开放识别框架D-S合成规则研究

通过引入开放识别框架的概念,运用证据距离和反映证据冲突分配权重的证据冲突强度,提出了一种证据合成的有效方法,该方法能明确区分识别框架的完备性,判断是否需要开放原识别框架。最后,通过实例验证了该方法的合理性与有效性。     

基于冲突系数和pignistic概率距离的改进证据组合方法

针对单一使用冲突系数描述证据冲突存在的不足,从冲突系数和pignistic概率距离2个组合元素的角度出发,在冲突系数和pignistic概率距离形成的平面直角坐标系中,通过计算证据源对应的点到直线的距离表征证据之间的冲突,将冲突信息转换为支持信息,通过权重系数修正证据源,采用Dempster组合规则合成.通过算例分析和对比,证明了改进证据组合方法的可行性和优越性.     

D-S证据理论在目标识别中的应用

D-S证据理论的应用过程与事件密切相关,针对目标识别这一特定领城,研究了证据理论中基本概率赋值的获取、组合规则及决策规则等问题.识别实例中,利用灰关联分析法来处理基本概率赋值的问题,对多传感器基本概率赋值进行组合后,最终通过决策完成对2种车型的类型识别,经计算仿真表明,D-S证据理论在目标识别中具有一定的有效性和优越性.     

一种度量广义基本概率赋值冲突的方法

Dempster-Shafer(D-S)证据理论在组合冲突信息时,会得出与直觉相反的结论,为此研究人员提出很多改进方法,但是如何度量证据之间的冲突程度却常常被人们忽视.在分析现有冲突系数的基础上,在推广幂集空间框架下提出了一个冲突距离系数,用来刻画证据之间的冲突程度.进一步提出了合理的冲突度量参数应满足的判定准则,通过两种典型的悖论情况,分析了经典冲突系数和二元组冲突表示方法的不足,说明所提出的冲突系数不但满足判定准则,而且在判断冲突使用Dempster组合规则方面可以达到二元组表示方法的同等效果,甚至在判断冲突使用Dezert-Smarandache theory(DSmT)方面可以超过二元组表示方法的效果.     

一种基于“交并集”和Pignistic概率的证据冲突的改进方法

针对D-S证据理论无法解决高冲突证据的缺陷，通过对现有几种证据冲突的改进方法进行分析，提出了基于“交并集”和Pignistic概率的改进方法。本文放宽D-S组合规则的假设，只要求证据在组合时至少有一条是真实的，如果证据A和B相互支持，说明它们都是真实的，可以用“交集”运算将证据的信度聚焦在它们的芟集上；如果证据A和B相互冲突，表明不知道哪一条证据是真实的，则用“并集”运算将信度聚焦在它们的并集上，即证据支持A或B中的一个，这种思路更符合人类的直觉。由于在目标识别系统中，最终决策是单个待识目标，因此以还要用Pignistic概率转换法将多元素命题的BPA再分配给它的各个组合元素，最后，信度最高的元素作为结果进行输出。实验表明，本文方法在解决证据冲突方面较其他方法拥有明显的优势。使用本文方法时，证据的融合顺序对融合结果没有影响，因此可以很方便地编程实现。     

A new divergence measure for basic probability assignment and its applications in extremely uncertain environments

Information fusion under extremely uncertain environments is an important issue in pattern classification and decision-making problems. The Dempster-Shafer evidence theory (D-S theory) is more and more extensively applied in dealing with uncertain information. However, the results contrary to common sense are often obtained when combining different evidence using Dempster's combination rule. How to measure the difference between different evidence is still an open issue. In this paper, a new divergence is proposed based on the Kullback-Leibler divergence to measure the difference between different basic probability assignments (BPAs). Numerical examples are used to illustrate the computational process of the proposed divergence. Then, the similarity for different BPAs is also defined based on the proposed divergence. The basic knowledge about pattern recognition is introduced, and a new classification algorithm is presented using the proposed divergence and similarity under extremely uncertain environments. The effectiveness of the classification algorithm is illustrated by a small example handling robot sensing. The proposed method is motivated by the urgent need to develop intelligent systems, such as sensor-based data fusion manipulators, which are required to work in complicated, extremely uncertain environments. Sensory data satisfy the conditions (1) fragmentary and (2) collected from multiple levels of resolution.     

A generalization of frame perturbation in Hilbert subspaces and its application to wavelet subspaces

A perturbation theorem for frames in a Hilbert space is obtained, which is a generalization of a result by Christensen. Based on this general result, an irregular sampling theorem for frames in wavelet subspaces is established, which has a previous result as a special case.