[关键词]
[摘要]
根据D-S理论在证据完全对立的情况下失效的原因,考虑证据间基本可信度分配函数(BPA)值在证据中所占比重以及相关程度两个方面对证据合成的影响,在D-S理论框架基础上构建了一种新的多信息融合算法——比亲算法。通过与两篇D-S理论应用论文中的实例进行比较表明:在证据不相对立的情况下,新算法与-S理论应用的结果是一致的;当证据对立时,新算法则具有纠错功能而优于D-S理论。结果验证了这种算法的可行性、有效性及实用性。
[Key word]
[Abstract]
Based on the reasoning that D-S evidence beco me s invalid when evidences are contradictory, a new multi-sensor data fusion meth od is proposed. The data fusion under the influence of proportion and correlatio n of BPA's numerical value is considered. Representative examples from two paper s are used to test the feasibility and availability of this new method.
[中图分类号]
TN958
[基金项目]