模糊神经网络在移动机器人信息融合中的应用  

Application of fuzzy neural networks in information fusion for mobile robot

作　　者：丁承君[1] 张明路[1] 段萍[1] 

机构地区：[1]河北工业大学机械学院,天津300130

出　　处：《控制理论与应用》2004年第1期59-62,共4页Control Theory & Applications

基　　金：河北省自然科学基金项目(501031);天津市自然科学基金项目(003601211).

摘　　要：针对移动机器人所用的传感器,提出了一种用于多传感器信息融合的方法,将模糊逻辑和神经网络结合起来,构建了模糊神经网络,并建立了网络的计算模型.通过建立的模糊神经网络对移动机器人的多传感器信息进行融合,实现了移动机器人对动态环境中障碍和环境类型的实时识别以及无冲突运动.网络的训练和试验表明该方法在移动机器人躲避运动物体中是可行的.An information fusion method of multi-sensor was proposed according to sensors used in mobile robots. The fuzzy logic and neural network were combined to construct fuzzy neural networks (FNN), and the calculating models of fuzzy neural networks were also put up. The mobile robot could recognize moving obstacles and environment types online, and could decide its action by using the information fusion method put forward, and thus realize mobile robot's movement without collision in dynamic environment. The training of network and test results indicated that the method which was used in avoiding moving obstacles of mobile robot is practicable and effective.

关 键 词：移动机器人 信息融合 模糊神经网络 模糊推理 学习算法 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TP183[自动化与计算机技术—控制科学与工程]