摘要详情

ID / 提交时间

32 / 2025-03-27 13:01:33

标题

Application of PLS methodology in fault diagnosis of industrial boilers

关键字

PLS,Multiple linear regression,Fault diagnosis,Boiler systems

主题及专题

T02 数据驱动的故障诊断

状态

全文待审

作者

红蛟王 / 青岛科技大学

亚星张 / 青岛科技大学

摘要

Partial Least Squares (PLS) is a dimensionality - reduction technique grounded in statistical principles. It extracts valuable information from process data and models the process. PLS not only reduces data dimensionality and extracts features but also accounts for the regression relationship between input and output data. In this study, the PLS method is applied to fault diagnosis. A regression model is constructed to elucidate the regression relationship between each independent variable (input) and the dependent variable (output). The PLS method is further employed to diagnose boiler faults. By determining whether the Q statistic and the

statistic exceed their respective control limits, the malfunction of the boiler system can be detected. Simulation results demonstrate that the PLS method can effectively detect faults with high accuracy in fault data recognition.