通信学报 ›› 2018, Vol. 39 ›› Issue (1): 117-125.doi: 10.11959/j.issn.1000-436x.2018003
朱京伟,王晓丹,宋亚飞,黄文龙
修回日期:
2017-12-02
出版日期:
2018-01-01
发布日期:
2018-02-07
作者简介:
朱京伟(1990-),男,陕西咸阳人,空军工程大学博士生,主要研究方向为模式识别、智能信息处理等。|王晓丹(1966-),女,陕西汉中人,空军工程大学教授、博士生导师,主要研究方向为机器学习、模式识别、智能信息处理等。|宋亚飞(1988-),男,河南汝州人,空军工程大学博士生,主要研究方向为模式识别、智能信息处理等。|黄文龙(1973-),男,重庆人,空军工程大学副教授,主要研究方向为智能信息处理、图像解译等。
基金资助:
Jingwei ZHU,Xiaodan WANG,Yafei SONG,Wenlong HUANG
Revised:
2017-12-02
Online:
2018-01-01
Published:
2018-02-07
Supported by:
摘要:
针对现有的相似性/相异性测度在量化证据冲突时存在的不足,定义一种新的被称为幂Pignistic概率距离的相异性测度,并提出基于幂Pignistic概率距离的加权证据组合方法。该方法通过幂Pignistic概率距离量化两证据之间的冲突程度,然后建立相似性矩阵并求得各证据的可信度,再用加权平均法修正证据,最后利用Dempster规则进行组合。数值算例的结果表明,所提方法是合理有效的。
中图分类号:
朱京伟,王晓丹,宋亚飞,黄文龙. 基于幂Pignistic概率距离的加权证据组合方法[J]. 通信学报, 2018, 39(1): 117-125.
Jingwei ZHU,Xiaodan WANG,Yafei SONG,Wenlong HUANG. Weighted evidence combination method based on power-Pignistic probability distance[J]. Journal on Communications, 2018, 39(1): 117-125.
表2
6种相似性测度的值"
事件 | 1?dBet | simTa | 1?dSP | 1?dJ | cosθ | 1?dPBet |
1 | 0.333 3 | 0.333 3 | 0.428 6 | 0.382 8 | 0.378 0 | 0.258 5 |
2 | 0.333 3 | 0.333 3 | 0.406 0 | 0.376 4 | 0.355 1 | 0.252 4 |
3 | 0.333 3 | 0.333 3 | 0.383 5 | 0.369 5 | 0.327 5 | 0.246 3 |
4 | 0.333 3 | 0.333 3 | 0.360 9 | 0.362 0 | 0.296 9 | 0.240 2 |
5 | 0.333 3 | 0.333 3 | 0.338 3 | 0.354 0 | 0.265 2 | 0.234 2 |
6 | 0.333 3 | 0.333 3 | 0.315 8 | 0.345 4 | 0.234 1 | 0.228 1 |
7 | 0.333 3 | 0.333 3 | 0.293 2 | 0.336 4 | 0.204 6 | 0.222 0 |
8 | 0.333 3 | 0.333 3 | 0.270 7 | 0.327 0 | 0.177 2 | 0.215 9 |
9 | 0.333 3 | 0.333 3 | 0.248 1 | 0.317 0 | 0.152 3 | 0.209 8 |
10 | 0.333 3 | 0.333 3 | 0.225 6 | 0.306 7 | 0.129 9 | 0.203 7 |
11 | 0.333 3 | 0.333 3 | 0.203 0 | 0.295 9 | 0.109 7 | 0.197 6 |
12 | 0.333 3 | 0.333 3 | 0.180 5 | 0.284 8 | 0.091 7 | 0.191 6 |
13 | 0.333 3 | 0.333 3 | 0.157 9 | 0.273 3 | 0.075 6 | 0.185 5 |
14 | 0.333 3 | 0.333 3 | 0.135 3 | 0.261 4 | 0.061 2 | 0.179 4 |
15 | 0.333 3 | 0.333 3 | 0.112 8 | 0.249 1 | 0.048 2 | 0.173 3 |
16 | 0.333 3 | 0.333 3 | 0.090 2 | 0.236 6 | 0.036 5 | 0.167 2 |
17 | 0.333 3 | 0.333 3 | 0.067 7 | 0.223 8 | 0.026 0 | 0.161 1 |
18 | 0.333 3 | 0.333 3 | 0.045 1 | 0.210 6 | 0.016 5 | 0.155 0 |
19 | 0.333 3 | 0.333 3 | 0.022 6 | 0.197 2 | 0.007 9 | 0.148 9 |
20 | 0.333 3 | 0.333 3 | 0.333 3 | 0.183 5 | 0 | 0.142 9 |
表5
8种证据组合方法的结果对比"
方法 | m1 ,m2 | m1 ,m2 ,m3 | m 1,m2,m3,m4 | m 1,m2,m3,m4,m5 |
m (A)=0 | m (A)=0 | m (A)=0 | m (A)=0 | |
Dempster规则 | m(B)=0.857 1 | m(B)=0.631 6 | m(B)=0.328 8 | m(B)=0.140 4 |
m(C)=0.142 9 | m(C)=0.368 4 | m(C)=0.671 2 | m(C)=0.859 6 | |
m (A)=0 | m (A)=0 | m (A)=0 | m (A)=0 | |
Yager法 | m(B)=0.18 | m(B)=0.018 | m(B)=0.001 8 | m(B)=0.000 18 |
m(C)=0.03 | m(C)=0.010 5 | m(C)=0.003 68 | m(C)=0.00110 | |
m(ABC)=0.79 | m(ABC)=0.971 5 | m(ABC)=0.994 52 | m(ABC)=0.998 72 | |
m (A)=0 | m(A)=0.637 3 | m(A)=0.902 6 | m(A)=0.963 1 | |
Wen法(cosθ法) | m(B)=0.857 1 | m(B)=0.143 6 | m(B)=0.007 3 | m(B)=0.001 1 |
m(C)=0.142 9 | m(C)=0.219 1 | m(C)=0.083 2 | m(C)=0.026 0 | |
m(AC)=0.007 0 | m(AC)=0.009 8 | |||
m (A)=0 | m(A)=0.636 4 | m(A)=0.892 4 | m(A)=0.960 5 | |
Yu法(dSP法) | m(B)=0.857 1 | m(B)=0.142 4 | m(B)=0.008 2 | m(B)=0.001 1 |
m(C)=0.142 9 | m(C)=0.2211 | m(C)=0.070 0 | m(C)=0.020 6 | |
m(AC)=0.029 3 | m(AC)=0.017 8 | |||
m (A)=0 | m(A)=0.706 3 | m(A)=0.908 4 | m(A)=0.965 9 | |
simTa法 | m(B)=0.857 1 | m(B)=0.087 0 | m(B)=0.005 7 | m(B)=0.000 8 |
m(C)=0.142 9 | m(C)=0.206 7 | m(C)=0.083 0 | m(C)=0.027 0 | |
m(AC)=0.003 0 | m(AC)=0.006 3 | |||
m(A)=0.154 3 | m(A)=0.556 8 | m(A)=0.865 3 | m(A)=0.968 8 | |
Murphy法 | m(B)=0.746 9 | m(B)=0.356 2 | m(B)=0.089 1 | m(B)=0.015 6 |
m(C)=0.098 8 | m(C)=0.078 1 | m(C)=0.038 2 | m(C)=0.012 7 | |
m(AC)=0.008 8 | m(AC)=0.007 5 | m(AC)=0.002 9 | ||
m(A)=0.154 3 | m(A)=0.736 9 | m(A)=0.948 4 | m(A)=0.986 9 | |
d J 法 | m(B)=0.746 9 | m(B)=0.161 8 | m(B)=0.012 0 | m(B)=0.001 0 |
m(C)=0.098 8 | m(C)=0.091 5 | m(C)=0.031 0 | m(C)=0.008 8 | |
m(AC)=0.009 8 | m(AC)=0.008 6 | m(AC)=0.003 2 | ||
m(A)=0.154 3 | m(A)=0.764 0 | m(A)=0.952 0 | m(A)=0.987 4 | |
本文(dPBet法) | m(B)=0.746 9 | m(B)=0.135 1 | m(B)=0.009 5 | m(B)=0.000 8 |
m(C)=0.098 8 | m(C)=0.090 1 | m(C)=0.029 6 | m(C)=0.008 5 | |
m(AC)=0.010 8 | m(AC)=0.009 0 | m(AC)=0.003 3 |
[1] | DEMPSTER A P . Upper and lower probabilities induced by a multi-valued mapping[J]. Annals of Mathematical Statistics, 1967,38(4): 325-339. |
[2] | ZADEH L . A simple view of the Dempster-Shafer theory of evidence and its implication for the rule of combination[J]. AI Magazine, 1986,7(2): 85-90. |
[3] | YAGER R R . On the Dempster-Shafer framework and new combination rules[J]. Information Sciences, 1987,41(2): 93-137. |
[4] | SMETS P . The combination of evidence in the transferable belief model[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990,12(5): 447-458. |
[5] | 孙全, 叶秀清, 顾伟康 . 一种新的基于证据理论的合成公式[J]. 电子学报, 2000,28(8): 117-119. |
SUN Q , YE X Q , GU W K . A new combination rule of evidence[J]. Acta Electronica Sinica, 2000,28(8): 117-119. | |
[6] | LEFEVRE E , COLOT O , VANNOORENBERGHE P . Belief functions combination and conflict management[J]. Information Fusion, 2002,3(2): 149-162. |
[7] | FU Y W , YANG W . Combination rule based on proportional redistribution of generalized conflict for evidence theory[J]. Chinese Journal of Electronics, 2014,23(3): 533-538. |
[8] | 权文, 王晓丹, 王坚 ,等. 一种基于局部冲突分配的 DST 组合规则[J]. 电子学报, 2012,40(9): 1880-1884. |
QUAN W , WANG X D , WANG J ,et al. New combination rule of DST based on local conflict distribution strategy[J]. Acta Electronica Sinica, 2012,40(9): 1880-1884. | |
[9] | 陈圣群, 王应明 . 证据的分组合成法[J]. 控制与决策, 2013,28(4): 574-578. |
CHEN S Q , WANG Y M . Grouping method for combining evidence[J]. Control and Decision, 2013,28(4): 574-578. | |
[10] | 张鑫, 牟龙华 . 基于局部冲突消除的证据合成法则[J]. 系统工程与电子技术, 2016,38(7): 1594-1599. |
ZHANG X , MU L H . Evidence combination rule based on local con-flict elimination[J]. Systems Engineering and Electronics, 2016,38(7): 1594-1599. | |
[11] | 陈雁飞, 夏学知, 胡丹桂 ,等. 基于试探折扣修正的证据组合方法[J]. 电子学报, 2014,42(4): 756-765. |
CHEN Y F , XA X Z , HU D G ,et al. Evidence combination based on tentative discount of evidences[J]. Acta Electronica Sinica, 2014,42(4): 756-765. | |
[12] | 罗贺, 尹艳平, 胡笑旋 ,等. 基于可信因子的证据源修正方法[J]. 系统工程与电子技术, 2015,37(6): 1459-1464. |
LUO H , YIN Y P , HU X X ,et al. Method to modify evidence source based on trustworthy factors[J]. Systems Engineering and Electronics, 2015,37(6): 1459-1464. | |
[13] | WEN C L , WANG Y C , XU X B . Fuzzy information fusion algorithm of fault diagnosis based on similarity measure of evidence[C]// Advances in Neural Networks Lecture Notes in Computer Science. 2008: 506-515. |
[14] | YU C , YANG J H , YANG D B ,et al. An improved conflicting evidence combination approach based on a new supporting probability distance[J]. Expert Systems with Applications, 2015,42(12): 5139-5149. |
[15] | 毕文豪, 张安, 李冲 . 基于新的证据冲突衡量的加权证据融合方法[J]. 控制与决策, 2016,31(1): 73-78. |
BI W H , ZHANG A , LI C . Weighted evidence combination method based on new evidence conflict measurement approach[J]. Control and Decision, 2016,31(1): 73-78. | |
[16] | MURPHY C K . Combining belief functions when evidence conflicts[J]. Decision Support Systems, 2000,29(1): 1-9. |
[17] | 邓勇, 施文康, 朱振福 . 一种有效处理冲突证据的组合方法[J]. 红外与毫米波学报, 2004,23(1): 27-32. |
DENG Y , SHI W K , ZHU Z F . Efficient combination approach of con-flict evidence[J]. J Infrared Millim Waves, 2004,23(1): 27-32. | |
[18] | 肖建于, 童敏明, 朱昌杰 ,等. 基于 Pignistic 概率距离的改进证据组合规则[J]. 上海交通大学学报, 2012,46(4): 636-645. |
XIAO J Y , TONG M M , ZHU C J ,et al. Improved combination rule of evidence based on Pignistic probability distance[J]. Journal of Shang-hai Jiaotong University, 2012,46(4): 636-645. | |
[19] | JOUSSELME A L , MAUPIN P . Distances in evidence theory:comprehensive survey and generalizations[J]. International Journal of Approximate Reasoning, 2012,53(2): 118-145. |
[20] | JOUSSELME A L , GRENIER D , BOSSE E . A new distance between two bodies of evidence[J]. Information Fusion, 2001,2(1): 91-101. |
[21] | TESSEM B . Approximations for efficient computation in the theory of evidence[J]. Artificial Intelligence, 1993,61(2): 315-329. |
[22] | 李昌玺, 周焰, 王盛超 ,等. 多源信息融合中一种新的证据合成算法[J]. 上海交通大学学报, 2016,50(7): 1125-1131. |
LI C X , ZHOU Y , WANG S C ,et al. A novel combination rule of evi-dence theory in multi-source information fusion[J]. Journal of Shang-hai Jiaotong University, 2016,50(7): 1125-1131. | |
[23] | 王亮, 吕卫民, 滕克难 ,等. 基于分类修正的多证据合成方法[J]. 控制与决策, 2015,30(1): 125-130. |
WANG L , LV W M , TENG K N ,et al. Combination method of mul-ti-evidence based on classification correction[J]. Control and Decision, 2015,30(1): 125-130. | |
[24] | HAN D Q , DENG Y , HAN C Z ,et al. Some notes on betting commitment distance in evidence theory[J]. Science China-Information Sciences, 2012,55(3): 558-565. |
[1] | 王欣, 付威. 基于推土机距离的证据冲突强度量方法[J]. 通信学报, 2022, 43(5): 204-213. |
[2] | 杨晓萍,廉伟健,李孟杰,钱志鸿. 基于Pignistic概率函数和相关系数的证据组合方法[J]. 通信学报, 2019, 40(3): 73-79. |
[3] | 宋亚飞,王晓丹,雷蕾,薛爱军. 基于信任度和虚假度的证据组合方法[J]. 通信学报, 2015, 36(5): 98-103. |
[4] | 刘卓超,杨力,姜奇,王巍,曹春杰. 基于信任评估的Ad Hoc网络关键节点选取算法[J]. 通信学报, 2014, 35(Z2): 213-219. |
[5] | 刘卓超,杨 力,姜 奇,王 巍,曹春杰. 基于信任评估的Ad Hoc网络关键节点选取算法[J]. 通信学报, 2014, 35(Z2): 29-219. |
[6] | 宋亚飞,王晓丹,雷蕾,薛爱军. 基于相关系数的证据冲突度量方法[J]. 通信学报, 2014, 35(5): 95-100. |
[7] | 宋亚飞, 王晓丹, 雷蕾, 薛爱军. 基于相关系数的证据冲突度量方法[J]. 通信学报, 2014, 35(5): 13-100. |
[8] | 张琳,刘婧文,王汝传,王海艳. 基于改进D-S证据理论的信任评估模型[J]. 通信学报, 2013, 34(7): 167-173. |
[9] | 张琳1,2,3,刘婧文1,王汝传1,2,3,王海艳1,2,3. 基于改进D-S证据理论的信任评估模型[J]. 通信学报, 2013, 34(7): 19-173. |
[10] | 秦艳琳,吴晓平,高键鑫. 分布式环境下信任路径选择性搜索及聚合研究[J]. 通信学报, 2012, 33(Z1): 148-156. |
[11] | 蒋黎明,张琨,徐建,廖俊,张宏. 证据信任模型中的信任传递与聚合研究[J]. 通信学报, 2011, 32(8): 91-100. |
[12] | 杨凯,马建峰,杨超. 无线网状网中基于D-S证据理论的可信路由[J]. 通信学报, 2011, 32(5): 89-96. |
[13] | 臧天宁,云晓春,张永铮,门朝光,崔翔. 基于通信特征和D-S证据理论分析僵尸网络相似度[J]. 通信学报, 2011, 32(4): 66-76. |
[14] | 陈深龙,张玉清. 增强ad hoc网络可生存性的健壮多维信任模型[J]. 通信学报, 2010, 31(5): 1-9. |
[15] | 田俊峰,赵卫东,杜瑞忠,蔡红云. 新的入侵检测数据融合模型——IDSFP[J]. 通信学报, 2006, 27(6): 115-120. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||
|