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

Multi-point prediction method of roller kiln temperature field based on deep learning

A technology of deep learning and prediction methods, applied in neural learning methods, biological neural network models, design optimization/simulation, etc., can solve the lack of multi-point prediction of ceramic roller kiln sintering temperature field space, and the slow time of temperature prediction data. , without considering the roller kiln and other issues, to achieve timely and accurate global control, improve model accuracy, and ensure the effect of accuracy

Active Publication Date: 2022-06-21
GUANGDONG UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The present invention overcomes that the prediction model of the roller kiln in the above prior art does not consider the spatial multi-point prediction of the sintering temperature field of the ceramic roller kiln, and does not consider the influence of the historical data of the roller kiln on the temperature, and the time of the temperature prediction data is slow. , low efficiency, and provide a multi-point prediction method for roller kiln temperature field based on deep learning

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
  • Multi-point prediction method of roller kiln temperature field based on deep learning
  • Multi-point prediction method of roller kiln temperature field based on deep learning
  • Multi-point prediction method of roller kiln temperature field based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] like Figure 1 to Figure 5 As shown, the multi-point prediction method of roller kiln temperature field based on deep learning includes the following steps:

[0043] S1: Use the roller kiln sensor to collect the input parameters in the roller kiln and the real sintering temperature in the roller kiln;

[0044] S2: According to the boundary value of the input parameters collected in step S1, use the simulation software to simulate the temperature of the roller kiln, use the real sintering temperature collected in step S1 to verify the accuracy of the simulation software, and output the position that cannot be measured by the sensor. The simulated sintering temperature of the point;

[0045] S3: Define the input data set according to the input parameters and the simulated sintering temperature, and the input data set includes the input parameter matrix X i and the input temperature matrix Y j , where the input parameter matrix X i There are i input parameter vectors i...

Embodiment 2

[0070] like Figure 2 to Figure 5 As shown, the multi-point prediction method of roller kiln temperature field based on deep learning includes the following steps:

[0071] S10: collect data and preprocess the data, including step S101 and step S102;

[0072] S101: Collecting data: the collected input variables include the gas inlet velocity of the upper and lower spray guns of each section of the roller kiln firing section; the temperature before gas combustion; the total gas pressure; the secondary gas pressure; Flow rate; air pressure of combustion main pipe; pressure before spray gun; temperature of combustion air before combustion; frequency and current of primary exhaust fan; frequency and current of secondary exhaust fan. The temperature of the roller kiln was simulated and verified by Fluent software. Take 8 points in the single-section geometric space of the simulation as output variables, and collect them every 10 seconds. Real-time measurement values ​​are collec...

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 multi-point prediction method of the roller kiln temperature field based on deep learning includes the following steps: using the roller kiln sensor to collect various input parameters in the roller kiln and the actual sintering temperature in the roller kiln; according to the collected input parameters, use the simulation software Simulate the temperature of the roller kiln, and at the same time output the simulated sintering temperature of the position that the sensor cannot measure; define the input data set according to the input parameters and simulated sintering temperature, which includes the input parameter matrix and the input temperature matrix, and based on Transformer, establish a roller In the multi-point prediction model of kiln temperature field, the encoder encodes the input parameter matrix, and the output matrix of the last encoder is converted into a query matrix key matrix and used as the input of each decoder. In the decoding stage, the input of the decoding component includes The input temperature matrix is ​​decoded by multiple decoders, and finally the multi-point temperature of the sintering section of the roller kiln at the current time point is output. The invention can improve the quality of ceramic products and reduce product energy consumption.

Description

technical field [0001] The invention relates to the technical field of roller kiln smelting, and more particularly, to a multi-point prediction method of roller kiln temperature field based on deep learning. Background technique [0002] The roller kiln has the advantages of uniform temperature, short firing time, fuel saving and high degree of automation, and is widely used in the ceramic production industry. Roller kiln is a typical continuous kiln, which consists of preheating section, firing section and cooling section, and each section is divided into several sections. The ceramic blanks are placed on the rollers, and the blanks are preheated, fired, and cooled sequentially through the continuous rotation of the rods. The ceramic firing mechanism is: first, the ceramic blanks are preheated and dried and crystals decomposed in the preheating section; several burners are arranged in units of nodes in the firing section, and gas fuels such as natural gas or coke oven gas ...

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): G06F30/20G06N3/08G06F119/08
CPCG06F30/20G06N3/08G06F2119/08
Inventor 杨海东王中琰徐康康朱成就印四华金熹王弦楷匡先云余炳圳黄梓伟
Owner GUANGDONG UNIV OF TECH