摘要： 多传感器数据融合在日益复杂的工业、军事等领域变得越来越重要，在实现方法上，神经网络由于其自身特点在该领域显示出广阔的应用前景。针对以往解决方案的不足，提出分散式的网络结构模型，详细描述了层次模型及各层次组成模块的工作原理及实现方法，然后提出高层规划层的概念，用于实现分散型的网络结构。最后，进行了仿真实验，用实际数据论述该算法。

关键词: 层次模型, 融合算法, 高层规划层

Abstract: Multisensor data fusion is getting more and more important in increasingly complicated fields of industry , military and so on.In the application, neural network(NN) shows a broad prospect in this field due to its features. In order to settle the deficiencies in former solutions, I hereby bring forward in this thesis the model of decentralized network structure. After describing in detail the work principle and realization method of level model and compositive modules in each level , I put forward the concept of high programming level to realize the decentralized network structure. In the end ,I demonstrate this algorithm with real data which I got from emulational expertiment.

中图分类号: 

• TP183