附录
CONTROL, PID CONTROL, AND
ADVANCED FUZZY CONTROL
FOR SIMULATING A NUCLEAR
REACTOR OPERATION
XIAOZHONG LI  and DA RUAN*
elgian Nuclear  Research  Centre  (SCKoCEN
Boeretang  200, 8-2400 Mol, Belgium
(Received 15 March  1999)
reactor then    Based  on  the  background  of  fuzzy  control  applications to  the  first  nuclear reactor  in Belgium (BRI) at  the Belgian Nuclear  Research Centre (SCK.CEN), we  have  made a real fuzzy logic control de
mo model. The demo model is suitable for us  to test and com- pare  some new  algorithms  of  fuzzy control  and  intelligent systems, which  is 
advantageous because it is always difficult and time-consuming, due to safety aspects, to do all
experiments in a  real nuclear environment. In  this  paper, we  first  report briefly on the
construction  of  the  demo model, and  then  introduce  the  results of  a  fuzzy control,
a proportional-integral-derivative (PID)  control and an advanced fuzzy control, in which
the advanced  fuzzy control  is a  fuzzy control with an adaptive function that can 
Self-regulate the  fuzzy control rules. Afterwards, we present a comparative study of those
three methods. The results have shown that fuzzy control has more advantages  in terms
of  flexibility, robustness,  and easily updated facilities with  respect to the PID control of
the demo model, but  that PID control has much  higher regulation  resolution due to its
integration term. The adaptive fuzzy control can dynamically adjust the rule base,
therefore it is more robust and suitable to  those very uncertain occasions.
Keywords:  Fuzzy  control; PID  control;  fuzzy adaptive control; nuclear  reactor
I  INTRODUCTION
    Today  the  techniques  of fuzzy  logic control are very mature  in most
engineering  areas,  but  not  in  nuclear  engineering,  though  some research has been done (Bernard, 1988; Hah and Lee,  1994; Lin et al. 1997; Matsuoka,  1990). The main  reason  is  that  it  is  impossible  to do experiments in nuclear  engineering as easily as in  other  industrial areas.  For  example, a  reactor  is  usually  not  available  to  any  individual.  Even  for  specialists  in  nuclear  engineering, an  official  licence for doing  any on-line test  is necessary. That  is why we  are still conducting projects such as "fuzzy  logic control application"  in BRl  (the first  nuclear  reactor  in  Belgium)  (Li  and  Ruan,  1997a; Ruan,  1995; Ruan and Li, 1997; 1998; Ruan and van der Wal,  1998).  In the framework  of  this  project, we  find  that  although  there  are  already many fuzzy  logic  control applications,  it  is difficult  to  select  the most  sui-
table for testing and comparison of our algorithms. Moreover, due to the  safety  regulations  of the nuc
lear  reactor,  it  is  not  realistic  to perform many experiments in BRl. In  this situation, we  have to conduct part of  the pre-processing experiments outside  the reactor, e.g., com-
parisons of  different  methods  and  the  preliminary  choices  of  the parameters.  One  solution  is  to  make  a  simulation  programme  in  a computer,  but  this  has  the  disadvantage  that  in  which,  however, the  real  time  property  cannot  be  well  reflected.  Therefore  another solution  has  adopted,  that  is,  we  designed  and  made  a  water-level
control  system, referred  to  as  the  demo model, which  is  suitable  for our  testing  and  experiments.  In  particular,  this demo model  (Fig.  1) is designed  to simulate  the power control principle of BRl  (Li et al., 1996a,b; Li  and Ruan, 1997b).
    In this demo model, our goal was to control the water level in tower TI at a desired level by means of  tuning VL  (the valve for large control tower  T2)  and VS  (the valve  for  small control  tower  T3). The pump keeps on working to supply water  to T2  and  T3.  All  taps are for manual tuning at this  time. VI  and  V2  valves are used  to control the water levels  in  T2  and T3  respectively. For example, when  the water  level  in T2  is lower than photoelectric switch sensor 1  then the on-off  valve V, will be opened (on), and when the water level  in T2  is higher  than photoelectric switch senso
r 2 then the on-off  valve Vl will be closed (off). The same is  true  of  V2.  Only when  both  VI and  V2  are closed  V3 will be  opened,  because  it  can  decrease  the  pressure  of  the  pump  and thereby prolong  its working life. The pressure sensor is used  to detect the height of water  level  in  TI.  So for TI,  it  is  a dynamic system with two entrances and one exit for water flow.