摘要
提出一种采用雷达图形实现设备运行状态的分类显示,应用模糊推理对雷达图形进行自动分类判别的方法,并采用主成分分析对数据进行降维处理以减少误判率和模糊推理的复杂度。对减速箱轴承运行状态模式的仿真结果表明,降维后的雷达图仍然能够合理表示故障状态。将数据回代判别,模糊推理的识别结果完全正确。
A novel method for recognizing different running states of equipment was proposed, which used radar chart expression to display and applied fuzzy rules to implement automatic classification. In order to decrease the rate of fault alarms and the complication of fuzzy reasoning, principal component analysis (PCA) was employed to reduce dimensionality. The experimental results of decelerating box data indicate after dimensionality reduction, the radar chart expression may exactly distinguish all states, results of fuzzy reasoning are correct.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2006年第z3期2098-2100,共3页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金项目
教育部博士学科点专项科研基金项目
关键词
状态识别
雷达图
模糊推理
主成分分析
降维
state recognition radar chart expression fuzzy reasoning PCA dimensionality reduction
