.png){kind=icon}
.png){kind=icon}
Return
Continuous caster final electromagnetic stirrers position optimization using genetic programming
.png){kind=icon}
PRE
AI
DOI：10.1080/10426914.2023.2219317.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In January 2022 final electromagnetic stirrers were installed to improve billets' macrostructure and to reduce defects in STORE STEEL, Slovenia. From January to September 2022 the material from 1121 batches, out of 2284 batches produced, was examined using ultrasonic testing. Among these 1121 batches, 270 batches were detected with defects larger than 2.5 mm. Based on various metallurgical reports made from samples taken after an ultrasonic inspection, the root causes were found to be shrinkage porosity and segregations. Based on the gathered parameters from this period (chemical composition, casting parameters, cooling or heating parameters of billets), the occurrence of inner defects was predicted using logistic regression and genetic programming. The genetic programming proved to be better than logistic regression. Based on modeling results, optimization of the final electromagnetic stirrers position for 51CrV4 spring steel was conducted. Consequently, out of 41 examined batches, only 2 were detected with defects larger than 2.5 mm.
Keywords:
Continuous
caster
final
electromagnetic
stirrers
position
casting
parameters
ultrasound
examination scrap
rate
shrinkage
porosity
segregations
optimization
logistic
regression
genetic
programming
Journal
M
IF:
4.7
Papers:
4.6K
Citations:
9.3K