基于聚类算法的轨道交通履带式消防机器人的应用

Application of Crawler-type Firefighting Robot in Rail Transit Based on Clustering Algorithm

对履带式消防机器人的构型进行分析,简化其构型设计为四轮机器人,并建立相应的坐标系。通过对其运动模型进行规律分析,得出单侧履带上所有点的运动速度规律,完成了功能模块的设计,再根据功能模块的设置完成对机器人的硬件结构设计。将履带式消防机器人所采集视频图像中每个像素点的色彩值作为特征向量,将所有采集到的视频图像构成一个样本集合,把图像分割任务转换为数据集合的聚类任务,运用K-means聚类算法进行图像区域分类,获取所需的分离图像。

This paper analyzes the configuration of a crawler-type firefighting robot,simplifies its configuration and design as a four wheeled robot,and establishes the corresponding coordinate system.By analyzing the laws of its motion model,the velocity patterns for all points on the single track are obtained.The design of the functional module is completed,and the hardware structure of the robot is designed according to the settings of the functional module.The color values of each pixel in the video image collected by the crawler-typefirefighting robot are used as feature vectors,and all collected video images are formed into a sample set.The image segmentation task is transformed into a clustering task of the data set,and the K-means clustering algorithm is used for image region classification to obtain the required separated images.