【摘要】 提出了一种基于径向基函数 ( RBF)神经网络的数字地图压缩方法 ,给出了网络训练的具体算法。与一般 RBF网络的构造方法不同 ,该网络结构的所有参数是通过学习方式同时获得的 ,因此大大增强了其适应性。数字仿真表明该方法具有自动适应地形、参数配置合理、在线计算量小等特点更多还原

【Abstract】 This paper presents a RBF neural network model and the algorithm of network learning which are suitable for the compression of digital map. The construction of this model is different from that of ordinary ones in that it constitutes all network parameter synchronously via learning, hence the applicability of it is strengthened greatly. This method possesses several characteristics such as adaptation to terrain, rational parameter collocation and low calculation cost by simulations.更多还原