论文部分内容阅读
提出一种用于移动机器人对当前感知环境进行识别和分类进而实现避障的方法。分析和研究反映当前感知环境的特征向量的抽取技术，并建立反映期望环境类别的特征向量训练样本数据库；建立基于ART2神经网络的分类器模型，根据多传感器信息实现移动机器人对其当前感知环境进行快速识别和分类；利用模糊逻辑技术，设计用于移动机器人在未知环境下避障的三维模糊控制器；在TIT-1型移动机器人上进行实验。
A method is proposed for mobile robots to recognize and classify the current perceived environment and then achieve obstacle avoidance. Analyze and research the extraction technology of eigenvectors that reflect current environment and establish eigenvector training sample database reflecting the desired environment category. Establish a classifier model based on ART2 neural network, and realize the current perception environment of mobile robot based on multi-sensor information Rapid identification and classification; the use of fuzzy logic technology designed for mobile robots in the unknown environment of the three-dimensional fuzzy controller; TIT-1-type mobile robot experiments.