Page 948 - The Mechatronics Handbook
P. 948

0066_frame_C31.fm  Page 11  Wednesday, January 9, 2002  7:49 PM









                       concept of passivity, which is really an abstract representation of the idea of energy dissipation in both
                       linear and nonlinear systems. Passive systems are most common in mechanical and electrical engineering
                       applications.  A mechanical system consisting of masses, springs, and viscous dashpots is a common
                       example for a passive system. We now give the following definitions.
                       Definition: Truncation of a signal
                       Let Y be the space of real-valued functions defined on [0,  ∞). Let  x be an element of Y. Then  the
                       truncation of x at some T > 0 is defined by

                                                          x t()  for 0 ≤ t ≤  T
                                                  x T t() =  
                                                          0    for t >  T


                       Definition: Extended space
                       If X is a normed linear subspace of Y, then the extended space X e  is defined by the set

                                              { x ∈  Y : x T ∈ X for some fixed T ≥ 0}

                       The extended L 2  space is denoted by L 2e .

                       Definition: Scalar product between two signals
                       The scalar product between two real-valued time signals x, y ∈L 2e  is defined as


                                                                       T
                                            〈 x | y〉 =  ∫ ∞ x t()y t() t =  ∫  T  x t()y t() t
                                                       T
                                                               d
                                                                              d
                                                     0               0
                       Definition: Passive systems
                       A system with input u(t) and output y(t) is passive if
                                                          〈|〉    0
                                                           yu ≥
                                                   ∃
                       The system is input strictly passive if     > 0 such that
                                                         yu ≥
                                                        〈|〉      u  2
                       The system is said to be output strictly passive if     > 0 such that∃


                                                         yu ≥
                                                        〈|〉      y  2
                       31.8 Adaptive Observers and Output Feedback Control

                                                                                    15
                       We now state the nonlinear adaptive observer problem formulated by Besancon :
                                                        (
                                                                 (
                                                   x ˙ =  fx, u, t) +  gx, u, t)q

                                                   y =  h x()                                   (31.15)
                                              ∞
                       where functions f and g are C  with respect to all their arguments and q is a constant and unknown
                       parameter. Variables x, u, and y respectively denote the state, input, and output vectors. The input signals
                       may be assumed to belong to some set of measurable and bounded functions. By the phrase adaptive
                       observer, we imply the problem of reconstructing a state estimate  (t) using the input u and output y
                                                                          x ˆ
                       in the presence of the unknown parameter q such that lim t→∞  x ˆ t() x t()–  =  0 . The conditions for the

                       ©2002 CRC Press LLC
   943   944   945   946   947   948   949   950   951   952   953