摘要
D-S证据理论是一种性能优越的信息融合方法,由于各传感器所提供的证据的重要程度不同,需要对各证据进行加权合成处理。目前的加权D-S证据理论限于合成规则的研究,较少讨论如何获取优化的证据权值。实际上,证据权值的确定是证据进行加权合成的基础和关键。针对加权证据理论的这一研究不足,提出了一种求取最佳证据权值的方法。首先,阐述了思想,再建立了优化模型;然后,改进了粒子群优化求解算法,利用其优越的求解非线性多峰值函数的能力,求解出了最优权值。通过实例仿真表明:这种证据理论的加权算法是有效的,与对比方法相比,具有更好的融合效果。
The D-S evidence theory is an excellent method of information fusion. Because of the difference which is caused by the sensors,it is essential to deal with the evidence with a method of weighed D-S theory Presently, the study of the weighed D-S theory is confined to synthetic methods, and the study on how to obtain better weight value is also rarely covered. In fact, the definition of the weight values is fundamental to the process of weightsynthesizing the evidence. Considering the disadvantages of the weighed D-S theory, a best method of obtaining evidence weigtit value is presented. After elaborating the thoughts of the new method, an excellent weight value is provided by establishing an optimized model and improving the particle swarm algorithm. Compared with the comparison methods, this evidence theory proves more effective by making simulation test.
出处
《传感器与微系统》
CSCD
北大核心
2007年第1期84-86,共3页
Transducer and Microsystem Technologies
基金
全国优秀博士学位论文作者专项资金资助项目(200237)
关键词
证据理论
权值
优化
粒子群
evidence theory
weight value
optimize
particle swarm