|
当然不是,你可以看看下面的比较,根据系统的实际情况选择相应的推理算法( M& ~: |/ |- t% ?/ T6 i
Comparison of Sugeno and Mamdani Systems
/ B5 T y0 P. R7 n+ O$ T5 Z" t% `, |+ Y/ @) ^" v* L; M0 L+ H6 H& ]+ J* n
Because it is a more compact and computationally efficient representationthan a Mamdani system, the Sugeno system lends itself to the use ofadaptive techniques for constructing fuzzy models. These adaptivetechniques can be used to customize the membership functions so thatthe fuzzy system best models the data.
6 S0 v2 _/ @: X) R7 P& M: M9 N M( W4 X) l- q0 X
Note You can use the MATLAB® command-line function mam2sug toconvert a Mamdani system into a Sugeno system (not necessarily witha single output) with constant output membership functions. It usesthe centroid associated with all of the output membership functionsof the Mamdani system.
, A' J2 B5 f( a5 K
1 Y4 q" F$ q1 ~* \$ ^/ ^: FThe following are some final considerations about the two differentmethods." Z% V$ b# R9 I: {/ C
" J* x9 G! ~0 X/ K& I4 lAdvantages of the Sugeno Method:
3 u. }6 U' h0 f6 ~! _8 c
& S( C+ Q. V& i' q. [+ vIt is computationally efficient.
v% J: f, a7 O9 y9 LIt works well with linear techniques (e.g., PID control).
4 ~( k- A$ m3 p3 K- ?' PIt works well with optimization and adaptive techniques.
Z u5 [ p6 e, f8 I2 HIt has guaranteed continuity of the output surface.8 p" N- c5 W$ `, J2 h
It is well suited to mathematical analysis.
$ y+ m6 L4 w- x0 ^% J% X6 z/ c( n# k9 ~) l, n& O
Advantages of the Mamdani Method:/ H" r7 T. s& h6 y+ R4 m9 m
& Y% O0 O: x$ @- H6 e) M! lIt is intuitive.9 r! z' j* i- F" d: _9 l Q
It has widespread acceptance.
3 a! U1 F( O3 S0 vIt is well suited to human input. |
|