基于粗糙集的矿井通风机故障诊断系统的研究  

The study of fault diagnosis system of mine ventilator based on rough set

作　　者：李春华[1] 石瑶[1] 赵杰[1] 

机构地区：[1]黑龙江科技学院电气与信息学院,哈尔滨150027

出　　处：《自动化与仪器仪表》2010年第6期30-32,34,共4页Automation & Instrumentation

基　　金：黑龙江省教育厅科学技术研究面上项目(编号:11541304)

摘　　要：针对矿井通风机故障诊断中数据量大和属性多的特点,采用粗糙集中基于改进的可辨识矩阵和逻辑运算的属性约简算法,对矿井通风机的故障特征进行属性约简。仿真结果表明:采用粗糙集方法对故障数据处理后,使训练样本数量明显减少,训练误差减小,训练步数与训练时间也相当减少,从而提高了矿井通风机故障诊断的效率和准确性。Aimed at the large amount of data and the more attributes in mine ventilator fault diagnosis,this paper use the attribute reduction algorithm based on the improved discernibility matrix and logic operation of rough set,and applies it to the fault characteristics of mine ventilator attribute reduction.Simulation results show that： after using the rough set to deal with failure data,made the number of samples of training reduced significantly,training error decreases,the training steps and the training time considerably reduced,thereby increasing the efficiency and accuracy of mine ventilator fault diagnosis.

关 键 词：矿井通风机 故障诊断 粗糙集理论 可辨识矩阵和逻辑运算 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]