Page 94 - Biosystems Engineering
P. 94

Biosystems Analysis and Optimization      75


                                          Step response
                       1.5
                                  Loop shaped control
                                  P-controller
                        1
                      Amplitude




                       0.5




                        0
                         0        5      10       15      20      25
                                            Time (s)
               FIGURE 2.22  Step response for the closed-loop control system with
               designed return ratio L(s) and with a proportional controller (P = 1).


                   The step response for the closed-loop control system is illustrated
               in Fig. 2.22, together with the step response of a closed-loop system
               with a simple proportional controller (P = 1).
                   Note that the system with the shaped return ratio moves more
               rapidly toward the desired value but needs a longer time to settle in
               the 2 percent region around the steady-state value. The reason for this
               high settling time can be found in the presence of the zero at 0.1 rad/s,
               which has been introduced to obtain sufficient disturbance attenua-
               tion in the midfrequency range. This attenuation is illustrated for
               both feedback-controlled systems in Fig. 2.23 for a sinusoidal distur-
               bance with amplitude of 5 cm and frequency of 0.5 rad/s. The loop
               gain–shaped controller performs considerably better with an attenu-
               ation factor of 10, whereas the attenuation factor for the system with
               a proportional controller is only 1.3. This controller has thus traded
               some settling time performance (Fig. 2.22) for a considerable increase
               in disturbance attenuation performance (Fig. 2.23).



          2.6  Nonlinear Optimization Example: Cruise Control
                 on a Combine Harvester
               In this section, a real-life optimization example will be elaborated,
               more specifically, a cruise control system on a combine harvester.
               This is a nonlinear optimization problem that will be solved by
               model-based predictive control. Because some of the concepts used in
   89   90   91   92   93   94   95   96   97   98   99