Open AccessFault detection of pressure sensor of blast furnace fan based on Chaos Sparrow Search Algorithm-Support vector machine regression
Author(s) -
Ming Zhang,
Qiong Liu,
Jihong Deng,
Chun Chen
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2010/1/012029
Subject(s) - blast furnace , fault (geology) , support vector machine , fault detection and isolation , engineering , control theory (sociology) , residual , algorithm , computer science , artificial intelligence , actuator , control (management) , seismology , geology , chemistry , organic chemistry , electrical engineering
The sensor diagnosis system of the blast furnace axial fan is very important to the safety of the blast furnace fan control system. In view of the fact that the sensor fault detection of the blast furnace fan is not considered in the existing research, this paper proposes to use SVR to establish a fault detection model for the blast furnace fan outlet pressure sensor, and then use the improved chaotic sparrow algorithm to optimize the selection of the SVR penalty parameters and kernel function parameters Find the optimal parameters. By comparing the model prediction value and the residual error of the diagnostic sensor output value, the sensor fault diagnosis is realized. When the fault is judged, the model prediction value is used instead of the fault sensor output value to be used by the blast furnace fan control system to realize the fault-tolerant control of the blast furnace fan control system. The simulation results show that this method realizes the fault detection of the pressure sensor of the blast furnace fan and improves the safety of the blast furnace production.