《一种基于信息源可信度的证据组合新方法》由会员分享，可在线阅读，更多相关《一种基于信息源可信度的证据组合新方法（4页珍藏版）》请在金锄头文库上搜索。
1、30null7d0/Vol. 30 null No. 72008M7Systems Engineering and Electronics Jul. 2008cI|: 1001null506X(2008) 07null1210null04l: 2007 10 31;: 2008 01 25bTe:f( 1959 ) ,3,q,pV,1Z_dy_bEnullm ail: kgdlxyzdl 163. comBVFZE张多林1 , 潘null 泉1 , 张洪才1 , 邢清华2( 1. 西北工业大学自动化学院,陕西西安 710072; 2. 空军工程大学导弹学院, 陕西三原 713800)null
2、nullKnull1: DnullS证据组合规则在处理高冲突度信息时,会得出与直觉相反的结论,进而遭到质疑,也引发了很多专家学者的研究b根据指挥员在作战中处理冲突证据的一般习惯,提出了一种基于信息源可信度的证据组合新方法b建立了信息源可信度的计算模型,给出了基于信息源可信度的按比例分配冲突度的证据组合推理规则,算例验证结果表明,新的组合推理规则在处理冲突方面的性能得到了明显改进b1oM: DnullS理论; 证据组合规则; 信息源可信度; 冲突信息; 信息融合ms|: T P 212. 9 null null null nullDS: ACombination rule of evidence
3、 theory based on credibility of sensorZH ANG Duonulllin1, PAN Quan1, ZH ANG H ongnullcai1 , XIN G Qingnullhua2(1. Coll. of Automation, NorthwesternPolytechnical Univ., Xinull an710072, China;2. MissileInst. of Air ForceEngineering Univ., Sanyuan713800, China)null null Abstract: The DnullS combinatio
4、n rule is doubted for concluding error result s in dealing wit h high conflict innullformat ion, therefore it is studied by many experts and scholars. According t o the common habit of commandersin dealing with conf lict evidence in operation, an improved combinat ion rule based on credibility of se
5、nsor ispresent ed. T he computing model of credibilit y of sensor is built first ly, then a modified combination rule of disnullt ributing conflict in proportion based on credibility of sensor is given. T he numerical example shows t hat thenew combinat ion rule has good performance in dealing w ith
6、 conflict and improves the reliability and rationalit y oft he combination results.Keywords: DnullS theory; evidence combination rule; credibility of sensor; conflict information; informat ionf usion0nullnullnull nullCLW,+Y5,Ri5,7)/)aa4robC)/,ZE1bDempnullsternullShafer,Dempsterd95Hn54,iShaferB?ZBiV7
7、S%wb(eDnullS)z$,a5,7OVUaFZB,X$ 00, crd1 nullm1(A)+ crd2 nullm2(Z) = 0(8)nullnull(A,Z)= crd2 nullm2(A) nullm2(A) nullm1(Z)crd2 nullm2(A)+ crd1 nullm1(Z), crd2 nullm2(A)+ crd1 nullm1(Z) 00, crd2 nullm2(A)+ crd1 nullm1(Z) = 0(9)null nullZED1,7O?8In4V,BzZEb2. 3nullLL!d.l 1m1(A) = 0. 98,m1(B) = 0.01,m1(C
8、) = 0.01m2(A) = 0,m2(B) = 0. 01,m2(C) = 0. 99m3(A) = 0. 90,m3(B) = 0,m3(C) = 0. 10m4(A) = 0. 90,m4(B) = 0,m4(C) = 0. 10null null+FZE9FTnV4UbZE9FTH,T(3)98W,TdBPA(m1,m2) = 0.461 976,dBPA(m1,m3) = 0.040 277,dBPA (m1,m4) = 0. 040 277,dBPA (m2,m3) = 0. 421 927,dBPA (m2,m4) = 0. 421 927,dBPA(m3,m4) = 0.00
9、0 00T(4)98supm1 = 2. 457 470,supm2= 1. 694 170,supm3= 2. 537 796,supm4 = 2.537 796798VDcrd(m1) = 0. 819 157,Dcrd(m2)= 0.564 723,Dcrd(m3)= 0. 845 932,Dcrd(m4) = 0.845 932null nullIn.VH,L!.V0,In.V,9|.VT8V,F?5(6)#M1T(7)- (9),9FTV4KBbnull7f:BVFZEnull 1213null null nullV4null+F?5lFTF?5m(A) m(B) m(C) m( n
10、ull)DnullSF?5m1,m2m1,m2,m3m1,m2,m3,m40000. 01000. 9911000Yagerm1,m2m1,m2,m3m1,m2,m3,m40000. 000 1000. 009 90. 000 990. 000 0990. 990. 999 010. 999 901D9m1,m2m1,m2,m3m1,m2,m3,m40. 180. 3210. 420. 0040. 0030. 0030. 1940. 1880. 1810. 6220. 4880. 396D10m1,m2m1,m2,m3m1,m2,m3,m40. 485 10. 6260. 694 90. 01
11、0. 006 70. 0050. 504 90. 367 30. 300 1000D8m1,m2m1,m2,m3m1,m2,m3,m40. 009 70. 486 30. 596 70. 000 30. 002 70. 002 30. 018 90. 0610. 071 30. 971 10. 467 90. 329 7D1m1,m2m1,m2,m3m1,m2,m3,m40. 492 338 570. 776 331 600. 928 704 860. 000 347 990. 000 000 240. 000 000 000. 507 313 430. 223 668 160. 071 295 11000ZEm1,m2m1,m2,m3m1,m2,m3,m40. 581 638 220. 844 486 420. 955 970 640. 000 352 230. 000 000 200. 000 000 000. 418 009 580. 155 513 410. 044 029 41000null nullV4VA, DnullSZE?lrF, YagerFTV,BBB50H,5C5v,FT90,AfbD 1, 10|ls,5AvC,FAv9v,5l,C,F5vhlbD 8 9|sls,ssbVTnV4KBbTV,InV,T|Fbl
