Transformation method of decision-making probability based on correlation degree

11

2015 11 Systems Engineering — Theory & Practice : 1000-6788(2015)11-2932-07

: TP391 ! " # $ % & ')()* + 35

Vol.35, No.11 Nov., 2015 : A

,.-0/ , 13204 , 5 6 , 73839 H IKJMLMNMOMPMQMRMSMTM(UM: VM; < WM= XM> PM?AQZ@CYMBE[]D \ FA,@ ^ZG _M, : `]; aM< b 150001) Y]NMcMdMeMfZY]WMXMPMQMTMUMgZh i gZh]jMkMlMmMnM\ZY]NMOMPMQMRMSMoMTMUMpZY]WMXMPMQMqZrMY]dMeMfMsMVMtMuMTMUMgZhMY]vMw xMyMzMLZ{]|M}MfMTMUMgZh]oZ{]|M~M MfMTMUMgZh]Â&#x20AC;MÂ MÂ&#x201A;MÂ&#x192;Â&#x2026;Â&#x201E;MÂ&#x2020; , Â&#x2021;Z_ nM\ZY]WMXMPMQ . Â&#x2C6;MÂ&#x2030;MÂ&#x160; | LZ{ b gZh]Â&#x2039;Â?Â&#x152;Â?Â&#x17D;MgZh , Â?MÂ&#x2018;MÂ&#x2019;ZÂ&#x201C;]b gZh]Â&#x201D;MÂ&#x2022;MÂ&#x2020;MÂ&#x2013;MÂ&#x17D;MÂ&#x2014; . Â&#x2DC;MÂ&#x2122;MÂ&#x161;Â&#x203A;NMOMPMQMRMS ; dMeMf ; }Mf ; ~M Mf ; WMXMPMQ

. ,

Transformation method of decision-making probability based on correlation degree ZHAO Yu-xin, JIA Ren-feng, LIU Chang, SHEN Zhi-feng (College of Automation, Harbin Engineering University, Harbin 150001, China)

Abstract To slove the problem in the transformation of basic probability assignment to decision-making probability, this paper proposed a novel transformation method based on correlation degree. The correlation degree between basic probability assignment of singleton proposition and decision-making probability was used to evaluate the transformation method, and the decision-making probability of each proposition was achieved by linear combination, which was the transformation method of decision-making probability based on proportional belief and proportional plausibility. The proposed method was compared to the other usual methods with an example. The experimental result shows that the proposed method is more reasonable and effective.

Keywords basic probability assignment; correlation degree; belief; plausibility; decision-making probability

Â&#x153;Â&#x17E;Â? Â&#x;Â&#x2026; Â&#x2026;ÂĄÂ&#x2026;¢Â&#x2026;ÂŁ Â&#x2026; ¼¤ , ÂŚÂ&#x2026;§Â&#x2026;¨Â&#x2026;Š Â&#x;Â&#x2026;ÂŞÂ&#x2026;ÂŤÂ&#x2026;ÂŹÂ&#x2026;­Â&#x2026;ÂŽÂ&#x2026;ÂŻÂ&#x2026;°Â&#x2026;ÂąÂ&#x2026;²Â&#x2026;ÂłÂ&#x2026;´Â&#x2026;¢Â&#x2026;ÂŁ , ÂľÂ&#x2026;ÂśÂ&#x2026;œ¼¡š¸ ÂŹÂ&#x2026;ÂşÂ&#x2026;ÂťÂ&#x2026;ÂźÂ&#x2026;½Â&#x2026;ž , Â&#x2026; ÂŹÂ&#x2026;ÂżÂ&#x2026;Ă&#x20AC; Ă MĂ&#x201A; ÂŻM¢MÂŁ . Ă&#x192; D-S Ă&#x201E;MĂ&#x2026;M M ÂżMĂ&#x2020;M½MĂ&#x2021;MĂ&#x2C6;MĂ&#x2030;M¢MĂ&#x160;MĂ&#x2039;MĂ&#x152;MĂ? , Ă&#x17D; ÂżMĂ&#x2020;MĂ?MĂ?MĂ&#x2C6;MĂ&#x2018; ÂŹMĂ&#x2019;MÂşĂ&#x201D;Ă&#x201C;Ă&#x2022;ÂŹMÂşMÂťĂ&#x201D;Ă&#x201C;Ă&#x2022;ÂŹMÂŽMĂ&#x2013; ÂŞMÂŤ , Ă&#x2014; ÂżMĂ&#x20AC; Ă MĂ&#x2DC; ¸ ÂŹM­MĂ&#x2122;MÂŞMÂŤMĂ&#x161;MĂ&#x203A;MÂŹMÂşMÂťMĂ&#x153;MĂ? ÂŻMĂ&#x17E;Â&#x2026;Ă&#x; , Ă MĂ&#x192;MĂĄMâMĂŁMäMĂĽ Â&#x;MÂŞMÂŤMĂŚ Ă&#x201A; Ă&#x201C;èçMĂŠMĂŞMĂŤ Ă&#x201C;èÏ Ă­MĂŽMĂŻ Ă&#x201C;èðMĂąMòMĂłMĂ´ ĂŽ ĂľMĂś ÂĄMáMø , ĂšÂ?ĂşZĂť]ĂźMĂ˝MĂžMĂż Ă&#x2018; ÂŹMÂşMÂťMĂ&#x152;MĂ?MÂŻ ¨M . D-S Ă&#x201E;MĂ&#x2026;M M Z¤ ÂŻ Ă&#x2DC; (basic probability assignment, BPA), ÂŹ

MĂ?MÂŞMÂŤMĂ&#x161;MĂ&#x203A; , MÂĄ MĂ?MÂŞMÂŤMĂ&#x161;MĂ&#x203A; . BPA ÂżMĂ&#x2020;M½ Ă&#x201A; Ă&#x2C6; MĂ?MÂŞMÂŤ , Âľ Â&#x; Ă&#x2DC; ¸ BPA Ă&#x161;MĂ&#x203A;M¢MÂŁ , M¨ Ă&#x2030;M¢MÂŻMĂ&#x152;MĂ? ÂĄ MĂ? MÂŻ BPA MĂ? , ! ÂşMÂŻM¢MÂŁ " . Ă&#x192; ¢MÂŁ # $ MĂž Ă&#x2039;Â&#x2026; % ĂłMÂŻMĂ&#x17E;MĂ&#x; , Ă & , ¢MÂŁ # $ ĂźMĂ˝ ' ( ) * + , ÂŻ - . . Â&#x; # / ÂŞMÂź 0 1 (transĂšÂ?Ăş , BPA Ă&#x161;MĂ&#x203A;M¢MÂŁ Ă&#x20AC;M# ´M$ ÂŻ ÂŻMĂ&#x17E;MĂ&#x; %M M ÂĄ # $ , §MĂĽ ÂŻMĂ&#x17E;MĂ&#x17E;MĂ&#x; Ă&#x; % Smets ÂĄ MĂ?" Â&#x2026;ÂŻ BPA 3

ferable belief model, TBM) ¤ Pignistic , ¾ 2

M� , 4 5 6Mß ªM 7 8 ,  9 ¸ ¢M£ . Cuzzolin ¥ M� M¯ BPA : ; MºM=< 3 >

MĂ? , Âľ ? ÂĄ MĂ? @ MĂ? A % B MÂŻ C , D 6MĂź ÂŹ Ă&#x201A; . Cobb Ă&#x20AC;M´MÂŻ Ă&#x2DC; ¸ E F G ÂŻ HJILKNM OJPJQLR SJTLUNVJWJXJYJZ J[J\J]JYJ^JWJ_J`JaJbdcdeJfdZ gJhJiJjJkJXJlJSJmJnJWdX YJoJZpJqJr J[J\J]JYJ^JWJ_J`JaJb sJtJu v w gJhJiJxJyJzJm {J| }J~J JÂ&#x20AC;J _JÂ&#x201A;JÂ&#x192;LÂ&#x201E; Â&#x2026;JÂ&#x2020;JWJX Â&#x2021;JÂ&#x2C6;JÂ&#x2030;JÂ&#x160; Â&#x2039;Â?Â&#x152;Â?Â&#x17D; v w gÂ?hÂ?iÂ?Â?Â?Â?Â?Â&#x2018;Â?m }Â?~Â?_Â?Â&#x201A;Â? _Â?Â&#x201A;Â?Â&#x192;Â&#x2019;Â&#x201E; Â&#x201C;Â?Â&#x201D;Â?Â&#x2022;Â?Â&#x2013; Â&#x2014;Â?Â&#x2DC; Â?Â&#x2122;

0

[1â&#x2C6;&#x2019;2]

[3]

[5]

[4]

[6]

[7â&#x2C6;&#x2019;16]

[7]

[8]

[9]

: 2014-02-28 : (51379049); (LC2013C21); : (1980â&#x20AC;&#x201C;), , , hrbeu.edu.cn; (1987â&#x20AC;&#x201C;), , ,

(HEUCFX41302);

(HEUCF041410) , , ,

,

,

:

:

,

,

, E-mail: zhaoyuxin@ , E-mail: zhuxijia@ 126.com.

sJtJu , : JY J J J J J J J J¡J¢J£J J¤ 2933 ¥ $ (plausibility function transform, PFT), ¦ § ¨ © E F G # $ ª ¯ « ¬ , à A % § ¨ ªM¼ F G ¯ « ¬ , ¯ ­ ®=<]½ ¯ ° . Dezert ÀM´M¯ DSmP ÞMß=<]½MÏMÐ , äMå ±=²]â , µ ³ G ¯ ´ µM¬MÆ ¶ · . ÀM´ © <º¹MªM¼ # $ ÞMß @=<º¹ E ¼ #"$ ÞMß , <º¹MªM¼ # $ ÞMß # $ M¯ »M¼=<]½ ¼ · , Daniel

¸ 4 5 ½ ¾¾M ¢M£M%Mì íÞM;ßM<º¯ ¹ Ø E ¿ ¼ # $ ÞMß # $ ¯= M<º¯ ¹ »M# ¼=$ <]ÞM½ ßM¯ ¯M° ¬ ,Á ¬ 9 ÀM¸ ´ ¢M£ . Ø ©MþMî ÿ Ç"# ¸ $ + ÞMÂ ß ¼M¯ ¯MÈ ¢M£ # $ À Daniel ,

#

$ Ã

Ä Å ÞMß . 2 ÞMß û MÍM¯ ,BPA M ¯ M ¢ £ ¯ " ¼ Æ <ʹ ª ¼ #"$ Þ ß"@Ë<ʹ E" ¼ #"$ Þ ß Ú Û"Ì Ü"Í Â , ´" +"Í Â ¯ ¢ £ ý " . Î Ã , Ï ª"Ð ¹"! Ñ , ×MÄ"å © É "¾ ÞMßM¯ Â Ü .

1 Daniel ÒÔÓÔÕÔÖÔ×ÙØÙÚÙÛÝÜ 1.1 Daniel Þ ß à á â ã ä <º¹MªM¼ # $ ÞMßM¯ # $ å æ : 11

[15]

[16]

P ropBelP (A) =

m(A) P m(X) m(B)

X

(1)

MÞ ß 9 å MÍM M¡ Í ¤ ¯ ªM¼Ë<ºç #"$ ¢ £ , # $ M¯ »M¼=<]½ ¼ · , 4 5 ½ ¾M¢ 2 M£ ì í . <º¹ E ¼ # $ ÞMßM¯ # $ å æ : A∈X⊆2Θ B∈X

P ropP lP (A) =

P l(A) P m(X) P l(B)

X

(2)

MÞ ß 9 å MÍM M¡ Í ¤ ¯ E ¼=<Êç # $ ¢M£ , # $ M¯ »M¼ è=<]½ ¯ ° , ¬ 9 ¸ 2 ¢M£ . é ê ë 1.2 Æ ¬MºM»MÜM¼Mñ —— î ¯ ï ð ñ , ¾ ò [17] ó ô © õ ö ¥ ñM ¯ ÷ î @ + Â î , ø É ÜMð × » ù © + ý ÂMì í G . ú û 1 ü õ ö ¥ ñ X Y ¯ ýMý A∈X⊆2Θ B∈X

þ ÿ õ ö ¥ ñ

X=

X

(

a1 , a2 , · · · , ak p1 , p2 , · · · , pk

+M¸ õ ö ¥ ñ

Y

)

,Y =

¯ ÷ î » ù ý

(

ú û

)

,

:

HY (X) = −

ú û

b1 , b2 , · · · , bk q1 , q2 , · · · , qk

H(X) = −

n X

qk log pk

(3)

pk log pk

(4)

k=1 n X

õ ö ¥ ñ X Y Ä Å ¯ + Â î » ù ýMÎ ( ¯ ÷ î Ä @ , H(X; Y ) = H (X) + H (Y ) ¥ M ñ

¯ ÷ õ ö + ÂM GM + ÂM G ¸ » ù ý : k=1

2

Y

3

X

H(Y ) HY (X) H(X) γX (Y ) = HX (Y )

γY (X) =

% M²MÜMð + M Â

G

γ(X; Y ) =

H(X ⊗ Y ) H(X) + H(Y ) = H(X; Y ) HY (X) + HX (Y )

(5) (6) (7) (8)

:

1 ∀X, Y

? ?

X, Y ý ý

0 ≤ γX (Y ), γY (X), γ(X; Y ) ≤ 1

(9)

γX (Y ) = γY (X) = γ(X; Y ) = 1

(10)

,

Ù Ù

& ÜMð % ¥ + ñ ÂM G γ¯ õ C ö ¥ ñ XÐ Y ¯´ ý Ä Mþ ÜÔÓ þ ¯M °MÜM± ¯ M¼M¾ ñ ò (0) = −∞ , ý & ¯ ïµ , õ, ? ö ¥ ñ Xý +M0¸ õ ö ¥ , ñ Y ¯ ÷ D î » ù log ý : 2934

HY (X) = H(X) =

n X

35

.

[18]

ÀM´ ©MþMÿ î

qk e−5pk

(11)

pk e−5pk

(12)

k=1 n X

¯ å æ ¿MÆM½MÇMÈMÉM¢ M¾M¯MÌMÍ , ¯ Ä © Ð ¯ ! ºMÜ .

2 ÔØÔÚ ! " #ÙÒÙÓÝÕÙÖ ÇMÈ % &=<º¹MªM¼ # $ ÞMß"@=<º¹ ¼ # $ ÞMß ¯ ' ( ¯M¬ Á ¾ 9 ¯ MÜý ,© ÀM$ ´ ©MþMÿ Ø ¸ +  ¼M¯M¢M£ # $ E ÞM ß . 2 ÞMß <º¹MªM,¼ ) # *M$ Î ÞM( ß @=<º¹ , E ¼ # å $+ ÞM ßM+ ÚG Û ÌMÜ Í Â , , # $ M¯ »M¼ -M¬ ¼ · ¬ ¯ ° , M´  . / 0 1 2 3 .

4 5 6 7 8 9 : 0 / ; < = > ? @BA C /BD E , A C / F G 0 E H I J K L / M N , O I J P Q > R S TVU W N I J 6 7 / P Q X Y Z , [ > R S TVU \ ] N I J 6 7 / P Q X Y Z . ^V_ X ` a b < > c d ` a b < e f / Y g h i , 0 1 2 3 I J > j k g h i l m n o p . q Q d ` a b < / BPA r F , s t c u / 0 1 2 3 v u w x F , y z { z | } ~ / k g V , D E ? @ A C . j d ` a b < / BPA I J / 0 1 2 3 / , } _ N o I J / 0 1 2 3 BPA / N v R I J 6 7 /B . q Q TVU W N I J 6 7 /B BN r F , w 6 7 / I J P

Q r , BPA / W , w I J 6 7 r , I J K / M N = u w Y Z ; q Q TVU \ ] N I J 6 7 / N r F , w 6 7 r TVU W N I J 6 7 , I J K / M N = u w Y Z . ¡ y ¢ , 4 5 6 7 / I J £ ¤ q ¥ : k=1

¦ §

m(A) P m(X) + α2 m(B)

X

N ewP (A) = α1

A∈X⊆2Θ B∈X

,

P l(A) P m(X) P l(B)

(14)

γP l = 1 − α1 γBel + γP l

(15)

γ γ ¨ © ª TVU W N I J 6 7 T«U \B] N IBJ 6B7¬§VdB` aBbB< / 3

/

£ ¤ (13) v { . z ­ ® Bel

¦ § ¯ ª

Pl

N ewP (A) = α

X

m(A) P m(X) + (1 − α) m(B)

A∈X⊆2Θ B∈X

,

α=

A

(13)

A∈X⊆2Θ B∈X

γBel γBel + γP l

α1 = α2 =

X

X

BPA

I J / 0 1 2

P l(A) P m(X) P l(B)

(16)

A∈X⊆2Θ B∈X

γBel γBel + γP l

(17)

> d ` a b < , y z I J £ ¤ ° { ­ ®  N ewP (A) = m(A) +

X

 m(A) P l(A) α P  m(X) + (1 − α) P m(B) P l(B)

4 5 6 7 ± ² 0 1 2 3 / ³ ´ q µ 1 y ¶ . / E ¹

l

·

n

0

1

2

3

I

J

¸

w E ¹ c 4 5 6 7 / @ . ½ ¾ ¿ Daniel P T (·) , ¥

º

» ¼ À f (A) = , g(A) = ,Á £ ¤ (18) { z I p ª X A∈X⊆2Θ ,|X|>1

B∈X

B∈X

[19]

Pm(A) m(B)

B∈X

(18)

.

PP l(A) P l(B)

B∈X

N ewP (A) = m(A) +

[αf (A) + (1 − α)g(A)]m(X)

A∈X⊆2Θ ,|X|>1

(19)

Â

11

Ã

ÄÆÅÆÇ , È : ÆÉ ÊÆËÆÌÆÍÆÎÆÏÆÐÆÑÆÒÆÓÆÔÆÕÆÖ

à á

1

× 1 ÆØ ÙÆÚÆÛÆÜÆÝÆÞÆß

Y

ç ½ ¡ P T (·) â ã ä ¥ å æ ,O

è éëê

Bel(A) ≤ P T (A) ≤ P l(A) X

P1 (A) = m(A) +

Á ¯ ª

(20)

f (A)m(X), P2 (A) = m(A) +

A∈X⊆2Θ ,|X|>1

° ¯ ª

2935

X

g(A)m(X),

A∈X⊆2Θ ,|X|>1

N ewP (A) = αP1 (A) + (1 − α)P2 (A), X

P1 (A) = m(A) +

f (A)m(X) ≥ m(A) = Bel(A),

A∈X⊆2Θ ,|X|>1

P1 (A) = m(A) +

y z ì . { í ° ¯ ª N ewP (A) >

X

f (A)m(X) ≤ m(A) +

A∈X⊆2Θ ,|X|>1

A∈X⊆2Θ ,|X|>1

Bel(A) ≤ P1 (A) ≤ P l(A),

P1 (A)

/ î ½ ? @ , y z

Bel(A) ≤ P2 (A) ≤ P l(A), P2 (A)

X

Bel(A) ≤ N ewP (A) ≤ P l(A).

m(X) = P l(A),

 35 ø ï ð ñ ò ó ô õ ö ÷

ù ú 1 û c ü ý þ BPA, P T (·) â ã ä 2 3 Y ç ½ , O P T (A) = m(A) (21) V ^

£ ¤ è éë¯ ª m(A) > ü ý þ BPA, y z m(X) (19) { ÿ N ewP (A) = m(A). ù ú 2 c _ { A, P T (·) â ã ä = 0, 0 (22) ^ èBé ¯ ª A > ª { , yBz Bel(A)P T=(A)0, P=l(A) = 0, Bel(A) ≤ N ewP (A) ≤ P l(A) {Bÿ N ewP (A) = 0. à á 2 c _ © Θ ¡ ¸ R(·), R(Θ) → Θ, P T (·) â ã ä 2936

¸ R(·) ª

è éë¯ ª

P T ∗ (R(A)) = P T (R(A))

, R(Θ) → Θ,

(23)

y z

R(A) = A, m∗ (R(A)) = m(A),

y z

X

N ewP ∗ (R(A)) = m∗ (R(A)) +

[αf (A) + (1 − α)g(A)]m(X),

R(A)∈X⊆2Θ ,|X|>1

y z

X

N ewP ∗ (R(A)) = m(A) +

[αf (A) + (1 − α)g(A)]m(X) = N ewP (A).

¡ y ¢ , 4 5 6 7 ã ä Daniel l · 0 1 2 3 I J ¸ E ¹ .

² U c 4 5 6 7 ± ² ¼ ´ ¾ ¿ , z c ¦ @ ½ ¾ ¿ í . ¥

º

» ¼ 1 À i © Θ = {θ , θ , θ , θ }, w ¥ 4 2 3 q h 1 y ¶ 1 ÆÚÆÛ A∈X⊆2Θ ,|X|>1

3

1

θ1 0.30

θ2 0.15

θ3 0.01

2

θ4 0.09

3

.

4

θ1 ∪ θ 2 ∪ θ 3 0.25

θ1 ∪ θ 2 ∪ θ 4 0.05

θ2 ∪ θ 3 ∪ θ 4 0.05

θ1 ∪ θ 2 ∪ θ 3 ∪ θ 4 0.10

4 5 6 7 ¾ ¿ q ¥ ± ² :

T«UBW NBIBJB0B1B2B3 6B7 , ¨B© ±B²B· θ , θ , θ , θ 0B1B2B3 P (θ ), P (θ ), P (θ ), 1) h 2 y ¶ .

P (θ ) q T«UB\B] NBIBJB0B1B2B3 6B7 , ¨B© ±B²B· θ , θ , θ , θ 0B1B2B3 P (θ ), P (θ ), P (θ ), h 3 y ¶ .

P (θ ) q j d ` a b < BPA ! ® > b < 2 3 ¨ " X, j TVU W N I J 6 7 TVU \ ] N I J 6 7 ÿ # 2) 0 1 2 3 ¨ © ! ® > 2 3 ¨ " Y $ Y q h 4 y ¶ .

4 % & 'ÆÚÆÛ ( ) 2 ØÆÙÆÚÆÛ P (θ ) 3 ØÆÙÆÚÆÛ P (θ ) θ θ θ θ m(·)

1

2

1

2

3

4

1

1

1

2

1

3

1

2

3

4

2

1

2

2

2

3

4

4

1

1

2

2

i

i

P1 (θ1 ) P1 (θ2 ) P1 (θ3 ) P1 (θ4 )

P2 (θ1 ) P2 (θ2 ) P2 (θ3 ) P2 (θ4 )

0.5453 0.3027 0.0193 0.1327

0.4740 0.3021 0.0918 0.1321

± ² T«UBW NBIBJB6B7 T«UB\B] NBIBJB6B7¬§VdB`Ba bB< B ¨ B © γ , γ . * ± ² { ÿ : 3)

Bel

1

2

BPA

Pl

HY1 (X) = 0.3676, HX (Y1 ) = 0.1081, HY2 (X) = 0.4200, HX (Y2 ) = 0.1140,

γBel =

4

IBJB 0B1B2B3 B

H(X) = 0.2047, H(Y1) = 0.1882, H(Y2) = 0.2372,

Á

3

X 0.3000 0.1500 0.0100 0.0900 Y1 0.5453 0.3027 0.0193 0.1327 Y2 0.4740 0.3021 0.0918 0.1321

H(X) + H(Y1 ) H(X) + H(Y2 ) = 0.8259, γP l = = 0.8277.

HY1 (X) + HX (Y1 ) HY2 (X) + HX (Y2 )

Â

11

Ã

4)

± ² ? @ A C

ÄÆÅÆÇ , È : ÆÉ ÊÆËÆÌÆÍÆÎÆÏÆÐÆÑÆÒÆÓÆÔÆÕÆÖ 5 ØÆÙÆÚÆÛ = 0.4995,

α. γBel α= γBel + γP l 1 − α = 0.5005.

5) P1 (θ1 ) P2 (θ1 ) P1 (θ2 ) P2 (θ2 ) P1 (θ3 ) P2 (θ3 ) P1 (θ4 ) P2 (θ4 ) , θ1 , θ2 , θ3 , θ4 P (θ1 ), P (θ2 ), P (θ3 ), P (θ4 ) 5 .

6 , DSmP ε , ε = 0.5.

6 , θ1 . BetP CuzzP , ,

Á

¨B© c

P (θi )

P (θ1 )

P (θ2 )

P (θ3 )

P (θ4 )

0.5096

0.3024

0.0556

0.1324

6 = > ? @ A B C D > ? E F $ ¾ ¿ î ½ ? @ $

ÿ $ 0 1 2 3 · θ θ θ θ γ

+ , h BetP(·) 0.4250 0.2917 0.1350 0.1483 0.8057 0.3500 0.3000 0.2050 0.1450 0.7441 q h § jy 4¶ 5 6 7 ¦.-./ } 6 7 ¾ ¿ n c T PFT(·) CuzzP(·) 0.4241 0.2897 0.1341 0.1521 0.8066 ^V_ 6 7 §10 2 3 4 D E k 5 3 § DSmP (·) 0.4530 0.2929 0.1096 0.1445 0.8219 6Bh PBQ 0.5453 0.3027 0.0193 0.1327 0.8259 · w U § 7 gB6B7 ÿ # PropBelP(·) B {

z !

PropPlP(·) 0.4740 0.3021 0.0918 0.1321 0.8277 6B7 G H ÕÆÖ (·) 0.5096 0.3024 0.0556 0.1324 0.8359 PBQ 8B>BbB< 0B1B2B3 + F 6 7 8 > j X ` a b < W 9 : l ; < d ` a b < k g ¨ I 6 ¤ J K ¨ | }ML1N ¨ W O 4 P ® W Q R , | _ 0 1 . PFT 6 7 ÿ # b < θ 0 1 2 3 P (θ ) = 0.35, b < θ 0 1 2 3 P (θ ) = 0.3 S T > x F , 4 _ 0 1 U V · @ 0 1 . DSmP 6 7 W ] X _ Y 7 g 6 7 , Z ÿ # P Q v > x [ . PropBelP 6 7 | } n L1N W , ÿ # PBQ \ \ >BxB , { >BwB6B7 IBJ ¼ ´ ¼ _B B , O 4 ]BF ^B1 _ ` . PropPlP 6B7 W ] v |B} n LaN W , Z ÿ # P Q T r , | _ V · @ ^ 1 . 4 5 6 7 ÿ # N γ T ¦ - / } 6 7 8 9 F , w 6 7 ÿ # P Q r BPA W , b c ÿ # ^ 1 2 3 Yed n ] $ | _ ^ 1 .

2937

1

2

1

2

3

4

ε=0.5

1

2

f g h i

_ D-S í j u } § , k k ^ 1 U ª n l X 6 m § 2 · Y < + X 6 m , n o p q } + F r U n Q R s + G , â 9 c ; < 6 m ¾ ¿ t , 6 b V · @ ^ 1 . Z > x X u v ¥ , y w x W > x y º , ^ 1 U x z V · ^ 1 , b ^ 1 2 3 I J r : ^ k < ; < , ª ^ 1 U { | } D $ { ~ P Q . 4 Y < ^ 1 2 3 I J 6 º u } , v z ^ 1 2 3 I J 9 ½ .

2 L1N n , c 6 4 < 6 x, y, z, w P ® nM VF , c L1w x9 W ¾ ¿ n t , ÿ # d kY 4 6 < 6 { # 4 2 3 i q 2 h 6 7m y ¶ h . ª n ° k 4 < Q 6 , â Y ¾ ¿ , Z , â 9 · Y E s 3 , z q Q , y ,z ì 6 ° ¡ P q ® ì8 y ´ ¶ N . Q N R , q s ¸ q h 9 y ¶ . t { ¢ N£ 4 5 6 7 ¾ ¿ ± ² , ÿ # k 4 < 6 { # 2 3 q h 10 y ¶ .

7 ¤ ¥ ¦ § ¨ © ' ÆÚÆÛ 8 ª « > x y z w x∪w x∪y∪z∪w ® ¯ ° ± {x,Õ ­ y, z}1 {x,Õ ­y, w}2 {x,Õ ­ z, w}3 {y,Õ ­z, w}4 m(·) 0.15 0.20 0.10 0.10 0.20 0.25 9 ² ³ ´ µ 10 ¤ ¥ ¦ § ¨ © 'ÆÚÆÛ x y z w ® ¯ −10 x y z w −10 −10 −10 ¶ ® ¯ −20 −30 −60 −100 p(·) 0.3364 0.2744 0.1452 0.2440 s t # · ¸ Y 7 < c 6 ½ m g ¹B6 > m+ X Q 6 R m s, ¾Bn ¿ ±B6 ² R Eº # < + 6 X m > , n X Q 6 R m » # + G ¼ ? { k | 5 @

, ¾ ¸ , ª + { z ^ » 1 ¼ e o f .p 6 m 1 Q R s : 0.3364 × (−10) + 0.2744 × (−10) + 0.1452 × (−10) + 0.2440 × (−100) = −31.96.

6 m 2 Q R s : 0.3364 × (−10) + 0.2744 × (−10) + 0.1452 × (−60) + 0.2440 × (−10) = −17.26.

6 m 3 Q R s : 0.3364 × (−10) + 0.2744 × (−30) + 0.1452 × (−10) + 0.2440 × (−10) = −15.49.

6 m 4 Q R s : 0.3364 × (−20) + 0.2744 × (−10) + 0.1452 × (−10) + 0.2440 × (−10) = −13.36.

6 ¡ º ±B²BPBQ z ! · , 6 m 4 Q R s + G , yBzB 6BuBw 2 ¿ 6 m 4 ¾B¿ , O c < 6 ¾ ¿ . y, z, w 3 4

2938

Ă&#x201A;

ï ð ù ò ó ô þ Ü á

35

ø

Ă&#x20AC; Ă 4 5 { ¡ ĂŽ ½ ? @ 6 7 Â&#x2022; Â&#x2013; @ Â&#x2014; c 4 2 3 ž Âż ^ Ă&#x201A; 2 3 I J , n I J K M N Ă&#x192; Â&#x203A; Â&#x153; v Â&#x17E; Â&#x; , Ăż # I J P Q Â? Ă&#x201E; @ q . w 6 7 Ă&#x2026; Â&#x2022; Â&#x2013; q Â&#x2014; | }ML1N W Â&#x161; , Âť Ă&#x2020; ^ Ă&#x201A; _ ` Ă&#x2021; X Ă&#x2C6; , Ă&#x2030; _ ^ Ă&#x201A; U V ¡ @ ^ Ă&#x201A; , { z r Â&#x2018; Â&#x2014; u } # W Â&#x161; Ă&#x160; @ $Ă&#x152;Ă&#x2039; Ă? Ă&#x17D; Ă? Ă&#x2021; Ă? X _ D-S Ă­ j Ă&#x2018; Ă&#x2019; ^ Ă&#x201A; Â&#x2026; Â&#x161; § . Ă&#x201C; Ă&#x201D; Ă&#x2022; Ă&#x2013; 5

[1] Dempster A P. Upper and lower probabilities induced by a multivalued mapping[J]. Annals of Mathematical Statistics, 1967, 38(2): 325â&#x20AC;&#x201C;339.

[2] Shafer G. A mathematical theory of evidence[M]. NewJersey: Princeton University Press, 1976.

[3] Leung Y, Ji N N, Ma J H. An integrated information fusion approach based on the theory of evidence and group decision-making[J]. Information Fusion, 2013, 14(4): 410â&#x20AC;&#x201C;422.

[4] Luo H, Yang S L, Hu X J, et al. Agent oriented intelligent fault diagnosis system using evidence theory[J].

Expert Systems with Applications, 2012, 39(3): 2524â&#x20AC;&#x201C;2531.

[5] Li B, Pang F W. An approach of vessel collision risk assessment based on the D-S evidence theory[J]. Ocean Engineering, 2013, 74: 16â&#x20AC;&#x201C;21. , .

[J]. , 2010, 30(11): 2013â&#x20AC;&#x201C;2018.

[6] He Jinfeng, Xu Songjie. Approach of quality performance measurement and assessment under uncertainty[J].

Systems Engineering â&#x20AC;&#x201D; Theory & Practice, 2010, 30(11): 2013â&#x20AC;&#x201C;2018.

[7] Smets P, Kennes R. The transferable belief model[J]. Artificial Intelligence, 1994, 66(4): 191â&#x20AC;&#x201C;234.

[8] Cuzzolin F. On the properties of the intersection probability[EB/OL]. [2013-12-01]. http://perception.inrialpes.fr/ people/Cuzzolin.

[9] Cobb B R, Shenoy P P. On the plausibility transformation method for translating belief function models to probability models[J]. International Journal of Approximate Reasoning, 2006, 41(3): 314â&#x20AC;&#x201C;330.

[10] , , .

[J]. , 2011, 39(3A): 121â&#x20AC;&#x201C;125.

Xu Peida, Hang Deqiang, Deng Yong. An optimal transformation of basic probability assignment to probability[J]. Acta Electronica Sinica, 2011, 39(3A): 121â&#x20AC;&#x201C;125.

[11] Hu L F, He Y, Guan X, et al. A new probabilistic transformation in generalized power space[J]. Chinese Journal of Aeronautics, 2011, 24(4): 449â&#x20AC;&#x201C;460.

[12] , , , .

[J]. , 2013, 28(8): 1214â&#x20AC;&#x201C;1218.

Wang Wanqing, Zhao Yongjun, Huang Jie, et al. Transformation of basic probability assignment to probability based on uncertainty degree[J]. Control and Decision, 2013, 28(8): 1214â&#x20AC;&#x201C;1218.

[13] , , , . D-S [J]. , 2013, 31(2): 295â&#x20AC;&#x201C; 299.

Jiang Wen, Wu Cuicui, Jia Jia, et al. A probabilistic transformation of basic probability assignment(BPA) in D-S evidence theory[J]. Journal of Northwestern Polytechnical University, 2013, 31(2): 295â&#x20AC;&#x201C;299.

[14] Sudano J J, Martin L. Yet another paradigm illustrating evidence fusion(YAPIEF)[C]// 2006 9th International Conference on Information Fusion, IEEE, 2006: 1â&#x20AC;&#x201C;7.

[15] Dezert J, Smarandache F. A new probabilistic transformation of belief mass assignment[C]// 2008 11th International Conference on Information Fusion, IEEE, 2008: 1â&#x20AC;&#x201C;8.

[16] Daniel M. Consistency of probabilistic transformations of belief functions[C]// Proceedings of the 10th International Conference IPMU. Perugia, 2004: 1135â&#x20AC;&#x201C;1142.

[17] , .

[J]. ( ), 1998, 2: 28â&#x20AC;&#x201C;35.

Wu Minjin, Bai Zhijiang. Relative entropy and its application[J]. Journal of East China Normal University (Natural Science), 1998, 2: 28â&#x20AC;&#x201C;35.

[18] , , , .

[J]. , 2011, 28(6): 839â&#x20AC;&#x201C;844.

Deng Yong, Wang Dong, Li Qi, et al. A new method to analyze evidence conflict[J]. Control Theory & Applications, 2011, 28(6): 839â&#x20AC;&#x201C;844.

[19] Daniel M. Probabilistic transformations of belief functions[C]// Symbolic and Quantitative Approaches to Reasoning with Uncertainty. Springer Berlin Heidelberg, 2005: 539â&#x20AC;&#x201C;551.

Ă&#x2014; Ă&#x2DC; Ă&#x2122; Ă&#x161; Ă&#x203A; Ă&#x153; Âś Ă? Ă&#x17E; Ă&#x; Ă ĂĄ Ă&#x17D; â ĂŁ ä ĂĽ ĂŚ çĂ&#x2020;Ă&#x2022;Ă&#x2020;Ă&#x2013;

ĂŻĂ&#x2020;Ă°Ă&#x2020;ĂąĂ&#x2020;òĂ&#x2020;ĂłĂ&#x2020;Ă´Ă&#x2020;ĂľĂ&#x2020;ĂśĂ&#x2020;á

è ĂŠ ĂŞ ĂŤ ĂŹ Ă­ ĂŽ ĂŻ Ă° ĂąĂ&#x2020;Ă&#x2030; G Ă&#x2018;Ă&#x2020;Ă&#x2019; ò ĂłĂ&#x2020;Ă&#x201C;Ă&#x2020;Ă&#x201D; Ă´Ă&#x2020;Ă&#x2018;Ă&#x2020;Ă&#x2019;Ă&#x2020;Ă&#x17D; Ăľ Ăś áĂ&#x2020;Ă&#x2022;Ă&#x2020;Ă&#x2013;

øĂ&#x152;Ăš Ăş Ăť

Ăź Ă˝ Ăž Ă&#x201E; Ăż Ă&#x2C6; Ă&#x2030;Ă&#x2020;Ă&#x160; Âś Ă? Ă&#x17E; Ă?Ă&#x2020;Ă&#x17D;Ă&#x2020;Ă&#x2030; G Ă&#x2020;Ă&#x2018; Ă&#x2019; ò ĂłĂ&#x2020;Ă&#x2018;Ă&#x2020;Ă&#x2019;Ă&#x2020;Ă&#x201C;Ă&#x2020;Ă&#x201D;Ă&#x2020;Ă&#x2022; Ă&#x2013; Ă&#x2020;ĂľĂ&#x2020;Ă?Ă&#x2020;Ă?

Ă&#x2C6;

Ă&#x2020;ĂłĂ&#x2020;Ă´ Ă&#x2020;Ă&#x17D;Ă&#x2020;Ă&#x2030; G Ă&#x2018;Ă&#x2020;Ă&#x2019; ò ĂłĂ&#x2020;Ă&#x201C; Ă&#x201D;Ă&#x2020;Ă&#x2018; Ă&#x2019;Ă&#x2020;Ă&#x2022; Ă&#x2013; Ă&#x2020;Ăą Ăş Ăş Ăť

Ă&#x2DC; Ă&#x2039;Ă&#x2020;Ă&#x152;

! " # Ăş Ăş Ăť $ % & Ăş '

ĂŽ ĂŻ Ăź ( ) * Ă&#x2C6; Ă° ĂąĂ&#x2020;Ă&#x2021;Ă&#x2020;Ă&#x17D; + , - .Ă&#x2020;Ă&#x2022;Ă&#x2020;Ă&#x2013;

Ă&#x2020;ĂłĂ&#x2020;Ă´Ă&#x2020;Ăľ