基于神经网络的移动机器人路径规划研究  

Research for mobile robot path planning based on neural network

作　　者：王仲民[1] 姚立卿[1] 张寒松[1] 

机构地区：[1]天津职业技术师范学院,天津300222

出　　处：《天津职业技术师范学院学报》2003年第1期10-12,共3页Journal of Tianjin Vocational Technical Teachers'college

摘　　要：利用一种并列连接的神经网络结构对移动机器人路径进行规划,并把神经网络与模拟退火算法相结合,解决了局部极值问题,最终收敛到全局最优解。计算机仿真研究表明:模拟退火算法具有计算简单,初值鲁棒性强以及通用易实现等优点。A kind of paratactically - connected neural networks for Mobile Robot Path Planning is used. Combined with the Simulated Annealing, the local minimum problems are solved, and at last the globe optimal value is achieved. The computer simulation results show that the simulated annealing's computation is simple, the initial value's robustness is good and it can be used widely and easily.

关 键 词：神经网络 移动机器人 路径规划 退火算法 权值 计算机仿真 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]