Page 103 - Dynamics and Control of Nuclear Reactors
P. 103

8.4 Control of a zero-power reactor   97




                     The closed-loop transfer function of a system with transfer function, 1/(s+a) and
                  with integral controller is given by

                                            YsðÞ     K i
                                               ¼                                (8.14)
                                                   2
                                            Xs ðÞ  ð s + as + K i Þ
                  Note that the closed-loop system becomes a second-order dynamic system, with the
                  potential for the system response to have an oscillatory characteristic, as K i becomes
                  large. The system error is given by
                                                  ð
                                                  ss + aÞXsðÞ
                                            EsðÞ ¼  2                           (8.15)
                                                 ð s + as + K i Þ
                  The steady-state error for a unit step input is
                                                 lim
                                           etðÞ ¼   sE sðÞ ¼ 0                  (8.16)
                                                s ! 0
                  The integral control thus results in a zero steady-state error for the above system.
                     For a¼0.02, the closed-loop transfer function is given by.

                                           Ys ðÞ     K i
                                              ¼                                 (8.17)
                                                 2
                                           XsðÞ  ð s +0:02s + K i Þ
                  For K i > 0.0001, the roots of the denominator polynomial are complex, resulting in
                  an oscillatory response of to a step input. Thus, extreme caution must be exercised in
                  choosing K i for a general system. A large value of K i results in a highly oscillatory
                  behavior, and quite likely makes the system unstable. Thus, the choice of an appro-
                  priate value of K i is important, rather than the use of an arbitrary absolute value.



                  8.3.8 Advanced controllers
                  Advanced control strategies use computed estimates of future consequences of cur-
                  rent control actions rather than current consequences as in classical control.
                  Advanced model-based controllers are necessary in some applications (such as con-
                  trolling missile trajectories). Sophisticated advanced control software is available,
                  but is not currently needed for reactor control. The next generation reactors may find
                  use for advanced control implementation.




                  8.4 Control of a zero-power reactor

                  This section illustrates system control by simulation of a zero-power reactor with
                  proportional control, integral control, and proportional plus integral (PI) control.
                  Included are responses to a reactivity step input and a power set point change.
                  The reader should note the effectiveness of the controller in driving the response
                  to the desired final value.
   98   99   100   101   102   103   104   105   106   107   108