Appeal No. 1996-1922
Application 08/110,506
regions represented by numerals I and III, and controller 34 is capable of assuring stability in the high
process gain region II without sacrificing performance in the low process gain regions I and III. In
addition, the auxiliary variable value 46 allows controller 34 to differentiate between a positive process
gain region represented by numeral I and a negative process gain region represented by numeral II
shown in Figure 2b. The controller is capable of providing a process control signal 50 that provides
negative feedback in one region and positive feedback in the other, respectively.
Fuzzy logic controller 34 provides a process input signal 50 designated u based on an
k
error signal 40 designated e a change in error signal 44 designated Îe and an auxiliary variable signalk, k
46 having an auxiliary variable value designated AV The error signal and change in error signal 40 and
k.
44, respectively, are derived in a conventional manner subtracting the process output signal 38 from a
set point signal or a system input signal 36.
Controller 34 contains a plurality of sets of fuzzy membership functions or inference
rules which are defined based on prior knowledge or predetermined characteristics of the process 32.
Each set of fuzzy membership functions is defined so that the process 32 is compensated for
undesirable system behavior in a particular region of operation. A particular set of inference rules for
the particular region of operation is selected using the auxiliary variable value. For example, an auxiliary
variable is associated with each of the regions I, II and III of process operations shown in Figure 2a.
Corresponding sets of fuzzy rules are then applied based on the auxiliary variable for providing fuzzy
3
Page: Previous 1 2 3 4 5 6 7 Next
Last modified: November 3, 2007