Aiming at the issue of target fusion identification in information fusion technology, the improved method of multi-sensor information self-adaptive fusion with weighted fusion center is proposed. The overall fusion structure is built first;then the data fusion process is combined with evidence theory and neural network, by adopting online learning of neural network to self-adaptively adjust the fusion weights; and the decision is made by evidence theory finally. The method using multi-sensor self-adaptation possesses better fault tolerance capability and higher accuracy of detection and identification, thus the flexibility and intelligence of data fusion are enhanced. The result of simulation proves the effectiveness of this method.%针对信息融合技术中目标融合识别的问题，提出了带有加权融合中心的多传感器信息自适应融合的改进方法。首先构建整体融合结构，将数据的融合过程与证据理论神经网络进行有机的结合，利用神经网络的在线学习、自适应调节融合权值、证据理论进行决策。这种采用多传感器自适应的方法具有较强的容错性、较高的检测与识别精度，增强了数据融合的灵活性和智能性。仿真结果证明了该方法的有效性。
展开▼
