DS理论在信息融合中的改进

Improvement of D-S theory in an information fusion system

多传感器数据的信息融合有许多算法,常用的有D S证据理论法,这是一种用于处理不确定性的方法。研究了证据理论组合规则的原理,通过深入分析,针对它的不足提出了一种改进组合方法,这样不仅能够组合冲突度大的证据,而且能根据各条证据所包含的不同信息量进行自适应加权组合。最后通过几个实例证明了该方法的有效性,提高了合成结果的可靠性。

There are many algorithms for multi-sensor information fusion. D-S evidence theory is an useful method to deal with uncertainty problems. The principle of the evidential combination formula is studied, and an improved combination method is put forward according to the shortage of the formula through in-depth analysis. Thus we can not merely make up conflicted evidence like this with different on-line weights, but also carry out adaptive weighting association according to the amount of information that an evidence includes. Some examples have been done to demonstrate that the new method is efficient and can improve the reliability of the combination results.