论文部分内容阅读
为了使老年人能够独立、安全、舒适地行走于户外环境,针对自行开发的智能助行机器人ZJU Walker,提出一种基于CMAC神经网络的在线学习控制算法。CMAC神经网络模拟人的小脑结构,具有学习速度快、结构简单等特点。采用CMAC神经网络对用户的行走习惯进行关联记忆和在线学习,并理解用户的行走意图。实验结果验证了所提方法的有效性,使助行机器人更加易于使用,提高了用户在使用过程中的舒适度。
In order to enable the elderly to independently, safely and comfortably walk in the outdoor environment, an online learning control algorithm based on the CMAC neural network is proposed for the self-developed intelligent walking robot ZJU Walker. CMAC neural network to simulate the human cerebellum structure, with learning speed, simple structure and so on. Using CMAC neural network to user’s walking habits associated memory and online learning, and understand the user’s intention to walk. The experimental results verify the effectiveness of the proposed method, make the robot more easy to use, and improve the comfort of the user during use.