Page 194 - Innovations in Intelligent Machines
P. 194

186    R.W. Beard
                                  1740
                                                                                  actual
                                  1730                                            estimated
                                  1720
                             h (m)
                                  1710
                                  1700
                                  1690
                                        0         5        10        15        20       25

                                    18
                                                                                  actual
                                    16                                            estimated
                                    14
                             V  (m/s)
                              a
                                    12
                                    10
                                    8
                                      0         5         10         15        20        25
                                                             time (sec)
                           Fig. 5. Actual and estimated values of h and V a after low pass filtering the pressure
                           sensors and inverting their models

                                                 ˙ v + ru − pw − g cos θ sin φ
                                        y accel,y =                     + η accel,y
                                                            g
                                                  ˙ w + pv − qu − g cos θ cos φ
                                         y accel,z =                     + η accel,z .
                                                            g
                           and that in unaccelerated flight ˙u =˙v =˙w = p = q = r = 0, we get that
                                                C(s){y accel,x } =sin θ

                                                C(s){y accel,y } = − cos θ sin φ
                                                C(s){y accel,z } = − cos θ cos φ.

                           Solving for φ and θ we get
                                                  !  C(s){y accel,y }  "
                                     ˆ
                                     φ accel = tan −1                                      (19)
                                                    C(s){y accel,z }

                                      ˆ
                                     θ accel = tan −1        C(s){y accel,x }      .       (20)
                                                                 2
                                                      C(s){y accel,y } + C(s){y accel,z } 2
                              Figure 6 shows the actual and estimated roll and pitch angles during the
                           sample trajectory using this scheme. Note that the sample trajectory severely
                           violates the unaccelerated flight assumptions. Clearly, model inversion does
   189   190   191   192   193   194   195   196   197   198   199