Page 304 - Adaptive Identification and Control of Uncertain Systems with Nonsmooth Dynamics
P. 304

Adaptive Estimation and Control of Magneto-Rheological Damper for Semi-Active Suspensions  307


                            Then the parameters η and P can be appropriately selected, such that the
                            performance requirement of the road holding (19.21) can be guaranteed.
                               Finally, we can obtain the upper bound of suspension spaces as





                                 |x 1 − x 3 | ≤ 2V 1   +  (V(0) + β α) λ min (P) ≤ z max  (19.42)
                               Hence, the suspension movement limitation (19.21) can be fulfilled if
                            the parameters  ,k s ,  1 ,σ 1 ,η,P are designed appropriately.


                            19.4 SIMULATIONS

                            In this section, numerical simulations are provided to illustrate the effec-
                            tiveness of the proposed estimation and control algorithms. The parameters
                            of the MR damper and quarter-car model are given as: m s = 600 kg,
                            m us = 60 kg, k s = 18000 N/m, k sn = 1000 N/m, k t = 200000 N/m, b f =
                            1000 Ns/m, b e = 2500 Ns/m, b c = 2200 Ns/m, c 0 = 810.78 Ns/m, c 1 =
                            13.76 s/m, θ I = 457.04 N/A, k 1 = 10.54 1/m, k 0 = 620.79 N/m, z max =
                            0.15 m.
                               The following two cases are simulated:

                            Case 1 (Adaptive parameter estimation of MR damper):Inthissim-
                            ulation, only the MR damper dynamics (19.11)isconsidered to show the
                            online modeling method (19.14). Thus, we set the velocity of piston as
                            ˙ z = 0.6cos(6t) and the input current as I = 2. The estimation performance
                            of the gradient method and the proposed method are compared. For fair
                                                                                           T
                            comparison, the initial simulation conditions are set as θ(0) =[0,0,0.001] .
                            The simulation parameters are set as   = 30diag[0.065,0.53,6.4] and k =
                            0.001,  = 1as [21]. One may find from Fig. 19.4 that the velocity-force
                            curves with the estimated parameters are very close to its nominal coun-
                            terparts. It is clearly shown that the estimated model based on (19.14)can
                            capture the essential dynamics of the realistic MR damper. This implies that
                            the estimated parameters converge to their true values.

                            Case 2 (Adaptive control with external road disturbance): The pro-
                            posed control and estimation are also simulated under the external road
                            disturbance, which is given as follows



                                                  h    1 − cos     2πV s  t , 0 ≤ t ≤  l

                                                  2          l
                                           z r =                        l  V s        (19.43)
                                                  0                 t ≥
                                                                       V s
   299   300   301   302   303   304   305   306   307   308   309