基于随机森林算法在通风机故障检测中的应用  

Application of Algorithm Based on Random Forest in Fault Detection of Ventilator

作　　者：王海川 李蓉[2] 郭英飞 Wang Haichuan;Li Rong;Guo Yingfei(Changcun Colliery,Shanxi Lu’an Green Energy Development Co.,Ltd.,Changzhi,Shanxi 046102;Lu’an Vocational&Technical College,Changzhi,Shanxi 046204)

机构地区：[1]山西省潞安环保能源开发股份有限公司常村煤矿,山西长治046102 [2]潞安职业技术学院,山西长治046204

出　　处：《江西煤炭科技》2024年第4期183-186,共4页Jiangxi Coal Science & Technology

摘　　要：矿井通风机在煤矿安全生产中占据重要地位。其主要作用是为井下输送新鲜空气,稀释井下瓦斯及其他有毒有害气体浓度,并同时排出有害的粉尘和气体。矿井通风机属于长时工作负载,一旦设备出现故障无法正常运行会直接影响井下作业人员的安全,因此对通风机异常进行早期检测,开展预测性维护、防止非计划停机意义重大。针对上述问题,提出一种基于随机森林的通风机系统故障预警方法。该方法利用通风机系统可检测的传感器数据构建了基于随机森林的异常工况预警模型,可以在线监控设备状态,并在故障发生早期及时准确预警。Mining ventilator plays an important role of transporting ventilating air,diluting gas and other poisonous gas concentration and emitting hazardous dust in safety production in mines.Aiming at the possible abnormality in ventilator,the authors propose a fault alarm method to carry out the early detection and predictive maintenance,and prevent the unplanned shutdown,in which an abnormal working condition alarm model based on random forest is established through the detective sensor data of ventilator system,monitoring the equipment status on line,making an early alarm in case of fault and ensuring the safety production.

关 键 词：通风机 故障诊断 故障预警 随机森林 

分 类 号：TD635[矿业工程—矿山机电]