基于改进型BP网络的自动寻路设计与实现  

Design and implementation of automatic path-finding based on improved BP neural network

作　　者：陈滔滔[1] 李坤和[1] 杜晓荣[1] 

机构地区：[1]中山大学光电材料与技术国家重点实验室珠海实验室,广东珠海519082

出　　处：《计算机工程与设计》2013年第11期3988-3995,共8页Computer Engineering and Design

基　　金：2010年广东省大学生创新实验基金项目(1055810023);广东省省部产学研重大专项基金项目(2009A090100025)

摘　　要：针对静态环境下机器人的自动寻路问题,进行了研究与探索,设计了一类改进型BP神经网络。该网络在结构上以按功能分类的方式取代传统的分层模式,在权值的设置方面参照了组合逻辑电路的基本原理;分析了时序逻辑电路的信息记忆过程,从而进一步赋予神经网络良好的记忆特性;提出一种基于人类寻路思想的路径规划方法。实验结果表明,通过将该寻路算法融入到改进的BP网络中,不仅使寻路的实现变得简单和方便,而且具有较高的实时性与适用性。The automatic path-finding problem of robot in static environment is studied and explored, a kind of improved BP neural network is designed. In this network, the traditional hierarchical model is replaced by a functional classification model, and the weights are set with reference to the basic principle of combinational logic. At the same time, the information memory process of sequential logic is analyzed so as to give the neural network fine memory characteristic. In addition, a path planning method based on the idea that how humans find ways is proposed. The result shows that the implementation of path-finding not only becomes simple and convenient, but also has high real-time performance and usability when putting this algorithm into the improved BP neural network.

关 键 词：自动寻路 BP神经网络 组合电路 时序电路 记忆特性 路径规划 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]