A calculation method for determining coke properties based on the properties of a single type of coal.
Thread Content
Many calculations are not very accurate.
The formulas proposed to the original poster are empirical equations developed by various factories based on their own coal usage patterns and through mathematical statistics; they can be used for qualitative analysis. However, for quantitative analysis, they certainly won’t be suitable for predicting the quality of coal and coke in your factory. Therefore, to obtain a suitable calculation method, you need to put in effort to collect data yourself and derive corresponding empirical equations
I agree with what was said on the second floor. However, to establish a calculation method for determining coke indicators based on single-type coal indicators, it is necessary to assume that the source and properties of the coal used in your factory are relatively stable; only then can the model developed be applicable!
A constant should be an empirical value. For example: Coke ash content = 1.3 * ash content of coal fed into the furnace ； Coke sulfur content = 0.85~0.9 of the sulfur content in the coal fed into the furnace. The M25 and M10 are a bit more complex, and their accuracy is also lower.
Constant values are obtained through multiple regression analysis of experimental data! The more experimental data there is, the more reliable this constant becomes!
Regarding the Wulong knife, I suggest you read Mr. Kun Jiao Zhang Guoqing’s article titled \"Using Multiple Linear Regression to Predict Coke Quality Based on Blended Coal Parameters.\" In it, he states that M40 = 51.16 + 0.534G – 0.247Ad – 0.238Vdaf.
I can’t remember which issue of ‘Fuel & Chemical Engineering’ it was; I’ll search online.
There is a software for multiple regression called SPSS, which is designed specifically for performing multiple linear regression; you might want to download it and give it a try.
It must be referring to this article. I don’t have it, but you can download it from VIP Information or China Journal Network. 【Title】Prediction of Coke Quality Using Multiple Linear Regression 【Authors】Zhang Guoqing, Zhang Qiongfang, Zhang Zhongqiong 【Institution】Kunming Coking and Gas Production Plant 【Journal Name】Kunsteel Science & Technology, 2006(2): 43-46 【Keywords】Multiple linear regression; Coal blending parameters; Prediction of coke quality 【Abstract】By using the statistical tool of multiple linear regression, and taking into account the influence of conventional coal blending parameters such as Ad, Vdaf, and GRI on key coke quality indicators like M40, CRI, and CSR, a mathematical model was established linking these coal blending parameters to the key coke quality indicators: y0=bi＋∑biXi. Theoretical verification showed that the established mathematical model is meaningful ； The predicted values were then compared with the actual test values from production, and the two matched well, verifying the accuracy of the model through practical production. Using multiple linear regression techniques to predict coke quality is useful for guiding production.