BÁO CÁO HỌC PHẦN

7/31/2019 BO CO HC PHN

1/31

I HC TH DU MT

KHOA CNG NGH THNG TIN

BO CO HC PHN : TR TU NHN TO

TI: Tm hiu logic m v ng dng

Sinh Vin thc hin : Nhm 4 C10TI01 - Cao Th Bo Trn

T Bo Thng

L Th Huyn Trang

Trn Th Thu Trang

Trn Thanh Trng

Nguyn V Bch Vn

***************************

Mc Lc

I. Tng quan v logic m

1. Gii thiu v suy lun khng chc chn

2. Gii thiu v logic m

II. Mt s khi nim c bn

III. ng dng ca logic m trong thc t

Ni Dung Bi Bo Co

Theo Zadeh, tnh ton mm nhm vo cc h thng c tnh cht: khm phnhng chnh xc c th chp nhn c, khng chc chn hon ton, ch mt

phn chnh xc nhng c th vn dng d dng, n nh, chi ph tnh ton thpv c th ng dng trong thc t. SC truyn thng bao gm bn k thut chnh:


2/31

Cc h thng lp lun theo xc sut (PR)

Cc h thng logic m (FL)

Mng noron (NC)

Tnh ton tin ho (EC-Evolutionary Computation) trong :EC=GP+ES+EP+GA

Trong cng ngh tnh ton mm, thnh phn pht trin vt bc v ng dngtrng ri nht l Logic m

Nm 1964, gio s Zadeh bt u suy ngh liu c th logic tt hn no dngtrong my mc. ng c tng liu ta c th bo my iu ho lm vic nhanh

hn khi tri nng ln, hay nhng vn tng t nh th, s hiu qu hn vict ra tng lut cho tng nhit . y chnh l bc i u tin ca logic mhin i nh chng ta hiu v ng dng ngy nay.

Phi mt mt thi gian di logic m mi c chp nhn, mc d ngay t umt s ngi rt quan tm. Bn cnh cc k s, nhng nh trit hc, tm lhc v x hi hc nhanh chng p dng logic m vo ngnh khoa hc ca mnh.

CHNG 0 : Gii Thiu

Ngy nay chng ta ang ng u vi nhiu ng dng th gii thc phc tpnh nhn dng mu (nh, ging ni hay ch vit tay), robot, d bo, v ra quytnh trong cc hon cnh khng r rng v.v . Cc loi ng dng ny da trnmt khi nim v nn tng mi c tn l Tnh Ton Mm (Soft Computing -

SC) do L.A. Zadeh, cha ca logic m a ra. tng c bn bn trong SCthuc v GS. Zadeh. Tm nhn ca ng v s tng tc gia cc k thut khcnhau ang nh hng mnh m n s pht trin ca th h h thng thngminh (da trn nhn thc) mi theo hng Tnh Ton Thng Minh (CI).

Cng Ngh Thng Minh (Intelligent Technologies) hay Tnh Ton Thng Minh(Computing Intelligent - CI) l kt qa chnh ca s kt hp khng ngng giaH M (Fuzzy Systems - FS) hay Logic M (Fuzzy Logic - FL), Mng Neural(NN) v Tnh Ton Tin Ha (Evolutionary Computation - EC). Cc cng ngh


3/31

ny khng ngng pht trin lm gia tng li ch v thng mi v cng nghiptrn th gii hin nay

Thc t cho thy khi nim m lun tn ti, hin hu trong cc bi ton ng

dng, trong cch suy lun ca con ngi. V d nh tr, rt-tr, hi-gi,... Hnna, B. Russel vit: Tt c logic c in lun gi s rng cc i tngc s dng l r rng. V th n khng th ng dng tt trong cuc sng trntri t ny.... Nh vy, rt cn mt tip cn nghin cu mi so vi logc cin.

Bng cc phng php tip cn khc nhau, cc nh nghin cu nhDubois, Prade , Mamdani , Tagaki, Sugeno, Ishibuchi , Herrera a ranhng kt qu c v l thuyt v ng dng trong cc bi ton iu khin m,

khai ph d liu m, c s d liu m, cc h h tr quyt nh,...

Nhng bi ton nh vy ngy mt nhiu hn trong cc lnh vc iukhin ti u, nhn dng h thng,... Ni chung l trong cc qu trnh quyt nhnhm gii cc bi ton vi cc d liu khng y , hoc khng c nhngha mt cch r rng (trong iu kin thiu thng tin chng hn).

Mt cch tip cn mi mang li nhiu kt qu thc tin v ang tip tcpht trin l cch tip cn ca l thuyt tp m (FUZZY SET THEORY), do

gio s Lotfi Zadeh ca trng i hc California - M ra nm 1965. Cngtrnh ny thc s khai sinh mt ngnh khoa hc mi l l thuyt tp m v nhanh chng c cc nh nghin cu cng ngh mi chp nhn tng. Mts kt qu bc u v hng nghin cu tip theo gp phn to nn nhng sn

phm cng nghip ang c tiu th trn th trng. L thuyt tp m ngycng phong ph v hon chnh, to nn vng chc pht trin logic m. Cth ni logic m (Fuzzy logic) l nn tng xy dng cc h m thc tin, v dtrong cng nghip sn xut xi mng, sn xut in nng, cc h chuyn gia trongy hc gip chun on v iu tr bnh, cc h chuyn gia trong x l ting ni,

nhn dng hnh nh,... Cng c ch cht ca logic m l tin ha v lp lunxp x vi php suy din m.

Trong phn ny, mc ch chnh l gii thiu khi nim tp m, logic m,tp trung i vo cc php ton c bn v bc u i vo lp lun xp x vi

php suy din m.

I.Tng quan v logic m


4/31

1. Gii thiu v suy lun khng chc chn .

Cc th tc suy din tun theo m hnh suy lun s dng trongphp tnh v t : t cc tin ng n , cc lut suy din vng

chc sinh ra nhng kt lun mi , m bo l ng n . Tuy nhin, trongthc t , c rt nhiu tnh hung chng ta phi rt ra nhng kt lun ttt nhng bng chng c xc nh ngho nn v khng chc chnthng qua vic s dng nhng suy din khng vng chc , v d : pht

biu cc chun on y hc ng n v xut cch iu tr t nhngtriu trng khng r rng ; hoc phn tch nhng trc trc ca t thngqua nhng biu hin ca n .

Nh vy s c hai loi thng tin khng chc chn : mt l d liu ban ukhng r rng, khng ng tin cy , khng hai l cc lut suy dinkhng chc chn , suy lun khng hp logic , suy lun ngc t kt quv iu kin , gi chung l suy lun theo kiu phng on . Suy lun

phng on khc vi suy lun vng chc .

V d : Ta p dng lut modus ponens trong suy lun sau

Nu c quy v dy cp b trc trc , th ng c khi ng c v nkhng sng

y l suy lun lun lun ng hay l mt suy lun vng chc , nhng nli khng gip ch cho qu trnh chun on trc trc ca xe . Tuy nhin ,nu ta o ngc li :

Nu ng c khng khi ng c v n khng sng th c quy hocdy cp b trc trc .

Suy lun ny li rt hu dng cho vic chun on trc trc ca xe . yl phng on t triu trng quan st c suy ngc li nguyn nhnca chng , nn gi l suy lun khng vng chc .

2. Gii thiu v Logic m ( Fuzzy logic )

L thuyt tp m (FUZZY SET THEORY) l mt phng php lunlinh hot n l mt cch tip cn x l suy din phng on v skhng chc chn , n mm do trong mi trng thng tin phc tp , dliu khng chc chn , thiu chnh xc , v bin ng kt hp vi ccchuyn gia a ra cc kt qu chnh xc v nhanh nht .


5/31

Cc chuyn gia s dng cc lp lun mt cch t nhin gii quyt ccbi ton .Cc tri thc ny thng l cc tri thc khng r rng v rtkh din t bng cc h thng logic truyn thng .

Suy lun theo logic m quan tm n mc tht ca mt khng nhHin nay , logic m ang c ng dng rng ri trong nhiu lnh vc,c bit l cc h thng iu khin m .

Logic m ang tr thnh mt trong nhng cng ngh thit k v phttrin h thng iu khin phc tp thnh cng nht hin nay . Chng tathng nghe ni nhiu n thut ng nh my git fuzzy , qut fuzzy , xemy fuzzy

II.Mt s khi nim c bn

Cc bin khng c xc nh mt cch chnh xc.

Cc bin c th c gi tr thc bt k trong khong 0 1 .

1. Tp r v hm thnh vin

Tp r l tp hp truyn thng theo quan im ca Cantor (crispset) . Gi A l mt tp hp r, mt phn t x c th c x A hocx A, C th s dng hm m t khi nim thuc v.

Nu x A, (x) = 1, nguc li nu x A, (x) = 0. Hm c gi lhm c trng ca tp hp A.

2. Tp m v hm thnh vi n


6/31

Cho S l mt tp hp v x l mt phn t ca tp hp . Mt tp conm F ca S c nh ngha bi mt hm t cch thnh vinF(x) o mc m theo x thuc v tp F.

Trong , 0 F(x) 1.- Khi F(x) = 0 => x F hon ton.

- Khi F(x) = 1 => x F hon ton.

Nu x, F(x) = 0 hoc 1th F c xem l gin

Hm thnh vin F(x) thng c biu din di dng th.

V d 1 : S l tp cc s nguyn dng v F l tp con m ca S vc gi l s nguyn nh

1 2 3S nguyn

V d 2: Mt s biu din tp m cho cc tp ngi n ngthp , trung bnh v cao .

1 short medium tall

Chiu cao 4 466 5 56 6 66

V d 3: Cho tp m Young

Lan 16 tui, (Lan)=1

Hng 25 tui, (Hng)=0.5

1


7/31

3 . Cc dng ca hm thnh vin

Cc hm thnh vin ca tp m c 3 dng c bn : dng tng , dng gim , vdng chung .


8/31

A, Dng S tng :

(x)=S(x, , , ) =

0 nu x


9/31

C, Dng hnh chung :

(x; , )=

S(x; - , - /2; ) if x

Hm dng chung

4 . Cc php ton trn tp m :

Cc php ton trn tp m c nh ngha thng qua cc hm thuc v cxy dng tng t nh cc php ton trong tp m kinh in , bao gm cc tpcon , php giao ,php hp v b .

Cho ba tp m A, B , C vi A(x), B(x), C(x) .

a, Php giao :

C=A B: C(x) = min( A(x), B(x))


10/31

b) Php hi

C=A B : C(x) = max( A(x), B(x))

c) Php b

C=A : C(x) = 1- A(x)


11/31

Nhn xt:

Logic m khng tun theo cc lut v tnh b ca logic truyn thng:

AA(x) 1 v AA(x) 05. Lut m

Mt lut m l mt biu thc if then c pht biu biu din dng ngnng t nhin th hin s ph thuc nhn qu gia cc bin .

C dng :

o IF x is A THEN y is B

o IF x is A AND y is B THEN z is C

o IF x is A OR y is B THEN z is C

Vi x,y,z : bin ngn ng (linguistic variable)

A,B,C : tp m (fuzzy set)

Lut m:

o Gip truyn t, m t rt t nhin nhng quy lut trong cuc sng.

o Th hin c nhng din t v chuyn mn hn

o Hiu lc i vi phm vi cc bin rng ln hn

o Mt Lut m c th thay th nhiu, thng l rt nhiu, nhng lut r

V d : Ifnhit l lnh v gi du l r Then si m nhiu .

Trong - 'nhit ' , 'gi du ' v 'si m' l cc bin .

- 'lnh', 'r' ,' nhiu' l cc gi tr hay chnh l tp m .

Hoc : Ifmt ngi c chiu cao l cao v c bp l lc lng Then chi bngr hay .

Cc bin y s l :' chiu cao','c bp', 'chi bng r'. Cc gi tr hay tp m l :'cao', 'lc lng', 'hay'.


12/31

6. C s tri thc m

L tp hp cc lut m lin quan n lnh vc no .

V d: c s tri thc m p dng cho Washing Machine

o if x is Large and y is Greasy then z is VeryLong;

o if x is Medium and y is Greasy then z is Long;

o if x is Small and y is Greasy then z is Long;

7.Logic m

L m rng ca thuyt tp m qua vic dng cc ton t logic AND, OR,NOT, ...

o Nhng pht biu l ngh, khng nh hoc lut.

o ngh v khng nh c gi tr m lin kt vi chng.

o Logic m p dng cc lut to ra gi tr mi hoc mc ng tng ng.

- nh gi s tht. Mc ng gia ng v sai. Khng phi mi th u l

ng/sai, bt/tt, trng/en.

- nh gi thnh vin: tp nhng ngi cao, tp nhng thnh ph xa xi, tpnhng vt t tin.

- Logic s dng nhng thut ng thuc v ngn t. Din t tri thc chuyn giamt cch t nhin.

8. Nguyn l x l cc bi ton m


13/31

Quy trnhxy dngh thng

suy lun m

Gi tr E c th c a vo h thng iu khin m thng qua b phn nhp .D liu vo s c chuyn thnh cc gi tr m . Qu trnh ny c gi l mha (fuzzification) .H thng iu khin s thi hnh qu trnh lp lun m(fuzzy processing) , ni b x l s so snh d liu u vo vi c s d liucha gi tr u ra .Qu trnh lp lun m lin quan n s thc hin cc lut cdng

IF THEN c nh ngha trong qu trnh thit k .Sau khi b iu khinm hon thnh lp lun m v t n kt qu u ra n chuyn sang giai on

gii m cho ra kt qu u ra U dng gi tr r .Cc h thng suy lun m (Fuzzy Inference System) thc hin vic suy lun to ra cc quyt nh t cc thng tin m h, khng y , thiu chnh xc.Qu trnh suy lun m bao gm 4 bc sau:

M ha :xc nh cc tp m c s v hm thuc ca chng . To lut :Xc nh cc quy tc hp thnh t bn cht ca ngdng v s dng kt hp cc tp m c s . Kt nhp : kt hp cc quy tc hp thnh . Gii m: gii m cho cc tp m kt qu .

A , Xc nh tp m c s v hm thuc .

i vi mt s ng dng n gin, cc tp m c s v hm thuc c th xcnh c d dng khng cn tham kho kin chuyn gia hoc kin cachuyn gia ch to ra cc gi tr khi to ban u. Phng php ny cn s dngcc k thut tnh ton mm hin i (v d nh cc gii thut di truyn hocmng nron).. i vi cc ng dng phc tp, xc nh cc tp m c s,

cc hm thuc lin quan thng da vo kinh nghim ca cc chuyn gia v ccquyt nh ch quan ca h.


14/31

B, To cc quy tc hp thnh .

Mt h thng m bao gm nhiu quy tc hp thnh. Quy tc hp thnh c tothnh t mi quan h ca cc thnh phn ca ng dng. Qu trnh to cc quy

tc hp thnh c th c thc hin bng mt chuyn gia hoc bng phngphp t ng dng k thut tnh ton m. Mi quy tc hp thnh c u vo lmt s tp m c bn v to ra kt qu mt tp m u ra.

C, Kt nhp cc quy tc hp thnh

Qu trnh ny tng hp kt qu ca cc quy tc hp thnh ring bit vo mt ktqu duy nht. u vo ca khu kt nhp l cc tp m u ra ca cc quy tchp thnh. u ra ca n l mt tp m cho mi bin u ra. Qu trnh kt nhpc thc hin nh sau: Vi mi i tng u tin trong u vo ca lut hp

thnh, tm gi tr nh nht ca hm thuc ti im xc nh bi d liu u vo.Tip tc thc hin vi cc i tng tip theo trong lut hp thnh. T ttc cc lut hp thnh, to mt tp m kt qu bng php ton max cacc gi tr thuc c c.

D, Gii m .

Sau qu trnh kt nhp cc quy tc hp thnh, chng ta thu c kt qu u ral mt tp m. Qu trnh gii m s xc nh r mt gi tr i din t hm

thuc ca gi tr m u ra. Gi tr c xc nh s l u ra ca ton b hthng. C hai phng php gii m chnh l phng php im cc i vphng php im trng tm. Vic la chn phng php gii m tu thuc votng ng dng c th. Vi cc ng dng phc tp th phng php im trngtm c s dng nhiu nht.

Phng php trng tm

y c(

x ) =

=

=

N

i i

N

i ii

y

yy

B

B1

1

)('

)(

a) Phng php cao : y h (

x) =

=

=

M

i

j

M

i

jj

y

yy

B

B

j

j

1

1

)('

)(


15/31

Vi j l ch s lut , y-j l im c lin thuc ln nht trong tp m

u ra B j , th j v Bj ( y-j ) c tnh theo cng thc A(x ) = Tn

(Aj

1 (x1) , .., Aj

n (x n) ) , An (x n) ) nh sau :

)('y

j

Bj

= )(yj

Bj

* Aj

1 (x1) * Aj

n (x n) ) * An (x n)

Sau khi bin i , ta c :

y m h (

x) =

=

=

M

i

jj

M

i

jjj

y

yy

B

B

j

j

1

2

1

2

/)('

/)(

vi j

l h s bin i ca lut j

Phng php tm ca cc tp (Center of - Sets ) :

Phng php ny mi lut c thay th bi tp singleton tm c j

y c o s (

x ) =

=

=

=

=M

i i

M

i i

j

xT

xTcni A

ni A

j

i

j

i

1

1

)(1

)(1

9 . Li ch ca vic s dng logic m .

- L mt la chn cho phng php thit k n gin hn v nhanh hn.

- Logic m n gin ha s phc tp trong thit k.

- Chp nhn c d liu khng chnh xc, x l cc bi ton phc tp rt tt.

- Cch tin li din t tri thc nhn thc bnh thng v tri thc chuyn gia.

- ng dng trong rt nhiu lnh vc nh:

+ X l nh, tn hiu,thit k v tng hp phn cng, tr tu nhn to v hchuyn gia/h h tr ra quyt nh, cc h thng iu khin

10 . Gii hn ca logic m .


16/31

Mt hn ch hin nhin ca lgic m l n khng phi lun lun chnh xc.Nhng kt qu c lnh hi nh mt phng on, v vy n c th hin khngrng ri c tin cy nh mt cu tr li t lgic c in.

Do vy n c nhng mt hn ch nh sau:- Khng phi l gii php cho mi trng hp.

- Nhng m hnh chnh xc / r c th hiu qu v tin li hn trong 1 s trnghp.

- Nhng tip cn khc c th thm nh chun hn.

11, Kt lun

Logic m t khi c nghin cu v pht trin n nay ngy cng c ngdng nhiu.

III.ng dng ca logic m trong thc t

ng dng my git

A , Gii thiu

Khi bn s dng my git, ni chung bn la chn thi gian git da vo slng qun o , tuy nhin , thc t th my git cn bit thm cc thng s khc nng cao hiu sut git , nh bn , loi cht bn , c x phng , lngnc . y , n gin , ta s quan tm ti hai i lng u l bn vloi bn . Khng may l khng c phng php ton hc no c th tnh chnhxc c mi quan h gia th tch qun o , bn vi thi gian git cn thit, bi vy , vn ny vn cn cha gii quyt c cho ti tn by gi . Mingi n gin ch da vo cm gic v kinh nghim ra thi gian git

t ng ha qu trnh xc nh thi gian git , logic m c th c s dng gii quyt vn .

Trong cc h thng my git ,vic trn c h thng gm hai phn m nhim

1. B cm bin2. B iu khin

B cm bin s ng vai tr l u vo cn b iu khin sa ra quyt nh thi gian git.

B iu khin : C hai tr nhp vo , c th cc gi tr ny c sensorgi ti


17/31

( 1) bn trn qun o

( 2) Loi cht bn trn qun o.

bn c xc nh bi s trong sut ca nc. Mt khc, loi cht bn cxc nh t s bo ha, thi gian n dng t n s bo ha.

Qun o du m chng hn cn lu hn cho s trong sut nc t n sbo ha bi v m l cht t ha tan trong nc hn nhng dng khc ca chtbn. Nh vy mt h thng phn t sensors (cm bin ) kh tt c th cung cpnhng input cn thit c nhp vo cho b iu khin m ca chng ta.

Nhng gi tr cho bn v loi cht bn l c chun ha ( phm vi t 0ti 100) c cho bi gi tr phn t sensors.

Vic tip theo l phi xc nh tp m :

- Vi bin ngn ng bn c cc tp m

+ Bn t (Dirty.Small)

+ Bn va (Dirty.Medium)

+ Bn nhiu (Dirty.Large)

- Vi bin ngn ng loi cht bn c cc tp m

+ M t (Greasy.NotGreasy)

+ M va (Greasy.Medium)

+ M nhiu (Greasy.Greasy)

- Vi bin ngn ng kt lun xc nh thi gian git c cc tp m

+ Git rt ngn (Time.VeryShort)

+ Git ngn (Time.Short)

+ Git va (Time.Medium)

+ Git lu (Time.Long)

+ Git rt lu (Time.Very Long)


18/31

B , Tp lut:

Mt quy tc trc gic tiu biu nh sau :

Nu bn t v khng dnh m th thi gian git nhanh

T nhng s kt hp khc nhau ca nhng lut v nhng iu kin khc,chng ta vit nhng quy tc cn thit xy dng b iu khin my git.

Gi :

x ch bn

y ch Loi cht bn

z Thi gian git

Dirty.Small Dirty.Medium Dirty.Large

Greasy.NotGreasy Time.VeryShort Time.Short Time.Medium

Greasy.Medium Time. Medium Time. Medium Time. Long

Greasy.Greasy Time. Long Time. Long Time.VeryLong

T ma trn quy tc trn , ta suy ra tp lut

if x is Dirty.Large and y is Greasy.Greasy then z is Time.VeryLong;

if x is Dirty.Medium and y is Greasy.Greasy then z is Time.Long;

if x is Dirty.Small and y is Greasy.Greasy then z is Time.Long;


19/31

if x is Dirty.Large and y is Greasy.Medium then z is Time.Long;

if x is Dirty.Medium and y is Greasy.Medium then z is Time.Medium;

if x is Dirty.Small and y is Greasy.Medium then z is Time.Medium;

if x is Dirty.Large and y is Greasy.NotGreasy then z is Time.Medium;

if x is Dirty.Medium and y is Greasy.NotGreasy then z is Time.Short;

if x is Dirty.Small and y is Greasy.NotGreasy then z is Time.VeryShortC , Hm thnh vin

Hm thnh vin ca bn:

Hm thnh vin ca Loi cht bn:


20/31

Cc hm thnh vin (membership function) , ca kt hp cho tng lut :


21/31


22/31


23/31


24/31

Ta c v d sau

BN : 30 CHT BN : 60

Ta c :

Dirty.Small(x0) = 0.4

Dirty.Medium(x0) = 0.6

Dirty.Large(x0) = 0

GreasyNotGreasy(y0) = 0

Greasy.Medium(y0) = 0.8

Greasy.Greasy(y0) = 0.2

W1 = min(Dirty.Large(x0), Greasy.Greasy(y0)) = min(0,0.2) = 0

W2 = min(Dirty.Medium(x0), Greasy.Greasy(y0)) = min(0.6, 0.2) = 0.2

W3 = min(Dirty.Small(x0), Greasy.Greasy(y0)) = min(0.4, 0.2) = 0.2


25/31

W4 = min(Dirty.Large(x0), Greasy.Medium(y0)) = min(0, 0.8) = 0

W5 = min(Dirty.Medium(x0), Greasy.Medium(y0)) = min(0.6, 0.8) = 0.6

W6 = min(Dirty.Small(x0), Greasy.Medium(y0)) = min(0.4, 0.8) = 0.4

W7 = min(Dirty.Large(x0), Greasy.NotGreasy(y0)) = min(0, 0) = 0

W8 = min(Dirty.Medium(x0), Greasy.NotGreasy(y0)) = min(0.6, 0) = 0

W9 = min(Dirty.Small(x0), GreasyNotGreasy(y0)) = min(0.4, 0) = 0

D , Bc cui : kt hp.C(z) = W2 x Time.Long(z) + W3 x Time.Long(z) + W5x Time.Medium(z)+ W6 x Time.Medium(z)

C(z) =

Ta tnh trng tm ca hm sau khi kt hp :

V moment tnh ca hm kt hp :


26/31

Vy , sau khi gii m ha , ta c thi gian git l :

CHNG 3 : M PHNG FUZZY LOGIC BNG MATLAB

3.1 Tng Quan v MatLab 7.0

3.1.1. Gii thiu Matlab 7.0

H thng tnh ton khoa hc k thut

Ngn ng lp trnh cp cao

Th vin hm phong ph (Tool Box)

M phng, v th, biu

Phn tch d liu

Pht trin phn mm k thut

Phin bn mi nht: Matlab 2007.

3.1.2 Cc ToolBox trong Matlab 7.0

Toolbox l cc th vin hm sn c h tr cho cc lnh vc tnh ton c th.

Cc toolbox thng dng

MatlabFuzzy Logic

Image Processing

Neural Network

Signal Processing

Simulink

Symbolic Math


27/31

y chng ti ch nghin cu toolbox Fuzzy logic m thi.

3.2 Bi ton Water Tank With Ruler Viewer (iu khin mc nc bm vothng nc)

Bi ton: C 1 thng cha nc.

Cn bm nc vo thng t ng bng my bm.

Ty vo mc nc trong thng bm.

3.2.1 Bi ton:

Hot ng valve (bm) da trn 5 Rules (lut) sau:

Rule 1: If (level is Okay) then (valve is no_change)

Rule 2: If (level is low) then (valve is open fast)

Rule 3: If (level is hight) then (valve is close fast)

Rule 4: If (level is Okay) and (rate is positive) then (valve is close_slow)

Rule 5: If (level is Okay)and (rate is negative) then(valve is open_slow)3.2.2 M Ho:

Bin ngn ng ng vo ca Valve: Rate, Level

Rate (tc valve) : negative, none, positive

Level (mc nc) : hight, okey, low

Bin ngn ng ng ra ca Valve : Valve


28/31

Valve (trng thi bm):close_fast, close_slow, no_change, open slow, open_fast.

3.2.3 S nguyn l bi ton m phng trn Matlab:

Trong : error: lch gia mc o c thc t v mc mong mun.

Fuzzy Controller with Ruleviewer : b iu khin trung tm dng kthut x l bng fuzzy logic.

PID: P = propootional (khu t l), I = intergral (khu tch phn),D= Differenrial (khu vi phn). Cng thc tnh ton:


29/31

Kp, KI, KD : h s t l ca 3 khu P, I, D

e(x): u vo

Switch: b cng tc chn x l theo hoc l : PID hoc l Fuzzy Logic hocl const ( -1).

du/dt : tc thay i ca valve

change: dng hn ch bin valve trnh valve quay qu nhanhhoc qu chm valve s chy.

3.2.4 Bn Demo s dng Matlab:

Trong vi 2 u vo : level v rate ta s c 1 u ra valve.

Mi 1 dng (1 n 5) ca level tng ng 1 lut.

ng ng vi dng th 6 ca valve l kt qu ra ca valve cn iu khin.


30/31


31/31

Documents

BÁO CÁO HỌC PHẦN