能量修正和决策集约简的证据理论优化算法

为了同时利用证据焦元的基本概率赋值和焦元的基数信息解决DS(Dempster Shafer, DS)证据理论高冲突问题,并考虑数据融合的抗噪声和干扰能力,提出采用基于焦元信息能量的先验信息比值演变函数对证据源和组合规则进行修正的算法。为了进一步降低〖JP2〗算法计算量,利用多元素焦元信任值向单元素焦元分配以及决策集约简方法,进一步优化算法性能,提高算法大数据量适应能力。实验结果验证了优化算法的正确性和有效性。

Optimization algorithm for DS evidence theory based on energy modification and decision set reduction

In order to improve the anti-noise or anti-jamming performance and to solve the problem of high conflict by using the basic probability assignment and the cardinality of focal elements, the evolution function of prior information ratio for focal elements’ energy is introduced to modify the evidence source and the combination rule of Dempster-Shafer(DS) evidence theory. For reducing the computation complexity, the trust value assignment of the multiple focal elements to single focal elements and the reduction of the decision set are presented to optimize the adaptability of the modification algorithm for big data. The experiment results prove that the optimization algorithm is accurate and effective.