an lpvsynOptions object initialized with default values.

The options are set using NAME, VALUE pairs, i.e. the options property specified by the character string NAME is set to VALUE. The setable options properties are specified below. The default choice is specified in brackets.

%--------------------------------------------------------------------------
%     NAME          |   VALUE     |    Description
%--------------------------------------------------------------------------
%   'Solver'        | ['lmilab']  | Optimization solver to be used.
%   -----------------------------------------------------------------------
%   'SolverOptions' | []          | Options passed directly to the solver.
%   -----------------------------------------------------------------------
%   'SolverInit'    | []          | Initial decision variables for LMI solver.
%   -----------------------------------------------------------------------
%   'Gammalb'       | [1e-6]      | Lower bound on closed-loop induced L2 norm.
%   -----------------------------------------------------------------------
%   'Gammaub'       | [1e6]       | Upper bound on closed-loop induced L2 norm.
%   -----------------------------------------------------------------------
%   'Xlb            | [1e-6]      | X Riccati variable lower bounded by Xlb*I
%   -----------------------------------------------------------------------
%   'Xub            | [1e6]       | X Riccati variable upper bounded by Xub*I
%   -----------------------------------------------------------------------
%   'Ylb            | [1e-6]      | Y Riccati variable lower bounded by Ylb*I
%   -----------------------------------------------------------------------
%   'Yub            | [1e6]       | Y Riccati variable upper bounded by Yub*I
%   -----------------------------------------------------------------------
%   'Method         | ['BackOff'] | String specifying the solution method:
%                   |             |----------------------------------------
%                   | 'MinGamma'  | Minimize the closed-loop induced L2 norm.
%                   |             |----------------------------------------
%                   | 'MaxFeas'   | Maximize the feasibility of the X Riccati
%                   |             | subject to contraints Gammalb and Gammaub
%                   |             | on the closed-loop induced L2 norm.
%                   |             |----------------------------------------
%                   | 'BackOff'   | First solve the 'MinGamma' problem for
%                   |             | GammaOpt and then solve a second stage
%                   |             | 'MaxFeas' problem with
%                   |             | Gammaub = BackOffFactor*GammaOpt.
%                   |             | This two-stage solution improves
%                   |             | the numerical conditioning of the
%                   |             | controller reconstruction.
%                   |             |----------------------------------------
%                   | 'PoleCon'   | Constrain the closed-loop poles.
%   -----------------------------------------------------------------------
%   'BackOffFactor' | [1.2]       | Multiplicative factor ( >= 1 ) to back off
%                   |             | the minimum gamma when Method = 'BackOff'.
%   -----------------------------------------------------------------------