Clinker free calcium content soft measurement method based on time lag XGBOOST model

Abstract

The invention discloses a clinker free calcium content soft measurement method based on a time lag XGBOOST model. The method comprises the following steps: acquiring production data of corresponding production parameters of a cement production process and quality data of a cement sample in a laboratory; confirming the production data and the quality data as input data, confirming the free calcium oxide content as output data, and performing data cleaning on the input data and the output data; respectively calculating the Pearson correlation between each input data and the output data; determining the time advance of each input data based on the production time node of the input data in the cement production process, and carrying out time delay alignment on each input data and the output data; a target XGBOOST model with the minimum mean square error is determined; and when new production data is received, inputting the production data into the regression model to obtain a predicted value of the content of the free calcium oxide. According to the method, better prediction capability and lower prediction errors are achieved on the complex time-varying problem, and the time-varying characteristics can be captured more accurately.

Description

technical field [0001] The present application relates to the technical field of cement clinker production, in particular, to a method for soft measurement of free calcium content in clinker based on a time-lag XGBOOST model. Background technique [0002] In the production process of cement rotary kiln, the production process is complex, and the main production links include raw material crushing, raw meal grinding, coal grinding, firing (preheater, precalciner, grate cooler), waste heat power generation and clinker storage, etc. process. The core of the sintering process is to control the quality of the clinker, of which the content of free calcium in the clinker is the most representative indicator of the quality of the cement clinker. At present, there is no instrument for real-time measurement of free calcium content in cement rotary kilns, most of which are measured by manual sampling every two hours. Such manual sampling and testing methods have many limitations, suc...

AI Technical Summary

Problems solved by technology

At present, the cement rotary kiln does not have an instrument for real-time measurement of free calcium content, and most of them use manual sampling and laboratory measurement every two hours.

This method of manual sampling and testing has many limitations, such as unrepresentative sampling, discontinuity, delay in obtaining results, etc., and it is impossible to measure the free calcium content of clinker more accurately.

Although there are currently some free calcium prediction methods, the commonly used free calcium prediction methods are generally linear regression and deep neural network. relationship, resulting in inaccurate models; deep neural networks have high applicability in the field of image text, but due to the small amount of data in the industrial field and the great changes in data distribution, it is difficult to have a better prediction effect

In summary, there is currently no method that can accurately predict the free calcium content of clinker

Images