基于结构模式识别的托辊状态与失效检验算法

An Algorithm for Verifying State and Failure of Coal Conveyor Roller in Thermal Power Corridor Based on Structural Pattern Recognition

在火力发电中为保证巡检机器人的有效实施,需要对廊道输煤传送带关键部件托辊状态监控提供理论和技术支持。根据巡检机器人的性能要求、运行方式和传送带工况,提出和采用基于结构模式识别,通过CCD探测、约定扫描方式、状态矢量表达、分类器设计和聚类范数算法,对传送带托辊状态、中心位移、失转、严重磨损和裂纹等进行在线监测和预警,实现机器人实时状态监控与异常情况的报警。

In order to ensure the effective implementation of replacing manual inspection by robot in thermal power generation,it is necessary to provide theoretical and technical support for the state monitoring of key parts of coal conveyor in corridor.According to the requirements of the inspection robot space orbit performance,operation mode and working condition of the conveyor,the cluster norm algorithm was put forward with the machine vision of CCD,specific scanning mode,the state vector expression,and classifier design,based on the structural pattern recognition,to inspect displacement of the center,loss in revolution,serious wear and visible crack,of the conveyor supporting roller,realizing the real-time status monitoring and abnormal alarm of the roller.