摘要
针对空调箱的温度、流量等传感器进行固定偏差和漂移故障的检测与诊断,提出了一种基于统计学的方法进行在线的故障检测和诊断。主成分分析法根据系统正常的历史数据建立数理统计模型,通过传感器实际测量数据与正常数据在故障子空间投影的比较,对传感器的故障进行在线检测。联合角度法改进了传统的贡献图的诊断方法,通过对故障库中的故障知识的利用,能够在线的分离出发生故障的传感器。
A statistic method combining principal component analysis （PCA） with joint angle plot was proposed to online detect and diagnose fixed or drift bias of temperature and flow rate sensors of air handling unit （AHU）. PCA model was set up based on the normal history data of the system to detect the sensor faults by comparing the projection on the residual subspace to which the real measurement vector and normal vector were projected. Whereas, the joint angle method, which could improve the Q-statistic plots method, was used to isolate the fault source on line employing the history fault knowledge.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2006年第6期1653-1656,共4页
Journal of System Simulation
关键词
空调箱
传感器
固定偏差
漂移故障
主成分分析
联合角度法
Air handling unit
Sensor
Fixed bias
Drift bias
Principal component analysis
Joint angle method
