A Direct Fusion Algorithm for Multiple Pieces of Evidence Based on Improved Conflict Measure
-
摘要:
针对Jousselme证据距离函数不能较好描述证据局部冲突和不能对高冲突证据进行准确冲突度量的不足,该文首先提出改进的Jousselme证据距离函数,该函数基于能够较好描述证据之间局部冲突情况的非重合度对Jousselme证据距离函数进行改进,使其冲突度量结果随非重合度取值及其取值范围的变化按适当比例进行变化;其次,基于冲突系数和新改进Jousselme证据距离函数共同构建改进的融合冲突度量函数。在此基础上,对焦元权系数计算式进行改进,并依此对局部多维冲突信息进行按比例分配。理论及应用分析结果表明,新算法是一种适用性广泛且抗干扰性能好的证据融合算法。
-
关键词:
- Dempster证据组合规则 /
- 冲突度量 /
- 证据距离 /
- 非重合度
Abstract:In the light of the disadvantages that Jousselme’s evidential distance function can not describe the local conflicting information of evidence well and can not measure the conflict of high conflicting evidence accurately, an improved Jousselme’s evidential distance function is proposed. In the new function, Jousselme’s evidence distance function is improved by using the non-coincidence degree, which can better describe the local conflict of evidence, so that the conflict measure result of evidence varies proportionally with the value of the non-coincidence degree and the scope of its change. Secondly, an improved fusion conflict measure function is constructed based on the conflict coefficient and the new improved Jousselme’s evidential distance function. On this basis, the weight coefficient formula of focal element is improved, and the local multi-dimensional conflicting information is assigned proportionately. Theoretical and application analysis results show that the new algorithm is a kind of evidence fusion algorithm with wide applicability and good anti-jamming performance.
-
表 1 不同参数对情况1中证据的冲突度量结果
冲突度量参数 ${m_1}$与${m_1}$ ${m_2}$与${m_2}$ ${m_1}$与${m_2}$ ${d_J}$ 0 0 0.4500 $k$ 0.4600 0.5050 0.6850 ${n_c}$ 0 0 0.4500 ${d_{{\rm{PJ}}}}$ 0 0 0.6255 ${d_{{\rm{NIJ}}}}(\gamma {\rm{ = 4)}}$ 0 0 0.7618 ${d^f}(\gamma {\rm{ = 4)}}$ 0.2652 0.2964 0.7241 表 2 不同参数对情况2中证据的冲突度量结果
冲突度量参数 ${m_1}$与${m_1}$ ${m_2}$与${m_2}$ ${m_1}$与${m_2}$ ${d_J}$ 0 0 0.8352 $k$ 0 0.4050 0.9000 ${n_c}$ 0 0 0.9000 ${d_{{\rm{PJ}}}}$ 0 0 0.9352 ${d_{{\rm{NIJ}}}}(\gamma {\rm{ = 8)}}$ 0 0 0.9886 ${d^f}(\gamma {\rm{ = 8)}}$ 0 0.2287 0.9662 -
SHAFER G. A mathematical theory of evidence turns 40[J]. International Journal of Approximate Reasoning, 2016, 79: 7–25. doi: 10.1016/j.ijar.2016.07.009 HAN Deqiang, DEZERT J, and YANG Yi. Belief interval-based distance measures in the theory of belief functions[J]. IEEE Transactions on Systems Man & Cybernetics Systems, 2016, 99: 1–18. doi: 10.1109/TSMC.2016.2628879 韩德强, 杨艺, 韩崇昭. DS证据理论研究进展及相关问题探讨[J]. 控制与决策, 2014, 29(1): 1–11. doi: 10.13195/j.kzyjc.2013.0517HAN Deqiang, YANG Yi, and HAN Chongzhao. Advances in DS evidence theory and related discussions[J]. Control and Decision, 2014, 29(1): 1–11. doi: 10.13195/j.kzyjc.2013.0517 DESTERCKE S and BURGER T. Toward an axiomatic definition of conflict between belief functions[J]. IEEE Transactions on Cybernetics, 2013, 43(2): 585–596. doi: 10.1109/TSMCB.2012.2212703 ZADEH L A. 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. doi: 10.1142/9789814261302-0033 SONG Yafei, WANG Xiaodan, LEI Lei, et al. Credibility decay model in temporal evidence combination[J]. Information Processing Letters, 2015, 115(2): 248–252. doi: 10.3969/j.issn.1001-506X.2015.07.04 BI Wenhao, ZHANG An, and YUAN Yuan. Combination method of conflict evidences based on evidence similarity[J]. Journal of Systems Engineering and Electronics, 2017, 28(3): 503–513. doi: 10.21629/JSEE.2017.03.10 QIAN Jin, GUO Xingfeng, and DENG Yong. A novel method for combining conflicting evidences based on information entropy[J]. Applied Intelligence, 2017, 46(4): 1–13. doi: 10.1007/s10489-016-0875-y 包甜甜, 谢新连, 魏照坤. 新的证据冲突度量算法[J]. 控制理论与应用, 2017, 34(1): 41–48. doi: 10.7641/CTA.2017.60275BAO Tiantian, XIE Xinlian, and WEI Zhaokun. Novel approach for measuring the conflict between evidence[J]. Control Theory &Applications, 2017, 34(1): 41–48. doi: 10.7641/CTA.2017.60275 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. doi: 10.3321/j.issn:1001-9014.2004.01.006DENG Yong, SHI Wenkang, and ZHU Zhenfu. Efficient combination approach of conflict evidence[J]. Journal of Infrared and Millimeter Wave, 2004, 23(1): 27–32. doi: 10.3321/j.issn:1001-9014.2004.01.006 宋亚飞, 王晓丹, 雷蕾, 等. 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. doi: 10.11959/j.issn.1000-436x.2015104SONG Yafei, WANG Xiaodan, LEI Lei, et al. Evidence combination based on the degree of credibility and falsity[J]. Journal on Communications, 2015, 36(5): 98–103. doi: 10.11959/j.issn.1000-436x.2015104 LIU Zhigang, HAN Zhiwei, ZHANG Yang, et al. Multiwavalet packet entropy and its application in transmission line fault recognition and classification[J]. IEEE Transactions on Neural Networks and Learning Systems, 2014, 25(11): 2043–2052. doi: 10.1109/TNNLS.2014.2303086 DENG Yong. Generalized evidence theory[J]. Applied Intelligence, 2015, 43(3): 530–543. doi: 10.1007/s10489-015-0661-2 郭强, 何友, 李新德. 一种快速DSmT-DS近似推理融合方法[J]. 电子与信息学报, 2015, 37(9): 2040–2046. doi: 10.11999/JEIT150086GUO Qiang, HE You, and LI Xinde. Fast DSmT-DS approximate reasoning method[J]. Journal of Electronics &Information Technology, 2015, 37(9): 2040–2046. doi: 10.11999/JEIT150086 周莉, 郭伟震, 曹珊珊. 改进的局部冲突信息再分配算法[J]. 光电子激光, 2018, 29(4): 396–404. doi: 10.16136/j.joel.2018.04.0205ZHOU Li, GUO Weizhen, and CAO Shanshan. An improved local conflict information redistribution algorithm[J]. Journal of Optoelectronics Laser, 2018, 29(4): 396–404. doi: 10.16136/j.joel.2018.04.0205 曹洁, 郭雷雷. 一种基于局部冲突分配的证据组合规则[J]. 计算机应用研究, 2013, 30(7): 2033–2035. doi: 10.3969/j.issn.1001-3695.2013.07.029CAO Jie and GUO Leilei. Evidence combination rule based on local conflict distribution strategy[J]. Application Research of Computers, 2013, 30(7): 2033–2035. doi: 10.3969/j.issn.1001-3695.2013.07.029 彭颖, 胡增辉, 沈怀荣. 一种修正证据距离[J]. 电子与信息学报, 2013, 35(7): 1624–1629. doi: 10.3724/SP.J.1146.2012.01680PENG Ying, HU Zenghui, and SHEN Huairong. A modified distance of evidence[J]. Journal of Electronics &Information Technology, 2013, 35(7): 1624–1629. doi: 10.3724/SP.J.1146.2012.01680 毛艺帆, 张多林, 王路. 基于重合度的证据冲突度量方法[J]. 控制与决策, 2017, 32(2): 293–298. doi: 10.13195/j.kzyjc.2016.1394MAO Yifan, ZHANG Duolin, and WANG Lu. Measurement of evidence conflict based on overlapping degree[J]. Control and Decision, 2017, 32(2): 293–298. doi: 10.13195/j.kzyjc.2016.1394 汤潮, 蒋雯, 陈运东, 等. 新不确定度量下的冲突证据融[J]. 系统工程理论与实践, 2015, 35(9): 2394–2400. doi: 10.12011/1000-6788(2015)9-2394TANG Chao, JIANG Wen, CHEN Yundong, et al. Conflicting evidence combination based on a new measure method of uncertainty[J]. Systems Engineering -Theory &Practice, 2015, 35(9): 2394–2400. doi: 10.12011/1000-6788(2015)9-2394 BURGER T. Geometric views on conflicting mass functions: From distances to angles[J]. International Journal of Approximate Reasoning, 2016, 70: 36–50. doi: 10.1016/j.ijar.2015.12.006