Regarding coke quality prediction!
Thread Content
The title of my master’s thesis is Prediction of Coke Quality under Dynamic Coal Blending, and my supervisor asked me to study the linear relationships between G, Y, V values of coal and the cold and hot strength of coke. I would like to ask everyone: besides using direct linear relationships, what other mathematical methods can be employed to build prediction models?
Take a look at this: http://bbs.hcbbs.com/thread-363919-1-1.html
And also this: http://bbs.hcbbs.com/viewthread.php?tid=73253&highlight=%D6%CA%C1%BF%D4%A4%B2%E2
Linear regression should be done! Establish a regression equation based on the various coal blending indicators and the cold strength data of coke!
Use a large number of experiments to conduct regression analysis. In fact, these things are not very useful in practice. However, the potential of coal chemical industry workers can be tapped. In fact, we make predictions every day. It is based on a large number of industrial ratios.
I’m wondering if some new mathematical methods can be used to build models; this way, the models resulting from such approaches might have a higher degree of correlation than those based on ordinary linear prediction, and they would also be more practical.