Unlock instant, AI-driven research and patent intelligence for your innovation.

On-line Prediction Method of Caustic Ratio of Outlet Solution in Alumina Double-flow Dissolution Process

A caustic ratio and process export technology, which is applied in special data processing applications, instruments, electrical digital data processing, etc., can solve problems such as strong randomness, long training time, prediction model generalization ability and prediction accuracy need to be improved

Active Publication Date: 2017-02-15
CENT SOUTH UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, there is already a method for predicting the causticity ratio of the dissolution solution that combines the mechanism model with the neural network model. Both generalization ability and prediction accuracy need to be improved, making it difficult to apply this method effectively in the actual production process for a long time

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • On-line Prediction Method of Caustic Ratio of Outlet Solution in Alumina Double-flow Dissolution Process
  • On-line Prediction Method of Caustic Ratio of Outlet Solution in Alumina Double-flow Dissolution Process
  • On-line Prediction Method of Caustic Ratio of Outlet Solution in Alumina Double-flow Dissolution Process

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] Such as figure 1 As shown, the present invention realizes the prediction of the causticity ratio of the eluate at the exit of six flashes by using the mechanism model, the method of correcting the prediction value of the mechanism model based on fuzzy expert rules, and the error time series prediction model based on the least squares support vector machine. Here, 422 data from a factory’s continuous production are selected for case analysis, in which the first 300 groups are used for training, and the last 122 groups are used for prediction.

[0042] 1. Predicted value of caustic ratio mechanism model

[0043] In this embodiment, the main chemical reactions in the stripping process are the reaction of lye and alumina to generate sodium aluminate; the reaction of alumina and silicon oxide and lye to generate sodium aluminosilicate; the reaction of titanium oxide and calcium oxide to generate insoluble titanium calcium oxide; a part of calcium oxide reacts with alumina a...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an online prediction method for RSA (Ratio of Soda to Aluminate) of an outlet solution of a two-stream-process alumina dissolution process. The method comprises the following steps: (1) obtaining a mechanism model predicted value based on material balance; (2) amending the mechanism model predicted value based on fuzzy expert rules according to the amounts of change of temperature and pressure difference under current work conditions, so as to obtain an amended mechanism model predicted value; (3) establishing an error time series prediction model based on a least squares support vector machine according to a historical error time series formed by an RSA measured value and the amended mechanism model predicted value, obtaining an error predicted value of next moment by using the prediction model, and compensating the amended mechanism model predicted value by using error predicted values, so as to obtain a predicted value for the RSA of the solution. The method has the advantages that the relative error of predicted outlet solution RSA is smaller than + / -2%, and the maximum absolute error is smaller than 0.025.

Description

technical field [0001] The invention relates to an online prediction method for the caustic ratio in the dissolution process of alumina double-flow method. Background technique [0002] In the production of Bayer process alumina, high pressure dissolution is an extremely critical process. The causticity ratio of the eluate in the high-pressure dissolution process is a very important economic and technical index. It not only reflects the alkali consumption of alumina, but also has a great impact on the subsequent production of alumina. In order to improve the subsequent production conditions of alumina and increase the output and quality of products, the key is to be able to detect the caustic ratio on-line and optimize the ratio on this basis. However, at present, there is no instrument that can directly detect the caustic ratio, but it is calculated on the basis of chemical analysis of the raw ore pulp and the dissolved ore pulp. Since the chemical analysis takes a long ti...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F19/00
Inventor 谢永芳王晓丽尉思谜阳春华桂卫华谢森徐德刚王峰
Owner CENT SOUTH UNIV